Skip to main content

Fast Company

A user told Meta’s Muse to sell his keyboard. What happened next was alarming

4 weeks 2 days ago

Matt Robb wanted to sell his computer keyboard.

So he made a listing on Facebook Marketplace — and, as an experiment late last month, he put Meta’s new artificial intelligence agent Muse in the driver’s seat.

“I kind of went into it like a guinea pig,” said Robb, a Toronto-based content creator who regularly reviews consumer tech. He had seen Muse’s plush toy-looking avatar pop up on Instagram and decided to see if the AI agent could spare him some online bargaining.

Muse got to work. But within a day, Robb got an alarming update: A buyer he didn’t know had shown up outside his house late at night. How? Muse had struck a deal on its own and shared Robb’s home address after agreeing to a pickup. It even sent an automated “Yep, I’m here!” message from his account when the buyer arrived — all without Robb’s knowledge.

“Not only did it give my address when the guy asked and it accepted the offer, but it literally was almost like teasing the guy (about me being there when I wasn’t),” said Robb, who also detailed the experience in his now widely-shared social media posts. “That really baffled me.”

Many have welcomed AI agents into their online lives, and raved about how personalized assistants like Muse can handle everyday tasks with ease — from snagging a dinner reservation or planning a trip, to organizing email inboxes. But like other AI bots, the agents are only as good as the information they have access to. They perform better the more they know about a user — and perhaps also that person’s friends, family and even colleagues. That raises privacy concerns.

Robb and the buyer eventually made amends, and Meta reached out to talk through what went wrong. But the episode showed real-world risks that can arise from sharing extensive personal information with the technology.

“Using an AI agent is similar to handing someone your keys and a to-do list,” said Veronica Rivera, assistant professor at Georgia Tech’s School of Cybersecurity and Privacy. “How helpful they are, and by extension how risky, depends on which keys you hand over.”

Companies have varied privacy policies for AI agents

Since its September launch, downloads of Meta’s Muse app have skyrocketed. And competition is only growing. OpenAI recently rolled out its own lineup of “always on” AI helpers, called dots, which can be integrated into ChatGPT and workplace platforms like Slack and Microsoft Teams.

Older generative AI chatbots, like the standard version of ChatGPT, respond to specific prompts. At the user’s request, they can write essays, generate new images or music, or even write lines of code.

Meta’s Muse and OpenAI’s dots go a step further to perform day-to-day tasks on a person’s behalf. AI agents are powered by the same large language models as chatbots, but they run in a loop. That allows an agent to take proactive steps, even when the user isn’t looking.

“One of the things that makes these AI agents really appealing is that they are unprecedentedly useful,” said Rivera, who is also a Microsoft-funded Computing Research Association Trustworthy AI Research Fellow.

But the convenience comes with the price of sharing more details about your life.

Once AI agents obtain users’ information, it’s very difficult to fully withdraw access, said Ebrahim Bagheri, a professor at the University of Toronto’s Faculty of Information.

“If you provide access to your credit card to an agent, the agent now immediately has access to all of the historical transactions … they will immediately — within less than seconds — ingest all of that data,” Bagheri said. And while users can later withdraw access, Bagheri said there’s “no taking back” data the agent has already seen, and how developers store historical data can be murky.

Both Meta and OpenAI outline privacy policies for their new AI agents, which operate on their own dedicated cloud computers. You can turn off memory sharing between dots and ChatGPT, but information the agent previously received isn’t automatically deleted. OpenAI gives users an option to request access to, delete or correct personal information, although rights can vary depending on where you live.

Meta’s Muse gives users options to delete message history and other information, too. For day-to-day use, the agent prompts users for access permissions — and Meta has said Muse learns about a user and who they interact with “based on public information and from what you’ve chosen to share.” But the company says that all stays secure for each user.

But the tech giant also notes that Muse can still “make mistakes or take unexpected actions.”

Risks of mishaps rise with AI agents’ level of access

In Robb’s case, he learned the disclosure of his address came down to “always allow” permissions he granted when he first let Muse on his Facebook Marketplace profile, where he had shared his pickup address. But he warned others could similarly misunderstand the extent of that access — and said he initially thought the AI agent would check with him before sharing that information or accepting offers, regardless.

His AI agent didn’t follow the rest of his instructions without error, either. For another item Robb was selling, Muse seemingly suggested he would accept a price $100 below the previously-set minimum. When reached by The Associated Press, Meta acknowledged the agent’s error without offering further comment.

Robb later largely disconnected Muse from his apps. In an interview earlier this month, he said it could’ve been a far worse situation, especially “if some stranger turned up to someone more vulnerable’s house.” Still, Robb said he expects these agents to be fine-tuned and he would consider re-enabling Muse in the future. He’s since been testing another AI agent — Grok Bot from Elon Musk’s X, which launched in beta in August.

Other tech companies have become wary of the influx of AI agents. Earlier this month, Apple said it would introduce new controls for its Mac operating system in response to AI agents’ broader access of apps like mail, messages and even browsing history without users’ “full knowledge and understanding.”

“As AI agents become increasingly capable and autonomous, the risks associated with this level of access will grow substantially,” Apple wrote in an Oct. 2 update.

Apple didn’t specify any developers or agents by name, but the update arrived after Inc. Magazine columnist Jason Aten reported that Muse watched notifications on his Mac — including Messages — without him giving the agent that specific access.

Meta spokesperson Andy Stone has maintained such integration is “entirely opt-in,” writing on social media last month that users would have to enable both the Apple security feature “Full Disk Access” and a Messages connector through Muse. Aten said he didn’t grant the agent those permissions.

Even those who do not join the AI agent boom aren’t necessarily shielded. Consumers who don’t download an AI agent might still interact with others who do. And beyond potential confusion when AI sends messages on behalf of a user, experts warn of ripple effects when it comes to privacy.

“When your agent reads your inbox or your texts, it’s also reading what your friends and family sent you, even if they never explicitly agreed to that,” Rivera said.

—Wyatte Grantham-Philips, AP Business Writer

Stefan Sagmeister’s new shoe designs are actually walking infographics

4 weeks 2 days ago

Data visualization turns charts, data, and statistics into something easier to comprehend. Turning that data viz into shoes makes the data into a functional conversation starter.

That’s what the Austrian graphic designer Stefan Sagmeister did with two shoes he designed for Ekn, a German footwear brand, that are on view at his exhibition Finally, Something Good at the Paris gallery Gaîté Lyrique through April 2027.

[Photo: © Marek Bäuerlein]

Sagmeister says the exhibition is meant to look at the world over the long term rather than through daily news. Each shoe uses data to tell a story: one about renewable energy, and the other about homelessness. The key to understanding what the shoes say is written in the footbed.

[Photo: © Marek Bäuerlein]

The Bring It On, a multicolored sneaker about renewable electricity, uses data from the Energy Institute’s 2024 Statistical Review of World Energy to inform the relative sizes of color panels on its sides, with each leather panel representing a different energy source. Yellow, the largest color, represents hydropower, followed by black for wind, pink for solar, and red for all other renewables.

“Wonderfully, the shoe is already out of date,” Sagmeister tells Fast Company. “In the 2025 figures, published this summer, solar passed wind for the first time, so today the pink panel should be slightly bigger than the black one.”

[Photo: © Felix Schöppner, Stephan Braubach, Mirko Müller]

His other shoe, Like a Complete Unknown, is a boot that uses circles to represent the relative homeless populations of six countries. The area of each circular appliqué corresponds to the number of people who are living in the streets and public spaces per 100,000 inhabitants, according to OECD data. The biggest circles are for the U.S., at 76, and Germany, at 46, followed by the U.K. (16), France (8), Mexico (5), and Japan (2).

[Photo: © Marek Bäuerlein]

“Countries count differently, so the numbers are not perfect, but Mexico, a much poorer country, reports a fifteenth of the American rate, and a gap that size is not only a counting difference,” he says. Sagmeister volunteered for the Coalition for the Homeless, an advocacy and service organization, and says what stayed with him from that experience was why some countries handle the same problem so much better than others.

[Photo: © Marek Bäuerlein]

The shoes are available for purchase on Ekn’s website. Both retail for more than $300, of which 10% goes either to the climate innovation think tank Future Cleantech Architects or the Coalition for the Homeless in New York, respectively. Sagmeister sees the inherent awkwardness between homelessness data and a boot costing 350 Euros, and says a donation “doesn’t make that go away.” He hopes people will take away that neither the data behind renewable energy nor homelessness is fate.

“When the number of people sleeping in the street runs from two to 76 per 100,000 across six countries,” he says, that shows “it’s about decisions, and decisions can be made differently.”

This startup just tested turning an underwater volcano into a power plant

4 weeks 2 days ago

Last month, a startup called Endurance Energy sailed 300 miles off the coast of Oregon to test something new: deploying geothermal power at an underwater volcano.

In a vent field around 5,000 feet underwater, they drilled a two-inch-wide, 52-foot-deep well that tapped into 500-degree heat beneath the ocean floor—the first fully encased well of its kind. Later, as 10-foot waves crashed around their ship, they lowered a three-ton generator to the sea floor to test its ability to create electricity from that heat.

[Photo: Endurance Energy]

The full process didn’t work exactly as planned, and the experiment stopped when a power cable failed. But when the team connected the generator to a natural vent in the sea floor, the generator briefly produced power. (For 8.7 seconds, it produced more power than it consumed.) The company says that it’s an important proof point for its ultimate vision: a network of undersea power plants that could provide baseload clean energy.

While tiny experiments happened in the past, that vision is new. “No one’s ever tried this even close to the commercial scale that we’re going after,” says Andrew Redd, founder and CEO of Endurance Energy.

Redd previously worked as an engineer at SpaceX. “I wanted to go after something that was going to be both really good for the world but also extremely compelling from an employee perspective,” he says. “I looked around me and saw all of the hyper-capable people at SpaceX and posed the question, how do you get people like this to work for you?”

The challenge of undersea power was incredibly complex, and had never been done before—the type of problem that engineers love to work on. It also had the potential, he realized, to be a major power source.

[Photo: Endurance Energy]

The Seattle-based startup, which has raised $63 million to date, launched at the beginning of 2025. Already, it’s had five undersea deployments, with the test in September the most complete, deploying the end-to-end steps in the company’s process.

The prototype system, called Adélie, is now back at the company’s Seattle lab, where it’s being tweaked and prepared to relaunch in a few months. It’s still too early to know whether the technology is truly viable. But Redd argues that the approach has advantages. On land, the best sites for geothermal power have permeable rock, hot temperatures, and access to water; all of those things are easily found offshore.

“You’ve got this vast, permeable network on the mid-ocean ridges,” he says. “You’ve obviously got unlimited water for your reservoirs. And then the crustal thickness is very low, so you’ve got this thin, oceanic crust that gives you a really high geothermal gradient.”

There are other questions, including environmental impact. The company, which has been working with the University of Washington in its early tests of prototypes, says that it’s carefully monitoring environmental impact and argues that it’s minimal (the sound of the drilling, for example, is quieter than the ship’s engine.) Still, it’s harder to know what the impact might be over time, and typical move-fast-and-break-things startup culture isn’t aligned with caution. It also isn’t clear what challenges the company might face with permitting when it scales up to full-sized commercial systems.

[Photo: Endurance Energy]

In theory, if it works, it could be a large power resource. Around the Pacific’s Ring of Fire, Redd says that there are around six terawatts of power that could be developed over the next five years—assuming, and it’s a big assumption, that the equipment was available to build it. That’s more power than current global demand.

“Where it’s cost-effective, where there are resources near to shore, we’ll be able to support a significant amount of the upcoming demand for baseload power, and then also start to displace existing power generation from non-renewable resources,” Redd says. “That’s really the mission of the company: to make offshore geothermal the economically obvious choice.”

Endurance Energy is building a modular system designed to be replicated in a factory to get to economies of scale. “That gets you into the realm of solar panel production, where it’s the same thing being built over and over again,” he says.

Now, as it retools the Adélie prototype, the startup is also developing its commercial-size, 2-megawatt generator. The first project is planned in the Kingdom of Tonga, where much of the grid runs on imported diesel. If all goes as planned, the new power plant could be running next year.

Trump posts plans to close this ‘community jewel’ in Washington, D.C.

4 weeks 2 days ago

President Donald Trump on Sunday announced plans to close a historic public golf course in the nation’s capital for 15 months and redevelop it despite a court challenge to his long-sought renovation.

Trump posted on social media about his intent to shutter the East Potomac Golf Course, not far from the White House, starting Nov. 15 to make it “one of the Best and Most Spectacular Golf Courses anywhere in the World.”

A judge last week, however, ordered a halt to his redevelopment plans until at least Oct. 26 after scolding the government for starting a large-scale tree removal at the course over a holiday weekend while providing little notice to the court or the plaintiffs trying to stop the makeover.

Trump did not acknowledge the legal fight in his post, which included an artist’s rendering for the course redesign.

“When completed, it will be open to the Public while, at the same time, designed to the Highest Level of Golf and able to hold World Class Tournaments such as the U.S. Open, the PGA Championship, the Ryder Cup, and the Presidents Cup,” Trump wrote.

Democracy Forward, which is representing the plaintiffs challenging his renovation plans, responded by saying that the course is part of a national park and recreation facility that is meant for everyone to enjoy.

“Our litigation is intended to protect this community jewel for the people and communities it serves and we will continue to work with our plaintiffs to defend it from unlawful and harmful intrusion,” Skye Perryman, the president and CEO of the group, said in a statement.

The judge in the case took notice of Trump’s online message and ordered the government to notify her by 8 p.m. EDT Sunday whether its position had changed from a court filing Friday that said “at most” Trump has given “preliminary thought” to the rebuilding plan and “there is no final agency action.”

The government responded by repeating that “there is no final agency action” because the National Park Service has not completed its review and the plan is not final. The filing, signed by Associate Attorney General Stanley Woodward, also said the government, in stating this position, had not been representing “the actual position or mental impressions” of Trump but instead had relied on his public statements.

Capital makeover extends well beyond the fairways

Trump, a golf aficionado who plays often, had already announced plans to makeover the links as part of his broader push to renovate and refurbish places around Washington, including the White House and the Kennedy Center for the Performing Arts. Many of his projects, including plans for a huge ballroom at the White House, the installation of his own name on the Kennedy Center and construction of a massive arch, have been challenged in courts.

The 106-year-old park and course has been the subject of many emergency hearings as the president has sought to renovate it. While it has been open to Washingtonians of all income levels, opponents of Trump’s plans to turn it into a high-end course have argued it will alter the character of the space and make it less affordable to people.

Trump in his post echoed arguments he has made to justify his plans to remake the Kennedy Center, arguing it is in such dangerous condition that it needs to close.

“The current Course is old, broken, dilapidated, and of no further use or effect. It is in very poor condition, with falling, dangerous, and disease infested trees and, importantly, a collapsing and hazardous seawall,” Trump wrote.

Debris from the demolition of the White House East Wing to make way for his ballroom project has also been dumped on the grounds. The National Park Service has said the debris tested positive for lead, chromium and other toxic metals.

Trump also sought to justify the National Park Service’s removal of hundreds of trees that the judge paused, writing in his announcement that when the golf course was originally built, “it was famous for no trees, with views of Washington, D.C., for all to see, but now, it is completely overgrown, dysfunctional, and unsafe.”

Legal case returns to court and the administration seeks to remove the judge

Trump’s post outlining a clear plan for construction complicates a federal court hearing scheduled for Monday in a lawsuit filed by the DC Preservation League and two golfers opposing the construction.

U.S. District Judge Ana Reyes has ordered key members of the Republican administration to testify on the recent tree-cutting and whether those actions are tied to the president’s plans for the historical public links.

The administration on Friday asked for Reyes to be replaced on the case, questioning her impartiality and citing what it called “egregious” behavior from the bench in response to an emergency hearing Oct. 12. Her frustration boiled over in a three-hour hearing when she lambasted the government’s attorney for the lack of notification she and the plaintiffs received before the trees were chopped down.

At the upcoming hearing, the judge is expected to address the request that she step aside. Meanwhile, over the weekend the plaintiffs asked that she grant them a more long-standing preliminary injunction in the case and/or a stay.

Associated Press writer Gary Fields contributed to this report.

—Michelle L. Price, Associated Press

New Elizabeth Holmes documentary scores thousands of sellout showings on its opening weekend

4 weeks 2 days ago

North American movie theaters were flooded with new films this weekend, including a video game reboot, a big budget survival thriller and a Jane Austen adaptation.

While Paramount’s “Street Fighter” took the first-place spot, earning $31 million from 3,575 theaters according to studio estimates on Sunday, most of the attention and cultural discourse was elsewhere: On the nearly three-hour Elizabeth Holmes documentary, “You Can See Everything.”

‘You Can See Everything’ had thousands of sellouts

Buzz and box office don’t always line up, but “You Can See Everything” made an extraordinary $15 million from only 855 screens, with over 2,000 sellout showings over the weekend, according to A24. It also had the biggest opening for a film playing on under 1,000 screens since “Borat” opened to $26.5 million in 2006, and the biggest nonmusic documentary opening in two decades, since “Fahrenheit 9/11” made nearly $24 million from 868 theaters. Those numbers do not account for inflation.

Like other hits this year, like the Gen-Z driven “Backrooms” and “Obsession,” ticket sales for “You Can See Everything” were driven by a largely younger audience. Over 70% of the opening weekend crowd was under 35.

The film, directed by Nathan Fielder and Lance Oppenheim, was made in secret over three years. Fielder spent over a month with the disgraced Theranos founder and her family, including two young children and partner Billy Evans, in her final days before reporting to federal prison. The reviews have been extremely positive, with 99% on Rotten Tomatoes and 99 on Metacritic. It also got an unprecedented 91% “definite recommend” from audiences, according to PostTrak exit polls.

“You Can See Everything” reportedly cost under $7 million to make and less than $10 million to market.

“Everybody underestimated the movie, because on paper this is like, wait, a three-hour documentary?” said Paul Dergarabedian, the head of marketplace trends for Rentrak. “But he (Fielder) is a huge draw, and he has a huge following on social media, and then Elizabeth Holmes is just a character that you couldn’t even make up. It’s stranger than fiction, and that’s really compelling.”

‘Street Fighter’ knocks out ‘Whalefall’

“Street Fighter” topped the charts as expected with a $51 million global opening, including $20 million from international showings. The movie starring Callina Liang, Noah Centineo and Andrew Koji cost a reported $82 million to make, not accounting for marketing and promotion. Audiences were 69% male.

Both “Street Fighter” and “You Can See Everything” may have contributed to “Whalefall’s” relative failure this weekend. The $60 million survival thriller starring Austin Abrams and Josh Brolin made $8 million from 3,065 locations in the U.S. and Canada, despite being better reviewed than “Street Fighter.” With $9.5 million from international showings, the Disney release had a $17.5 million global launch.

“Whalefall” opened in fourth place, behind the Jessica Chastain horror “Other Mommy.” The Universal Pictures release added $8.4 million in its second weekend, bringing its domestic total to $44.9 million.

Fifth place went to “Sense and Sensibility,” which earned an estimated $7.3 million from 2,545 theaters. The well-reviewed Austen adaptation starring Daisy Edgar-Jones drew a predominantly younger (58% between the ages of 14 and 18) and female crowd (84%). It also got the strongest CinemaScore of the week (A-).

Warner Bros. Clockwork also released a restored presentation of Ken Russell’s 1971 film “The Devils” in 350 locations, where it made $1.6 million. It’s the first release from the studio’s specialty label.

‘Musk’ documentary facing advertising blocks and Netflix shares box office for ‘La Bola Negra’

Alex Gibney’s Elon Musk documentary “Musk” also expanded to 472 screens in the U.S. and Canada, where it made $306,702. The film, which runs nearly four hours, has faced issues advertising on major social media platforms including X, which Musk owns, and TikTok, according to distributor Bleecker Street, which said TikTok this weekend permanently suspended the film’s ad account.

Netflix also shared ticket sales for the first time for the queer Spanish epic “La Bola Negra,” which made $277,500 from four locations this weekend. The streamer has not historically reported box office for its theatrical releases, but this film is breaking new ground as it’s being positioned as a possible best picture contender. It will be in theaters for 47 days before it will be available to watch on Netflix.

According to Rentrak, the weekend overall generated about $101.6 million in ticket sales, up 45% from the same weekend last year. The year to date stands at $8.2 billion, which is up almost 20% from 2025.

Top 10 movies by domestic box office

With final domestic figures being released Monday, this list factors in the estimated ticket sales for Friday through Sunday at U.S. and Canadian theaters, according to Rentrak:

  1. “Street Fighter,” $31 million.
  2. “You Can See Everything,” $15 million.
  3. “Other Mommy,” $8.4 million.
  4. “Whalefall,” $8 million.
  5. “Sense and Sensibility,” $7.3 million.
  6. “Verity,” $6.2 million.
  7. “Resident Evil,” $5.5 million.
  8. “Heart of the Beast,” $4.8 million.
  9. “Jailer 2,” $3.2 million.
  10. “Forgotten Island,” $2.3 million.

—Lindsey Bahr, AP Film Writer

Nationwide meat recall underway as ‘temperature abuse’ impacts 1.5 million pounds of beef, pork, and chicken

4 weeks 2 days ago

A massive recall involving more than 1.5 million pounds of meat product was announced this weekend after the manufacturer discovered what the U.S. Department of Agriculture’s Food Safety and Inspection Service (FSIS) deemed to be possible “temperature abuse.”

Here’s what that means, what meat products are affected, and what you should do if you have any of the recalled items.

What’s happened?

On Saturday, the USDA’s Food Safety and Inspection Service alerted the public about a voluntary recall by Kayem Foods Inc., a more-than-a-century-old meat products company based in Chelsea, Massachusetts.

The recall involves more than 1.5 million pounds of meat products produced by Kayem Foods. According to the FSIS, Kayem is recalling 1,584,361 pounds of ready-to-eat (RTE) meat and poultry products over fears that impacted products “may have experienced temperature abuse during storage.”

The problem was discovered after one of Kayem Foods’ retail customers complained about warm product temperatures related to some of the meat products, the notice states.

Kayem then investigated the claims and found one of the warehouse coolers where the products were stored “experienced temperature deviations,” according to the FSIS.

What is “temperature abuse”?

While the term “temperature abuse” makes it sound like someone knowingly altered the meats’ temperature, the situation is actually less nefarious.

In relation to this recall, “temperature abuse” simply means that the products’ temperatures deviated from what they should have been stored at, potentially making the products unsafe to eat.

In a recall notice posted directly on Kayem Foods’ website, the company says that the issue was due to “a cooling system malfunction at one of the company’s warehouses that resulted in storage temperature deviations.” 

Announcing the recall, Kayem CEO Matt Monkiewicz said, “Our business depends on providing consumers with safe food of the highest quality. Out of an abundance of caution, we are undertaking this voluntary recall.”

He added: “We take food safety very seriously and apologize to our valued consumers. Our management team takes full responsibility. We have learned from and corrected the issue and will not allow this to happen again.” 

What products are included in the recall?

The more than 1.5 million pounds of meat products impacted by the recall include items sold under nine different brands:

  • Al Fresco
  • Deutschmacher
  • Essem
  • Genoa
  • Jordan’s
  • Kayem
  • McKenzie
  • Triple M
  • Winnisimmet

The products include numerous ready-to-eat items, such as sausages, chicken, pork, bacon, and other products. 

The FSIS says the recalled products bear the establishment numbers “EST. M/P 7839,” “EST. P 20935A” and “EST. M 9520” inside the USDA mark of inspection.

Items included in the recall “were produced approximately March 31, 2026, to Oct. 7, 2026.” Kayem Foods says the items were “stored and shipped between June 7, 2026, and October 14, 2026.”

The FSIS has published a full list of the items on its website. The agency has also published photos of the product packaging.

Kayem Foods has published its own list of recalled items here.

Where were the recalled products sold?

Neither the FSIS nor Kayem Foods has announced which retailers sold the recalled products. However, the FSIS says that the “items were shipped to retail and foodservice locations Nationwide and Hong Kong.”

The agency says that, as of the time of its notification, there have been no confirmed reports of anyone becoming ill from eating the recalled items.

What should I do if I have the recalled products?

The FSIS says consumers should check their refrigerators, as the items could still reside there even if purchased long ago. 

Consumers who have the recalled items are being urged not to consume them, but instead dispose of them or return them to their place of purchase.

Those with questions about the recall can contact Kayem Foods Customer Service at 800-426-6100 or email the company at info@kayem.com.

How to grow your most valuable asset in the age of AI

4 weeks 2 days ago

A brand is still a company’s most valuable asset. It’s an intangible collection of signs, symbols, colors, behaviors, and voice, which never shows up cleanly on a balance sheet. You build it cumulatively, and when it finally comes together, it drives real business value. It’s why consumers pay more for a certain phone, sneaker, or pill. It’s why people choose you, again and again.

Until recently, people built this asset. Founders, business leaders, marketing teams, and the people on the front line interacting with customers shaped what the brand stood for and how it showed up.

AI is rapidly changing this relationship.

Increasingly, people don’t meet your brand directly. They meet an AI’s version of it through an LLM’s summary, an agent’s comparison, a generated image, or an answer that shows up before your website loads. According to recent research from R/GA’s forthcoming Agentic Commerce report, 64% of consumers have used AI to research a product or service in the past three months. A brand used to live almost entirely in what a company said and made. Today, AI agents increasingly influence how brands are represented and how consumers experience them.

THE SOUL IS MISSING

The marketing industry’s first instinct has been to treat AI as a production opportunity. Large holding companies and creative infrastructure players have spent the last couple of years plugging brands into generative pipelines to churn out assets, driving gains in output and efficiency that, when you look at the bottom line alone, might look like success.

But when you look at the outputs, there’s a problem. The experiences and content created are just okay. They are technically on-brand, but also completely forgettable. In a recent study, we found about 60% of consumers say a brand’s use of low-quality AI-generated content would negatively affect their opinion of that brand and likelihood to purchase from them, and this was most pronounced with Gen Z.

We’re creating more, but not necessarily better brand assets. Experiences that might clear an automated compliance check give nobody a reason to pick you over your competitor. The marketing industry prioritized scale over substance and soul, and consumers can feel it.

The natural response from organizations has been to add more guardrails through tighter prompts, stricter review, and better filters on the pipeline, but this doesn’t always translate to better outcomes. They’re trying to fine-tune existing systems instead of asking what else is needed. Creating agentic systems that can deliver quality outcomes and add value to your greatest asset gives people a reason to choose you first.

DESIGN FOR DUAL AUDIENCES

At R/GA, we’ve found better outcomes start with defining and designing brands for humans and agents. That means designing brands to be machine-readable and writable, with designers shaping the systems that create experiences as a creative expression. It goes beyond a brand’s typeface or color palette to define what context matters most and training agents to make decisions in keeping with the brand.

We’ve been calling this approach Brand DNA or BRDNA. It’s a simple metaphor. Just as our DNA shapes who we are and how we behave, a brand’s distinct and salient traits become living code that guides people, machines, and experiences.

This approach also centers on designers and strategists actively training agents and deploying them within an organization’s existing creative workflows and tech stack. The goal is to create agents that have enough context about what the brand values, what it prioritizes, and what it would never do, so that they can make a judgment call the way a good creative director would. This connects character and code, scales what is most valuable about the brand, and helps protect it. All of this serves to make sure the experience itself is worth choosing.

We’ve been building and fine-tuning this approach with our partners. For example, for G42 we built a new type of AI-enabled adaptive interface that replaced a traditional website. When visitors land on their homepage, an AI agent greets users and orchestrates the experience in real time around their specific questions or needs, reshaping content and format for each.

We applied this same approach in a very different way for FOX One to help them deepen engagement around the World Cup. For millions of viewers, fandom often ends the moment their national team crashes out of the tournament. We created an AI-powered experience to pair fans with new teams based on their interests and behaviors. The agent became an active part of the experience, creating personalized content streams and the agent even watched games alongside fans, delivering real-time context, cultural insight, and commentary that turns casual viewers into committed supporters. The agent made decisions and acted on behalf of the brand, all steered by the brand’s unique DNA.

Both examples point to the same shift. Brands need to be machine-readable and writable, with agents crafted to make decisions that reflect the brand’s integrity and create experiences that feel additive to your brand’s assets.

We believe this new approach is necessary to help brands get ahead. Not by AI-ifying your workflows for efficiency’s sake, but by ensuring your greatest asset is elevated by technology, not diminished by it.

Melissa Jackson Parsey is the global chief strategy officer at R/GA.

America can manufacture at cost-parity with China

4 weeks 2 days ago

In a recent NYSE interview, a reporter asked me if my focus on Made-in-America manufacturing makes me afraid of having to compete with the cost of China manufacturing goods. My response surprised her.

“Not at all,” I said. “As a matter of fact, it is highly possible for Made-in-America pricing to be at parity or even lower than the cost of China-made goods.”

Here’s why: Automation addresses an old competitive disadvantage while simultaneously creating a new advantage.

THE NEW ADVANTAGE

America has spent decades assuming that manufacturing migrated overseas because American labor was too expensive. That assumption was once true, but automation has changed the equation. When labor becomes a smaller percentage of unit cost, the issues of location, transportation, inventory, quality, supply-chain resilience, speed, and intellectual-property control (all of which we can control) become much more important.

But we face a conundrum: We still need people, but automation changes the nature of the people we need.

That includes the humans who manage and control the automation. It also includes the electricians, welders, engineers, and other skilled trades that form the base of the manufacturing pyramid. The need for these roles is greater than ever before.

Consider the current statistics:

U.S. industrial robot installations increased 11% in 2025 to roughly 38,000. The Bureau of Labor Statistics (BLS) projects nearly 1 million production-job openings annually from 2024 through 2034, largely because workers retire or leave occupations. BLS also expects industrial machinery mechanics—the people who maintain increasingly automated plants—to be one of manufacturing’s strongest growth occupations. 

AUTOMATION PLUS HUMAN INFRASTRUCTURE

There is an interesting paradox at play. America is becoming more automated, but China is not standing still. China installed 295,000 industrial robots in 2024, versus about 34,000 in the U.S., and China now has at least 2 million deployed. So, automation alone will not give America a competitive advantage. What automation does is give U.S. manufacturers the ability to neutralize disadvantages that once appeared structural, but only if we deploy it aggressively and build sufficient human infrastructure around it.

The process works like this: Historically, the assumption has been that if labor costs are dramatically lower somewhere else, that’s where manufacturing belongs. But automation compresses the labor component of unit economics. Now the calculation bypasses wages and moves to the total cost of production.

Executives should understand that a lights-out factory is not a people-free factory. Someone must design, integrate, power, program, repair, and continuously improve the factory to ensure its success. The result:

  • Specialized workers are a new constraint: Due to automation, nearshoring, and other key factors, American manufacturing may outrun its supply of electricians, engineers, technicians, welders, and other specialized workers.
  • Skilled trades as critical infrastructure: U.S. vulnerability lies in treating skilled trades as yesterday’s economy when, in fact, they’re critical infrastructure for the AI/robotics economy.

What to do?Companies thinking about reshoring today should evaluate automation + talent + supply chain simultaneously, rather than first deciding where to manufacture and then figuring out automation later.

THE FACTORY OF THE FUTURE

This country needs differently skilledworkers to fuel our next growth era. This requires continuing education and training of today’s workforce to prepare as many as possible for increasingly lucrative roles.

Tomorrow’s winner will be the system that converts raw materials into finished goods at the lowest risk-adjusted total cost, not simply the factory with the cheapest labor.

Parity with China’s manufacturing costs will come from automation, talent, and total-cost considerations. That is a competition America can enter on very different terms than it could 20 years ago. If we address this well, and I believe we can, the issue of parity with China’s manufacturing prices will resolve itself.

Matthew Chang is founder and principal engineer of Chang Robotics.

Stop sanding down your brand

4 weeks 2 days ago

We’ve all seen it. Sometimes it’s subtle. Sometimes it’s three competitors using the same word in 48-point type.

An entire category starts competing with itself instead of competing for the customer. The claim wars begin. More this. Less that. Faster, cleaner, smarter, better. Cruise control kicks in. Everyone benchmarks one another until they start looking like one another.

Data must become the adult in the room.

DATA AND COMFORT

I like data. It shows us what people clicked, what they bought, and what they said in a focus group while eating a stale sandwich. It helps us make smarter decisions and avoid expensive mistakes.

But data is much better at protecting a decision than creating distinction. If every brand is reading the same reports and staring at the same dashboard, why would anyone expect a different answer?

Data can lead to comfort, a feeling that is great when you’re taking a break on a beach somewhere. In a category or a boardroom, comfort gets expensive.

The danger is that comfortable decisions typically produce safe campaign language, with images and graphics that are an evolution, not a revolution. Maybe they even move the numbers a bit. But that doesn’t create memory.

A campaign can rack up millions of impressions without making one. A brand line can be read on a billboard and never repeated in a group chat, across the dinner table, or at a bar. Awareness puts your name in front of people. The soul is what sticks.

So, do we fire the data? No. We just stop letting it run the meeting.

Data should inform and challenge conviction, not dictate it. Leaders still have to look past the consensus, spot what others miss, and make a call.

At Six+One, we use “truth, tension, and transformation” to find a brand’s soul. Alliteration always makes things feel more official.


START WITH THE TRUTH

Most brand exercises begin with what the company sells, what it can claim, and what it can prove. Truth starts somewhere less comfortable.

What does the brand actually believe? What does the customer really feel? What is the category refusing to admit?

This goes deeper than a positioning statement. Customers can tell when a brand is hiding behind copy. Honesty builds trust and loyalty, and that is worth more than a momentary click.

When we worked with Jovē, we could have joined the bottled-water claim wars. pH. Purity. Process. Source. All true and all useful. It was also where every other water brand was standing.

The audience didn’t need another science class in the beverage aisle. They wanted a brand that understood them and the role hydration played in how they wanted to look, feel, and live. That led us to “The water that gets you.”

The product facts didn’t disappear. They just stopped talking.

FIND THE TENSION

Every category has a repeat button in messaging that gets tossed around. With Jovē, water brands were busy talking about themselves: where they came from, which celebrity drank them, or the latest packaging gimmick. The person standing in the aisle had a much simpler question: What does any of this have to do with me?

That gap gave us our way in. We stopped trying to make Jovē sound like a more impressive bottle of water and built the brand around the person holding it.

This is where brands often lose their nerve. They pull out the sandpaper, soften the edges, and fall back on data-approved words like premium, better for you, authentic, and customer-first. The words anybody can approve are usually the words nobody remembers.

Tension gives a brand a point of view and distinction. For Jovē, that led us to hydrationships: a water brand that wants to go deeper down to the cellular level, and we made it comical, romantic, and with fewer empty promises. The brand stopped acting like another bottle on the shelf and started understanding the person holding it.

BUILD THE TRANSFORMATION

Truth and tension can still die in a strategy deck. Transformation gets them into the real world.

It isn’t a makeover. It carries the point of view into the campaign, content, retail experience, service, and culture. The pieces can flex. They just need to feel like they came from the same place.

Once we landed on “The water that gets you,” we used it to judge everything that followed. We invited people into a deep, healthy hydrationship with their water brand. The idea could flex for cities with “The water that gets Miami,” and retailers with “The water that gets Publix.”

You can usually feel when an idea has some bite. The room shifts, but not everyone is on board right away. That’s often a good sign. The job is to keep the idea intact through research, reviews, and, yes, edit number 17.

Some of the strongest brands barely raise their voices. You still know who is speaking. We must protect the brand at all costs so that by round seven, the sharp thing everyone loved doesn’t turn into a spoon.

There will always be a request for another proof point, focus group, or example of a competitor doing it first. Use the information. Learn from it. But don’t ask research for permission to lead. A dashboard tells you what has happened, but brand leaders should decide what should happen next.

Comfort can be tricky to spot, but you see the cost in media dollars used to force a forgettable message, discounts that make up for weak preference, and customers who know the category but can’t remember your name.

Find the truth. Name the tension. Build the transformation. Then hold the line.

Because the most expensive place for a brand to live is where every other brand feels comfortable.

Eric Rojas is founder and chief creative officer of Six+One.

Brands should lead with meaning instead of product details

4 weeks 2 days ago

Walk through New York City this fall and count the advertisements not selling artificial intelligence. Subway cars, bus shelters, billboards: Every surface pitches a model, an agent, a muse. And the images themselves are increasingly generated by AI rather than photographed. Every panel I have sat on this year has asked the same questions: How much AI should go into the work, and where?

My brand consultancy firm, the bread and butter, spent months this year on a project where those questions never came up. That absence is the most interesting thing about the work.

NOT ALL CUSTOMERS CARE ABOUT AI

The client is a premium eyeglass lens manufacturer with an exceptional product, measurably the best in its category. For its entire history, the company communicated that fact through specifications: refractive index figures, precision measurements, and comparative charts.

The people who buy premium lenses are not waiting to hear whether a brand uses AI. They want to understand what makes the product better and worth paying for, in language that respects their intelligence without demanding their research time. Call it sober authority: the confidence to state little and be believed anyway.

A DIFFICULT TRANSITION

Our proposal mapped personas of high-end consumers, then designed the brand they aspired to: the Hermès of lenses. Hermès sells handbags, but its imagery sells a fairy tale luxury that touches longing instead of listing features. Our concept, “profound vision,” treated the lens as a way of seeing rather than a product to describe. The photography showed what lies beyond it—details visible only to those with the sensibility to notice them.

So, we did not photograph the product. We photographed stone, close enough that the grain records how long each plane took to form, and a leaf’s ridge, close enough to see how the line was made. No claim, no specification. The message was simple: This company pays the kind of attention you do not have time to pay, and that attention is what you are buying. Work this subtle is difficult to make, and harder to approve.

The greatest challenge is the middle phase. When a company stops leading with specifications and starts leading with meaning, the old language switches off before the new one takes hold. Sales soften. Someone in a quarterly review asks why the numbers dipped. The honest answer: The dip is the price of the transition, and few organizations hold their nerve through it. Most reach for the specification sheet, because specifications are safe, defensible, and legible to a sales team.

WHY MOST BRANDS DON’T MAKE THE JUMP

Our client has not made the jump yet. The work was received well, parts were produced, and then the specification sheet came back. The people who would have carried the decision are measured quarterly on a change that pays back over years. That is unfair to ask of anyone, and our proposal did not make the intervening period survivable enough. We showed them where to land. We did not build the bridge.

Ask a chairman which brand they want to become, and the most common answer is Apple. Apple became Apple by challenging convention, a bet that could only be placed at the board level. So I now set one condition before creative work begins: The chief executive must be in the room where decisions are argued, not in a review at the end.

A rebrand changes what a company says while it is selling. That makes it a revenue decision, not a marketing decision. No marketing director can authorize a decision whose downside lands on this quarter and whose upside arrives after they have been reassigned. Ask them to, and they will produce safe marketing—the old version with better photography. With the chief executive in the room, the dip becomes a shared decision rather than one person’s risk, and everyone below stops hedging.

THREE CONDITIONS PRIOR TO LEAPING

Prior to beginning this process, consider these conditions. The first is heritage, translated for the present. No matter the company’s age, the history should still speak. The task is to find the choice from the company’s past that still explains what it stands for—a standard the company set, a shortcut it refused, a process it kept when cheaper options appeared—and to say it plainly.

The second is that the brand has to stand in the direction where it is pointing. If the people running a premium brand know their customer only from research decks, the work will not land. Taste is not a set of data points, and you cannot direct a photographer toward a world you have never entered. The brands that manage it put someone who lives in that world in charge, or give them time and budget to learn it first. The person leading the work has to know luxury from the inside.

The third is elasticity. A brand has to appear in a dozen places that behave nothing alike. The instinct is to protect the core by repeating it identically everywhere. The better structure holds the core fixed and lets the storylines stretch. What must never move is the reason the brand exists. Movement occurs in its storyline form. Most brand systems have this backwards.

A final irony. The product language these brands are trying to leave behind is what machines now generate at no cost, in volume. In April 2025 Ahrefs analyzed 900,000 newly created web pages and found that 74% contained AI-generated content. The advertisements covering New York City sell the ability to produce more of it. When every company can publish unlimited product descriptions, those descriptions stop being the differentiator. Meaning is the remaining move a machine cannot make on your behalf.

Sooyoung Cho is the founder and CEO of the bread and butter.

Companies waiting on AI will fall behind

4 weeks 2 days ago

I don’t believe that every company will survive the transition to artificial intelligence.

Companies that learn how to use AI will move faster, make better decisions, and produce more with the same resources. Those that wait will create a gap with competitors so wide that it may become impossible to close.

AI adoption is here. Are leaders ready to help their people start using it?

The transition has been one of the hardest changes I’ve led at Geotab. Reshaping a company around a complete rethinking of what business is, is not easy. It changes the way people work, how teams are organized, and how leaders measure performance.

Avoiding that disruption is a decision to fall behind.

How we deployed AI to 3,400 employees in one day

We rolled out a custom AI (called GIA) to 3,400 Geotab employees in one day. I believed we would learn more and get further ahead by putting GIA at people’s fingertips immediately rather than designing a perfect rollout.

Every person, from technical and non-technical teams alike, had full access to a secure AI gateway with pre-loaded skills right from their laptops. Within seven days, we had a 97% daily active rate.

Here are six key lessons we learned about getting employees on board with AI.

1. Be the first adopter. 

    Leaders cannot ask employees to embrace a technology they have not taken the time to understand themselves. I have followed and tested new technology throughout my career. Using AI personally helps me speak honestly about what it can do, where it falls short, and how it can improve the way we work.

    2. Make the first step simple.

      Onboarding is the highest friction point of the process in any technology rollout. If the first experience is too complicated, many people will not get far enough to discover the value. In our case, employees could simply type “GIA” and begin. The easier it was to start, the faster people could move from uncertainty to experimentation.

      3. Communicate clearly and continuously.

        Leaders need to communicate regularly, share concrete examples, celebrate useful experiments, and create opportunities for employees to learn from one another. Our monthly town halls provide an opportunity for employees to hear about and share how they are using AI in their work. I also led a hackathon with over 2,000 participants to help employees build their skills and discover new ways to work more effectively with AI.

        4. Measure output, not usage.

          AI usage is not the same as business value. It is easy to measure log-ins, prompts, tokens, or the number of employees who opened the tool. But are teams delivering more value with the same resources?

          We make usage information available to all employees immediately, including token consumption. Transparency helps people understand how they use AI and encourages responsible, purposeful adoption.

          5. Set up guardrails.

            Adoption requires privacy, security, data protection, and human judgment. Guardrails should make innovation safer, not impossible. Employees must know what information can be used, what outputs must be reviewed, and when human expertise must take over.

            The companies that succeed will understand the risk and establish sensible boundaries. They will keep learning.

            6. Do it before you feel ready.

              The biggest mistake leaders can make is waiting for AI adoption to feel neat and comfortable.

              It is messy. Let the first version be imperfect. Rapid implementation is not optional with technology moving at lightning speed. The companies adopting AI now will outpace their competitors and build the confidence, skills, and adaptability required for a quickly transforming future.

              Neil Cawse is founder and CEO of Geotab.

              Recognition is a talent strategy

              4 weeks 2 days ago

              When organizations talk about attracting and retaining talent, the conversation generally revolves around compensation, benefits, and workplace culture. Yet one of the most overlooked leadership practices is visibility: ensuring the people behind great work get recognized for their contributions. Recognition in this sense isn’t an empty accolade or “participation trophy,” but is about how strong organizations build trust, engage teams, and create environments where talented people want to stay and give their best.

              As employers, we spend enormous amounts of time and money recruiting exceptional people, only to too often hide the very expertise we worked so hard to attract. At the same time, our communities increasingly want to know who’s behind the work. They trust people more than polished corporate messaging, and they’re looking for transparency into the ideas and expertise that shape the brands they support.

              Leaders often see building a strong internal team as a separate challenge from building external trust. In reality, they are deeply connected. Visibility is the key to strengthening both.

              The strongest teams are built by leaders who consistently create opportunities for their people to be seen, heard, and recognized. Visibility is a leadership practice that builds confidence, strengthens collaboration, and signals to both prospective and legacy talent that great people get to do great work here. This is what it looks like in practice.

              THE BEST IDEAS HAVE FACES

              Early in my career, I noticed that the people doing the most influential work were often the least visible inside, and certainly outside, their organizations. We celebrate innovation, but often overlook the people responsible for it. The experts shaping products, solving complex problems, managing change, and driving innovations often remain behind the scenes, while only a handful of executives or chosen “cool kids” become the public face of the work.

              One reason I chose to leave the fashion industry in 2007, after eight years at its creative pinnacle, was the growing emphasis on celebrity talent. While so many brands and luxury houses were delivering beautiful work, as an industry, we were increasingly relying on highly paid Hollywood-esque fame to sell our goods and ideas. I feared that the result was a diminishment of the craft and an acceleration of hype and triviality.

              Visibility isn’t about turning everyone into an influencer, nor does everyone need to become a public spokesperson. What visibility can do is create appropriate opportunities to share expertise, whether that’s speaking at an industry event, publishing an article, presenting to clients, mentoring colleagues, appearing as a spokesperson, or contributing to conversations that move a business and industry forward.

              Visibility gives a reason to retain the people already on your team and attract the kind of people you want to hire next. Exceptional talent is always drawn to environments where expertise is recognized and where the people behind the work are trusted to build, grow, and share it. When leaders make those individuals and teams visible, they celebrate great work while signaling that this is a place where contributions count.

              TREAT AWARDS AS MORE THAN MARKETING

              Through my work with the iF DESIGN AWARD these past three years, I’ve seen that recognition is most powerful when it goes beyond celebrating finished products. It should also acknowledge the people and effort behind them. The best award programs recognize years of curiosity, collaboration, persistence, and experimentation that make meaningful innovation possible.

              Each year at iF Design, we offer our 140 international jurors two “wildcards” for projects they think are “wow!” These are essentially a “pass go, collect $200” kind of opportunity, where the wildcard recipient bypasses our preliminary jury round, saving €400 and ensuring placement as a finalist (top 50% of global entries). In the past few years, I’ve seen multiple Gold winners (less than 0.4% of all entries) emerge from wildcard recipients. Notably, even a few are still grad students—very unusual! To say that this changes their career trajectories is a major understatement.

              Moments like these reinforce that recognition creates momentum. Externally, it signals to prospective employees that this is a place where great ideas are valued and where talented people are given the opportunity to thrive. Internally, it provides a level of validation that managers and colleagues alone cannot. It tells people their work has meaning beyond business goals, that their ideas are contributing something valuable to their industry and, in many cases, to society more broadly. And, of course, at the highest leadership levels, this kind of recognition also begets more resources!

              VISIBILITY CREATES A VIRTUOUS CYCLE

              There is a natural connection between visibility, retention, and recruitment. When people see their colleagues recognized, they see evidence that good work has a platform. When employees themselves are recognized, they gain a stronger sense of ownership and professional momentum. When prospective talent sees an organization consistently elevating the people behind its work, they gain insight into the culture before ever joining it. And when senior leadership sees their teams elevated and celebrated, it reinforces why the organization matters.

              That creates a virtuous cycle: Talented people produce strong work; leaders make that work and the people behind it visible; recognition strengthens pride and reputation; and that reputation attracts more talented people.

              Organizations often ask how to build cultures where talented people can thrive. The answer is partly found in leaders who consistently bring expertise to the surface, elevate meaningful contributions, and create opportunities for people to grow and expand. Recognition is one of the conditions that makes great work possible.

              THE TALENT IS ALREADY THERE

              In the talent race, perhaps the answer is simpler than we make it. Often, the talent organizations work so hard to attract is already there. What they need is the visibility that validates their work and gives them a reason to stay, while giving others a reason to join.

              Lisa Gralnek is the U.S. managing director and global head of sustainability at iF DESIGN, and host of the award-winning podcast FUTURE OF XYZ.

              Harvard Business School’s former dean has studied CEOs for three decades. He says the role remains remarkably unchanged

              4 weeks 2 days ago

              Hello and welcome to Modern CEO! I’m Stephanie Mehta, CEO and chief content officer of Mansueto Ventures. Each week this newsletter explores inclusive approaches to leadership drawn from conversations with executives and entrepreneurs, and from the pages of Inc. and Fast Company. If you received this newsletter from a friend, you can sign up to get it yourself every Monday morning. 

              Nitin Nohria has studied and educated CEOs for three decades. Since 1997, the former dean of Harvard Business School, now executive chairman of Thrive Capital, has taught the university’s New CEO Workshop, an invitation-only program offered through its executive education arm. In 2018, he and the economist Michael E. Porter wrote a groundbreaking study on how CEOs spend their time. And last month he published a new book titled The CEO: The Role, the Reality, the Responsibility.  

              Given this body of work, I was eager to speak with Nohria about how the CEO role has changed over the past 30 years, especially when so many executives feel the job has become more complex and demanding. Nohria says the context in which CEOs lead changes all the time—technology, social mores, regulations all evolve—but the role remains largely unchanged.  

              “The role of the CEO is surprisingly enduring,” he says. “It’s still to craft a vision for how the company will navigate the moment. It is to select an extraordinary senior team, [and] it is to align the organization [to] allow everybody else to make decisions that you can celebrate.” 

              Same role, different demands

              There’s also a consistency to the experience of becoming a CEO. Nohria says for 30 years he’s been asking newbies to identify the things that surprised them upon taking the job, and the same answers rise to the top: They’re unprepared to always be in the spotlight; they quickly come to the realization that they cannot run the company by making every decision themselves; and their schedules, already packed and demanding, become even more fragmented.

              Nohria says they need to remind themselves: “I can’t outwork this job. And my job is to shape the context where others can make decisions that I celebrate, as opposed to being the person who feels like they are driving the performance of this company.” 

              Nohria sees small differences in the way chief executives spend their time today. Since the COVID-19 pandemic, virtual meetings are more commonplace today than they were when his research about how CEOs spend their time was first published. Chief executives may find themselves engaging with regulators more than they did a decade ago.  

              Still, other stakeholders, such as customers and frontline employees, continue to get short shrift. (Nohria and Porter’s research, which entailed coding CEOs’ time in 15-minute increments, 24 hours a day, found that CEOs spent just 3% of time with customers.) “There are so many demands on a CEO’s time that the squeaky wheel gets the grease, and these are the two [constituencies] that don’t always make demands,” Noria observes. His advice for CEOs who want to make customer and employee visits a priority: “Put it on your calendar. If it isn’t scheduled, the risk that it’s going to get lost is very high in a CEO’s life.”  

              Efficiency—and an Emanuel shout-out

              Nohria says artificial intelligence tools are just starting to have an impact on the ways CEOs manage their time, with some using tools such as Copilot to help them analyze their calendars, perhaps reminding them to consider how they’re allocating their time. For example, an AI assistant could encourage moving a few things around to allow for more meetings with employees and customers. “Almost like a little bit of a good chief of staff that’s nudging you to do the right thing,” he notes. 

              Nohria makes a somewhat unexpected appearance in Roll the Calls, the autobiography of the Hollywood agent Ari Emanuel. Nohria once spent a day and a half advising Emanuel as he contemplated merging his Endeavor talent agency with another firm. While I surmised that the combustible and argumentative Emanuel is unlike most of the corporate CEOs who participate in HBS’s executive education programs, Nohria says, “He has a constancy and a driven purpose about getting things done, and I’ve always admired that about him.” As for Nohria’s observations on how Emanuel manages his time? “I haven’t studied his calendar, but if I had to guess, I think Ari would probably have more meetings in a single day than almost anybody I’ve studied.” 

              CEO time

              CEOs, how do you divide your time between internal and external constituencies? How much of your work is done in person versus virtual? And are you spending more time with regulators than in the past? Send your breakdowns—and calendar screenshots!—to me at stephaniemehta@mansueto.com, and we’ll see how Modern CEO readers stack up to the executives Nohria has studied. 

              Read more: the CEO calendar 

              A glimpse inside the painstaking process of bringing a subway car to life

              4 weeks 2 days ago

              Although the subway system is New York City’s bloodstream, most riders rarely slow down to wonder where the giant metal cars came from, or where they go after their last stop. Now, a new photography exhibit is transporting passersby into the spaces where the city’s subways and buses reside when they’re off duty, and even the spaces where they’re made.

              Commissioned by MTA Arts & Design, In Service: MTA Photographs by Christopher Payne is the agency’s first systemwide photography exhibition capturing a behind-the-scenes look into the machines of NYC’s transit system.

              Presented in 52 light boxes, the large photographs are displayed across five locations in Manhattan and Brooklyn: Grand Central Madison, Grand Central Dining Concourse, 42 St-Bryant Park, Bowling Green, and Atlantic Av-Barclays Ctr.

              Installation view of In Service: MTA Photographs by Christopher Payne, 2026 [Photo: Nicholas Knight/MTA]

              A resident of Inwood, Payne was captivated by the 207th Street Overhaul Shop, a nearby 43-acre complex where subway cars go to be repaired. Built from the late 1920s through the early 1930s, the shop services the R62 trains—the stainless steel orange-seated cars that ride the numbered lines. He became fixated on entering 207th Street and other overhaul shops around the city to see the cars stripped of their elements.

              Lift system used to move subway cars around in the maintenance shops. New York City Transit-Subways, Coney Island Overhaul Shop, Coney Island, Brooklyn. [Photo: Christopher Payne]

              The documenting of these spaces began as a personal project, before becoming the basis for a story for The New York Times. Eventually, it led him to work with the Metropolitan Transportation Authority (MTA) to get unprecedented access to shops across the city as well as the other factories around the country that help build New York’s transit cars.

              John Reveron torquing valves on a new train engine. Long Island Rail Road, Morris Park, Diesel Locomotive Shop, Richmond Hill, Queens. [Photo: Christopher Payne]

              The exhibition features images like a partially assembled car at a subway train factory in Nebraska, a moment of friendship between workers in a bus factory in Minnesota, and simpler photographs from a locomotive factory in California for the Long Island Rail Road (LIRR) and MetroNorth.

              Jennifer Evans (left) and Ashton Petschen organizing wiring within a bus frame. New Flyer, St. Cloud, Minnesota. [Photo: Christopher Payne]

              “I loved sort of being an ambassador of New York City and being able to say to the people making these vehicles that we use every day, you know ‘I ride that every morning’ and ‘that’s where I sit’,” Payne tells Fast Company.

              Ashley Torres applying bracket for rear door hatch. New Flyer, St. Cloud, Minnesota. [Photo: Christopher Payne]

              Payne’s work often takes him behind the scenes of factories and restricted areas. These places are filled with large equipment and create a visual chaos that translates to interesting photographs. He describe his approach as finding the “aha, that’s how they did it” moment.

              The resulting photographs are carefully composed vignettes. They’re not candid, but rather “styled, not staged,” as Payne puts it.

              Rolina Alejo (left) and Joseph Whitcanack routing cables for a new Metro-North SC42 locomotive. Siemens, Siemens Mobility, Sacramento, California. [Photo: Christopher Payne]

              He often walks into rooms, trying to capture the large size of the machines, looking for shorter factory workers who might make the scale more pronounced. Other times, the photographs are meant to capture the human moments among the great machines. In one case he spotted a man wearing a bright tie-dye shirt, standing out against other workers.

              “There was a young woman working with him and she piped in saying, ‘Oh, I made that shirt for him,'” Payne recalls. “All of a sudden I had this great shot. Without them, it would have just been this empty bus, but with them, it was like these two people who have this relationship and it all sort of fit.”

              Installation view of In Service: MTA Photographs by Christopher Payne, 2026 [Photo: Nicholas Knight/MTA]

              The photographs at each subway stop tell a different sliver of these machines’ story. At Grand Central Dining Concourse, the photographs illustrate the humans in the system’s infrastructure, displaying portraits of workers at overhauls and factories. At Grand Central Madison, the images showcase scale, with images from the MTA’s five operating agencies. At Bryant Park, the photos focus on everyday components of the system; at Bowling Green, the images capture the stages of a bus from manufacturing to maintenance. And at Barclays Center, the photographs follow the lifecycle of a Kawasaki R211 car, made in Nebraska and maintained in Yonkers.

              Lyndon Block (left) and Mike Schlosser welding a subway car frame. Kawasaki, Kawasaki Rail Car, Lincoln, Nebraska. [Photo: Christopher Payne]

              But within those different stories, the installations aim to capture a system at a “peak moment of elegance,” to honor the workers’ craft, Payne says.

              “If you think about someone who works in a factory, no one really cares what they do most of the time. When I show up, to me, they’re the most important people in the world,” he says. “If I don’t honor them, then I’m not doing my job.”

              Voters think politicians don’t listen. Could AI help?

              4 weeks 2 days ago

              The student who answers the door looks a little hungover, and she may well be. It’s late August at East Stroudsburg University, a midsize public college in northeast Pennsylvania, and students are soaking in the final days of summer.

              But Kim Allen is undeterred by the girl’s half-opened eyes. A 64-year-old middle school Spanish teacher from New York City, Allen is particularly immune to the sullen stares of youth.

              “Hi, my name is Kim,” she says cheerily. “I’m a volunteer with Ground Truth, and we’re here to find out how people are feeling about how things are going in the country. Would you say we’re on the wrong track or the right track?”

              The student shrugs a disinterested “I dunno,” but Allen presses on. She asks about her concerns, how likely she is to vote on a scale of 1 to 10, and whether it would be for Republicans or Democrats. On voting, the student ranks herself a 7. On Republicans, an 8.

              Ordinarily, that would be Allen’s cue to leave. She’d traveled to East Stroudsburg that Saturday, as she had about a dozen Saturdays before, to help Democrats reclaim a crucial House seat they lost in 2024. For most Democratic groups and campaigns, talking to a Republican voter would seem a waste of time.

              But Allen is part of a novel outreach effort attempting to do things differently. Her mission, according to the script on her clipboard, is to help Democrats “do a better job of listening to everyone, whether we always agree or not.”

              [Photo: Anna Goldstein / Swing Left]

              And so, Allen listens as the student talks about, among other things, her concerns regarding immigration. Allen shares her own perspective and discusses the candidates in the district: Scranton’s Democratic mayor, Paige Cognetti, and the Republican incumbent, Rob Bresnahan. But over eight minutes, she never gives the hard sell. As the sticker on her phone reads, Allen is only “here to listen.”

              Allen is one of more than 3,000 volunteers dispatched to more than 50 House districts across 23 states since January as part of the Ground Truth project. Launched late last year by the progressive grassroots group Swing Left, Ground Truth aims to knock on every door and engage voters of all political leanings in open conversations.

              This approach, known as “deep canvassing,” isn’t entirely new; it has existed primarily in small pilot programs since at least 2008. What’s new is the scale it set out to achieve, and the technology Ground Truth is using to make sense of the conversations.

              If one of Ground Truth’s goals is to rehabilitate Democrats’ image as the party of out-of-touch elites, the other is to turn those conversations into actionable data about the people on whose doors Democrats wouldn’t usually knock. That requires technology that can extract meaning from many thousands of open-ended conversations that can stretch 10, 15, even 30 meandering minutes. Making sense of all of that is something traditional Democratic infrastructure was not designed to do.

              But large language models can, and Swing Left is leaning heavily on them to undergird its listening tour. Canvassers like Allen use an AI-powered app, where they can record voice-to-text recaps of their conversations with potential voters. Those conversations get stored in a central database where Swing Left organizers can query the records in plain English (“What were voters talking about in Pennsylvania’s 8th district this weekend?”) and distill them into useful or surprising findings for campaigns. 

              That may mean helping a campaign find the repeat Trump voters in their district who are open to voting for a Democrat or unearthing hyperlocal concerns that haven’t yet crossed a candidate’s desk. In one community, this system unearthed widespread worries about a rusting water tower in the district—an issue the Democratic candidate wasn’t aware of and that certainly wouldn’t have been on any traditional canvassing questionnaire.

              Swing Left feeds those nuggets directly to candidates and funnels all of the information it collects into the Democratic Party’s central data warehouses so campaigns and third party groups can dig through it on their own—now and in the future. “Democrats have a real trust problem,” says Yasmin Radjy, executive director of Swing Left. “Our perspective is that the way that you build trust has to start by listening.”

              Ground Truth is far from the only effort to modernize campaigning with artificial intelligence. According to a survey in March from the American Association of Political Consultants, more than 80% of consultants from both parties report using AI weekly. It’s being used to match campaigns with online content creators who can amplify their message and to craft campaign ads and talking points. For its part, the Democratic National Committee has undertaken its own phone-banking listening effort called Local Listeners, which set out with much the same aim as Ground Truth. 

              Pulling off this technological shift has been no easy task. For one thing, it carries political risk. One survey found that most voters disapprove of the use of AI for many campaign related activities—including uses as simple as drafting fundraising emails. 

              Pervasive fundraising hurdles on the left and the slow pace of change within politics more broadly have also made it difficult to reach the kind of scale Swing Left hoped to achieve. The group initially set out to talk to 500,000 potential voters before Election Day. By the end of September, it had talked to just over 40,000.

              Even so, Radjy sees Ground Truth as a “proof of concept” for the 2028 election and beyond. It’s a test of whether AI—despite being an unpopular technology among most voters—can become a potent tool for understanding them.  The hope is that with this information in hand, people won’t just feel like some stranger was willing to listen, but that they might actually be heard.

              ‘It is too transactional’

              The United States is facing a crisis of political disillusionment. According to Pew Research, more than 80% of Americans say politicians don’t care what people like them think. That’s a level of political cynicism that is among the highest of dozens of countries Pew surveyed and is shared across just about every demographic. During the 2024 election, working class voters in particular viewed Democrats as being more “out of touch” than Republicans.

              The Ground Truth project emerged out of the rubble of that election, in hopes of upending the extractive nature of voter outreach, where every two years, the same voters get pummeled by text messages, emails and door knocks asking them for things—a vote or a dollar—but rarely, if ever, get asked whether and how they’re getting by.

              That was Allen’s experience knocking on doors for Kamala Harris in the outskirts of Philadelphia the Sunday before the 2024 election. Tasked with turning out reliable Democrats, she found that, “apart from Black women,” people simply weren’t interested in voting. Racing from house to house, she learned little about the people she met or why they’d become so cynical about democracy. The process felt transactional and demoralizing—“one of the most depressing experiences in my life,” she recalls. Allen went home convinced that “something has to be different.”

              She was hardly the only one feeling that way. For decades, Democrats have dominated the so-called political “ground game,” compared with Republicans. With rare exception, they open more field offices and engage more volunteers than Republicans, who have begun to embrace canvassing, but tend to do it through paid campaigns. “Most of our canvassing efforts are focused on ‘paid doors’ where you hire one of these firms and you knock as many doors as possible, as cheaply as possible,” says Eric Wilson, a Republican strategist and executive director of the Center for Campaign Innovation.

              During the 2024 cycle, Democrats went particularly big, mounting an historically huge voter outreach effort that included 1.9 billion voter contact attempts, including some 270 million door knocks, according to one analysis. And yet, the number of actual conversations volunteers were having with voters was simultaneously collapsing, as fewer people answered their phones and doorbells. 

              That mismatch led to a crisis of faith among the party’s volunteers and field organizers. “Every campaign operative believes that organizing is dead and or dying,” says Radjy, the Swing Left executive director. “It is too late. It is too transactional, and the scripts we use are highly robotic.”

              It was clear to Radjy after Harris lost that Democrats didn’t have a quantity problem, but a quality problem. Deep canvassing emerged as a possible solution. Over the years, researchers have found that canvassing efforts focused on more extensive, empathic conversations about anything from abortion to trans rights can have their own persuasive effects. But one reason this style of canvassing hasn’t taken off is that it’s difficult to track and scale.

              Listening, it turns out, is not just a human challenge, but a technical one. It’s one thing to sift through a few hundred unstructured conversations about people’s political views. It’s another to do that across many thousands of them. 

              Instead, Democrats try to maximize efficiency by focusing on a campaign’s likeliest supporters, or at least the ones that the data suggests could be persuaded. Canvassing scripts include quick and dirty questions—do you plan to vote and for whom?—that can be easily collated into what are known as support scores. Those scores drive decisions about which people campaigns and PACs will spend their money trying to convince. And the cycle continues.

              The problem is that support scores are blunt instruments. They’re often wrong and even when they’re right, they obscure the reasons underpinning people’s political beliefs. Even worse, Radjy says, there are plenty of people for whom Democrats have no robust data, so they wind up being overlooked.

              But the proliferation of large language models changed things, Radjy argues. Suddenly, there was a tool that could crawl through endless sprawling conversations and find the common themes and surprising insights within them in minutes. With a tool like that, Swing Left believed deep canvassing could work at a much greater scale. 

              “They’re the only ones who have really scaled deep canvassing that have a national volunteer team that has the ability to have boots on the ground in every single state,” says Ella Barrett, one of the organizers who helped develop deep canvassing in the late 2000s and now trains other groups, including Swing Left, through the organization she cofounded, New Conversation Consulting.

              The timing of this technological revolution was fortuitous. In 2024, the Federal Election Commission lifted a ban that had prohibited canvassing groups like Ground Truth from sharing their doorstep data with campaigns, state parties, and outside groups. Under the ruling, Swing Left would be able to funnel all of the data it collects through a central repository called the Democratic Data Exchange, ensuring important findings got disseminated throughout the party. “The purpose here is to be able to move this data about these conversations back to the party so the campaigns can improve their targeting and build their messaging,” says Amanda Coulombe, CEO of the exchange. 

              For its canvassing technology, Swing Left worked with an outside group, Everybody Techworks, which had built its own app, called People. Rather than relying on the Democratic party’s longstanding canvassing app, miniVAN, Ground Truth organizers use People to knock doors, collect data like names and phone numbers, and record voice-to-text notes recapping the conversations they’re having. Volunteers can attach these records to a person who’s already in the voter file, or they can add new entries, based on whoever they bump into along their routes. 

              All that data gets fed into the central database Swing Left uses to sift through these records and pull out findings. The system can be buggy: During one demo Swing Left showed Fast Company, it stalled out while crawling data from Pennsylvania’s 10th district. But it has delivered some valuable discoveries, says Brooke Butler, national political director for the Democratic Congressional Campaign Committee (DCCC), which works to elect Democrats to the House. 

              In the first half of this year, about one in three people canvassers talked to weren’t listed as likely Democrats in the voter file. But in person, about half of them identified themselves as either toss-ups or leaning toward Democrats. One in five identified as strong Democrats. That means Ground Truth is finding voters that traditional canvassing, focused on support scores, would have skipped. The DCCC is now adding all of the voters Ground Truth has identified as undecided or mildly supportive to its list of targets. “We didn’t have access to this kind of rich voter data at scale before,” says Butler.

              But pulling off a national deep canvassing campaign didn’t just require new technology. It also required training volunteers in a new way of speaking to voters. To develop a script, Swing Left volunteers and organizers spent six months working with New Conversation Consulting, painstakingly planning the best way to approach voters, how and when to disclose their own political leanings and plotting out decision-tree style follow up questions. If a voter says they’re feeling hopeless about the political process, the August script read, “Validate what is hard,” and share your side, but don’t focus “solely on Trump.” 

              Volunteers for Ground Truth get a brief training in how to listen deeply. On the East Stroudsburg trip, this training happens on the Swing Left bus, which shuttles dozens of volunteers from New York. “The point is not to persuade people to the Democratic side, but to listen to people’s concerns and figure out how Democrats can best serve the people in this district,” says one of the trainers that day, gripping the back of a bus seat. “A lot of the reason why we’re in this mess is that we were not listening.”

              Last November, Swing Left piloted the program with a small group of volunteers before launching the nationwide effort in January. Violet Kopp, Swing Left’s East Coast organizing program manager, admits she was doubtful it would work. “We were like, there’s no way people are going to answer,” she says, let alone engage in long, soul-searching conversations. 

              But it turned out the opposite was true. Over the last year, Swing Left says one out of every three door knocks have resulted in a conversation, and two-thirds of those conversations were what the group considers “substantive” ones, often lasting 10 minutes or more. 

              These conversations validate Swing Left’s instinct, Kopp says, that “there’s this hunger to be heard.”

              ‘I hit the jackpot’

              As Allen walks the streets of East Stroudsburg, dressed in a purple paisley dress and matching sneakers, she encounters voters of every persuasion. There’s the reluctant voter from a family of loyal Democrats, whom Allen encourages to call his mother; the woman who had voted Republican in the past but swears she never would again; and the college guy sitting on the porch with his buddies, who calls Democrats “corrupt as shit,” even as he acknowledges some of Cognetti’s mayoral accomplishments in nearby Scranton. 

              In a few weeks, all Ground Truth scripts will shift toward more targeted persuasion, but for now, in the final days of August, Allen mostly listens. The guys appreciate the approach. “A lot of people our age, you can’t have a conversation,” one says afterward.

              Her most promising encounter comes with a young woman I’ll call “Amy,” who says she rarely votes because she doesn’t know enough about the candidates, but who worries about disappearing abortion rights and wildfires. When Allen shows her photos of Bresnahan and Cognetti, a departure from Ground Truth’s script, Amy perks up. She likes the idea of voting for a young woman, she says, and she tells Allen she is now more inclined to vote for Cognetti. This exchange isn’t entirely unusual. In one small September sample collected by Swing Left—albeit after volunteers shifted to more persuasion-focused outreach—about a third of the people volunteers spoke with moved toward Democrats by the end of their conversations. 

              Allen shares her phone number with Amy, in hopes of convincing her to volunteer. “I hit the jackpot,” Allen says later.

              Still, the day’s canvassing also reveals the limitations of Ground Truth’s approach. Some volunteers struggle with the technology, while others veer off script. One repeatedly—and erroneously—introduces herself as a representative of the Democratic Party. And not everyone was receptive to conversation. There are people who don’t want to talk, people who talk a lot but say little, and at least one person who quickly reveals extremist views that make him an unlikely recruit to the coalition. One risk of knocking on every door is the possibility of wasting limited volunteers’ time or turning out the other side.

              Another question hanging over the entire project is whether and how campaigns will actually use this data. A select few have become early adopters. In Oregon’s 5th district, for instance, Congresswoman Janelle Bynum’s reelection campaign has used Ground Truth’s data to find Republican voters who are open to new candidates. “Ground Truth’s initial efforts showed us the evidence we needed to broaden our outreach beyond traditional narrow voter targeting,” says Quinton Berkompas, Bynum’s campaign manager.

              In Arizona, where Dr. Amish Shah edged out the establishment candidate in a hard-fought primary, Ground Truth’s early canvassing “built a strong foundation and helped us hit the ground running in the general election,” says Eve Zhurbinskiy, a campaign operative.

              Yet this kind of embrace is far from universal. According to Swing Left’s most recent data, only about 29% of the campaigns Ground Truth targeted are fully utilizing and acting on the data volunteers are collecting. About 40% say they lack the capacity to fully implement the program, and 5% declined outright to work with Ground Truth. The rest are somewhere in the middle. The Cognetti campaign in Pennsylvania didn’t respond to multiple requests for comment, so it is unclear how extensively it’s using the data.

              Ground Truth is asking campaigns to fundamentally change the way they operate, including by talking to people the voter file suggests they shouldn’t and looking beyond support scores. But in a system as fragmented as political campaigning, Radjy says, that change is admittedly slow moving. “We have a set of traditions that have become baked in as habits for the ways we campaign,” she says. She noted that one campaign cited its canvassing packet printing schedule as a barrier to working with Ground Truth. 

              And Swing Left is not immune to the other forces afflicting the Democratic party, including a pervasive fundraising shortfall that made it difficult for the group to hire organizers and achieve its 500,000 conversation goal. Still, Radjy says she remains optimistic about what Ground Truth was able to prove this cycle, even given the limitations. “Building something this new and complex took longer than we planned,” she admits. “While we’re not on pace to reach it this cycle, our goal galvanized our community and the ecosystem to establish crucial infrastructure and prove what is possible.”

              For now, the most immediate effect of this work may be on volunteers like Allen, who, after 2024, were beginning to wonder whether all of this work—the sweaty sneakers, the tired legs—was even worth it. Now, Allen says she’s planning to retire at the end of the school year and devote herself full time to Ground Truth volunteering, not just because she wants Democrats to win in 2028, but because of how the whole process makes her feel. “This feels different,” she says. “This feels like rebuilding something that’s been broken.”

              Why the grocery store of the future could have more mini cans and personal packs

              4 weeks 2 days ago

              The future of food packaging is petit.

              While warehouse stores offering food and household items in bulk will always have a draw among consumers looking for volume discounts, the grocery store of 2035 could look different from today’s, with tinier packaging on its shelves, according to a new report. Unlike stealth shrinkflation, which packs less product into the same sized package for the same price, the new era of smaller portions is part of the selling point.

              Traditional packaging strategies for consumer packaged goods (CPG) prioritize things like upselling, pantry loading, and increased consumption, but those strategies are becoming less relevant, warns the report, from the consumer intelligence company NielsenIQ. Instead, manufacturers, retailers, and designers must rethink their packaging formats to meet the needs of multiple shifting consumer trends at once.

              Household sizes keep going down, living costs keep going up, and the growing adoption of GLP-1s means more consumers are buying less. In other words, packaging design for high consumption and big families with deep, full pantries is packaging design for a world that no longer exists. Instead, companies need to think smaller.

              “If the American household is shrinking, the way it shops is shrinking right along with it,” the report says.

              Already, CPG giants like PepsiCo have begun testing smaller units. Mike Del Pozzo, president of the company’s U.S. beverages division, said in June the company was introducing mini cans at a low price point and considering “new price points and pack sizes” for other beverages too. And last year Campbell’s rolled out 2.5-ounce single packs of Goldfish that retail for less than $2. The report says smaller packaging sizes are showing up especially across soft drinks, ready-to-eat cereal, and salty snacks, but building in other categories too.

              Average household size has fallen from 2.6 people in 2000 to 2.5 people today due to factors like economic pressure and delayed family formation, the report says. And 21% of households now have at least one GLP-1 user. At the same time, food-at-home prices have risen 2.7% year over year as of June.

              That means things like smaller packs, trial sizes, mini cans, and portioned servings are appealing to a new generation of shoppers. The unit price might be higher than buying in bulk, but smaller sizing lowers the cost at checkout for inflation-pressed consumers.

              “Value traditionally sits at the intersection of unit price and product attributes that are important to the shopper,” the report says. “Smaller packages can create value in a different way by lowering the upfront cost, improving storage of the items, reducing the risk of waste, and better matching the amount a shopper needs based on what they know they will fully consume.”

              The family Costco run is giving way to a new way of grocery shopping that’s about buying fewer things at once. The report notes that fill-in shopping makes up most grocery trips, and changing consumer habits mean those quick trips could over time come with even smaller portion sizes.

              Idealab’s Bill Gross is building infrastructure for AI-assisted startups 

              4 weeks 2 days ago

              For decades, the entrepreneur and investor Bill Gross has been celebrated for shaping the way tech businesses get off the ground.  

              Idealab, which he founded in 1996, is considered the industry’s longest-running tech incubator; the directory of companies it’s backed—including CitySearch, Coinbase, Picasa, and Twilio—doubles as a museum of ever-shifting trends in logo design. Over the years, he’s also founded businesses in fields ranging from internet advertising to clean energy. 

              “My whole life, I’ve been looking at the world, seeing problems, and trying to start companies to solve them,” Gross says.  

              And for him, finding problems in search of a solution has typically involved extensive reading, including most of the major finance and tech publications. Recently, while reading an article about copper shortages, Gross reflected on how much of what he reads contains some sort of implicit problem statement and wondered how artificial intelligence might fare at generating business ideas based on news reporting. 

              In an initial test, Gross supplied Anthropic’s Claude with about 7,000 articles he had recently read, along with material he had put together about coming up with business ideas. “It came back with 32,000 company ideas, and they were incredible ideas,” he says. “They blew my mind at how good they are.” 

              At first, Gross wondered whether that meant the technology would make his role of finding exciting new business ideas obsolete. But instead, he realized, it meant AI could help further his longstanding goals of solving more problems at scale. And when he expanded the dataset provided to the AI, including documents like expired patents, SEC filings, consumer complaints from social media, and plenty of material from science publications, the AI system was able to generate a database of more than 1 million ideas. 

              “I’ve got a database of 1.2 million ideas that basically solve pretty much every world problem because it was trained by reading every world problem,” he says.  

              Gross intends to keep updating that database with the help of AI, and his plan is to make it available to anyone looking to start a company. Would-be entrepreneurs, Gross announced on October 19, can search the database for projects that match their interest and skillset, filtering based on notions of how much investment a particular idea might require, how much of the business can be automated with current technology, and whether there are any geographical constraints. 

              In addition, Gross is putting together funding designed to back entrepreneurs around the world starting a business based on one of the ideas. He’s doing this through a new venture he calls Idealab.ai. The organization indicates it expects to launch that side of the program by the end of 2026 and hopes, in the long term, to help connect founders from around the world with $100 million in potential funding from various sources. The company also anticipates offering features to help connect entrepreneurs with collaborators.  

              In addition to the database, Gross is offering entrepreneurs additional AI-driven resources, including a chatbot informed by lectures and writing he’s delivered over the years. He is also offering what he calls a periodic table, with tiles listing 118 functions a startup needs to perform, from email marketing to facilities management, along with software tools that assist with them. Where software doesn’t yet exist, Idealab.ai may work to build it out. 

              “All the tiles that nobody does, we’re going to build out so businesses can use this periodic table to help them grow their business,” Gross says. 

              In essence, he says, this follows from what the original Idealab incubator offers startups, handling back-office functions like HR management for the companies it backs. Now, increasingly, those functions will be managed by software, including AI, meaning many companies will no longer need to physically be based out of an accelerator that can offer such support. The technology also won’t replace humans’ role in businesses, Gross emphasizes, saying they’ll still be needed to make decisions and guide the technical operations. 

              “The human is still making all the decisions,” he says. “It’s just the mechanics that we can take away from them.” 

              Cloudflare, which is cited as useful for a dozen functions on the table, also announced plans to offer $100,000 in credits to businesses started through the Idealab.ai platform. Other businesses are expected to follow, Gross says. The company is also working with educational institutions, including Arizona State University and Imperial College London, as they support student entrepreneurship facilitated by the idea database and periodic table. 

              One hope, Gross says, is that the technology’s potential for helping founders might appeal to young people and others concerned about AI-driven job losses. If the technology can assist in creating new businesses, that may help overcome the potential loss of entry-level jobs at existing companies, while delivering new opportunities to young people with an entrepreneurial bent.  

              “This is the way maybe AI could make up for that and actually give you something with more purpose than taking an entry-level job,” he says. 

              A more accurate weather forecast could start with these high-flying drones

              4 weeks 2 days ago

              For nearly a century, reading the upper sky has meant tying a box of sensors to a balloon, letting go, and kissing the hardware goodbye. And they only go up at fixed times. If a storm starts brewing between launches, nobody’s up there to see it.

              Meteomatics, a Swiss weather technology company, thinks we can get better weather data. It recently introduced Strato, a multicopter drone made to study Earth’s atmosphere. It is designed to reach higher altitudes of the troposphere—the part of the atmosphere where weather happens—than weather balloons, at a fraction of the price. And once its mission is complete, it returns to its launch site ready to go back up again.

              A weather instrument you can fly over and over, wherever and whenever you need it, could change how we watch the sky and predict the weather.

              [Image: Courtesy Meteomatics] Data on demand

              According to Meteomatics, no drone before Strato could profile the atmosphere all the way from the ground into the stratosphere on demand.

              As Strato climbs, it reads temperature, humidity, pressure, and wind at every altitude, building a sort of core sample of the air above us. Meteorologists call that a vertical profile. Until now, only a radiosonde, the instrument package that hangs from weather balloons, could collect that data.

              Using a radiosonde has its flaws. Weather balloons need roughly 40 minutes just to rise to 12 kilometers, or about 39,000 feet. Additionally, 90% of weather balloons never return to the people who launched them, their radiosondes lost forever, essentially becoming pollution. Some new balloon models incorporate GPS and controlled descent that allow for recovery of the sensor packages, but scientists still have to drive for miles to pick them up. Plus, they must be refitted to fly again, which takes time. They’re impossible to operate continuously.

              Strato solves these problems. The drone rises to 34,000 feet in about 25 minutes, Meteomatics CEO Martin Fengler told me, finishing a full atmospheric profile in about half the time a weather balloon would. Plus, Strato always returns to its operators’ exact location, ready to ascend again and again. 

              [Image: Courtesy Meteomatics] More reliable forecasts

              Strato’s speed and its continuous-launch ability are the keys to get more data about the jet stream—the atmospheric highway that weather systems travel across—in a timely manner.

              “If the data isn’t located correctly relative to the jet stream, the forecast won’t be reliable either,” Fengler said when he announced Strato.

              Jet streams—there are several, in both Northern and Southern Hemispheres—are rivers of wind racing around the planet near the top of the troposphere. (The stratosphere starts right above this.) Up there, a sudden twist in the wind or a jump in temperature can push an airliner off its route, keep a rocket on the pad, and ruin your picnic in the park—or worse, as extreme weather events have shown.

              And we know the least about what happens up there. Meteomatics says government observation networks don’t reliably cover jet streams’ highest altitudes. This is because balloons typically launch on a fixed timetable from their fixed bases. Nobody deploys an extra one out of the blue because a weather system is coming through. So airlines, military pilots, and launch teams often make key decisions with data that’s stale or was gathered elsewhere.

              Strato’s readings feed Meteomatics’ own high-resolution forecast models—along with data from the company’s older, lower-flying “meteodrone” called Tropo—to provide much more accurate atmospheric readings.

              [Image: Courtesy Meteomatics] Excuse me, coming through

              As promising as the tech sounds, there could be potential problems with Strato’s approach.

              At 34,000 feet, Strato shares the sky with airliners at cruising altitude. I asked Fengler how they keep Strato out of the way of a jet full of passengers. According to him, there are no automatic, worldwide handshake protocols with air traffic control for drones, so Strato follows whatever rules each country sets.

              There are some safety stopgaps, Fengler says. Strato can carry a transponder, a radio beacon that puts it on controllers’ radar like any other aircraft, and the drone’s operators could issue warnings that tell pilots ahead of time to steer clear of an area. They could also coordinate with local air traffic control. 

              Since the mix of requirements varies with each country, launch site, and airspace, Fenglar notes that Meteomatics helps its customers handle the regulators. In other words, filing lots of paperwork.

              [Image: Courtesy Meteomatics]

              The atmosphere itself poses another hurdle. Flying that high is a nightmare for a drone. More than 6 miles up, the air is so thin that propellers barely have anything to grab onto. And if you change the shape of the blades to account for the lack of air density, then they’ll guzzle battery near the ground, where the air is thick, shortening the flight time and making it impractical for weather operations.

              Fengler tells me that Meteomatics’ engineering team had to test multiple combinations of propellers and motors to find the sweet spot. First, they used a flight simulation to predict how each design variation would behave as the air thinned. The most promising of those went into a vacuum chamber that mimicked high-altitude air, and into a wind tunnel that hit them with brutal gusts.

              The survivors in the tests, Fengler says, keep the drone climbing in thin air without draining it down low, and still have power to spare when the wind fights back.

              Another showstopper that they had to overcome? The weather itself. But Strato’s design, Fengler says, takes this into account. It can withstand sustained winds of up to 80 mph and its body is shaped to slice through the air. “An automatic wind mode swings its nose into the headwind like a weather vane,” he says. Rain doesn’t bother it either. It’s waterproof. Finally, if ice starts to form, Strato heats its own propellers to melt it. The operator, Fengler adds, monitors altitude, wind, temperature, and icing alerts live for the entire flight.

              [Image: Courtesy Meteomatics] The fine print

              Ice is not the only enemy. At high altitude, cold air is really the worst adversary of drones’ batteries. Strato runs on lithium-ion cells 0.83 by 2.76 inches in size. They are capable of charging anywhere from 32 to 140 degrees Fahrenheit, delivering power to the drone at minus 40 degrees.

              This doesn’t mean that the batteries won’t suffer. Fengler can’t yet say how many charges a pack survives before it starts to fade but he claims that Strato’s cells have beaten earlier generations in his team’s cycle-life tests. He tells me that Strato drains them in such a peculiar way that Meteomatics wants more data before publishing a number.

              The sensors can also be tricky. Balloons float up slowly with the wind, which gives their instruments time to settle as temperature and conditions change. Strato goes up at almost double the speed of the old balloons.

              According to the company, the drone goes quickly through temperature inversions, layers where the air gets warmer as you climb instead of colder, and through wind shear, the sudden shifts in wind speed or direction from one altitude to the next. If the sensor reacts too late, there could be bad readings due to a phenomenon called “dynamic heating.”

              Fengler claims his company’s sensors are built for the fastest possible response and sits out of reach of the rotors’ turbulence. Dynamic heating, he insists, is “not measurable in our testing.” 

              Better and faster?

              Whether Strato lives up to its promise of faster, better weather data and predictions is yet to be seen. It’s a great proposition, however, especially if you factor in the operation costs. By Meteomatics’ own estimate, ground stations spend $150,000 to $200,000 a year on weather balloons that fly once and disappear. A drone that comes home could be an easy sell, if the price is right.

              Fengler wouldn’t say how much Strato costs—“it depends on how each system is configured,” he told me—but it will definitely be much cheaper in the long run. At least that’s true for traditional disposable weather balloons, which still dominate roughly 80% to 90% of high-altitude survey operations.

              “Ten years ago, a drone that effectively flew into the stratosphere to collect weather data was a pipe dream,” Fengler says. “We’ve made it a reality.”

              If the company’s claims hold up, the weather balloon can finally retire after nearly a century of one-way trips.

              What a scary boot camp incident taught Marine veteran and Aflac president Virgil Miller about leadership

              4 weeks 2 days ago

              After more than 22 years at Aflac, the 11th grade feels like another lifetime. But that’s when I decided to join the United States Marines. That “first job” eventually found me in foreign land, a bit scared but ready to do my duty during Operation Desert Storm. But the most important lesson I carry from these formative years might be the one I learned before I ever left boot camp at Parris Island.

              The leadership philosophy I follow today centers on people, then process, then performance. I call it The Formula. But there’s something deeper behind it: Authentic leadership means recognizing and meeting the moment. Sometimes the job needs a disciplined executive. Sometimes it just needs a caring human. Most times it’s a combination of both.

              The night that changed everything

              When you’re a Marine recruit, you take a turn pulling what they call “fire watch”: overnight guard duty in the squad bay while everyone else sleeps. You’re not armed; you’re just aware . . . and responsible. It’s a test of your discipline to stay alert even when you know nothing will happen. Until it does. One night when I had the duty, the thing I became aware of, and suddenly responsible for, was a squad mate who was trying to take his own life.

              I knew I had to wake up my DI, my drill instructor. If you’ve ever been a kid who had to wake up a grumpy parent in the middle of the night to tell them you wrecked the car, multiply that apprehension by at least 10. But I had no choice, so I banged on the door. The response was about what you’d expect. The DI came out in his underwear and went from zero to 60 in a heartbeat, jumping in my face and berating me. I finally managed to blurt out the news. The DI muttered something I can’t repeat, disappeared back into his room, and shut the door.

              I barely had time to process my confusion before the DI reemerged. This time he was dressed. He looked me in the eye. You stay here.

              By now, I imagine many of you are thinking of a scene in a popular movie that started out the same way. But that’s where the comparison ends—because at that moment, everything changed. The sergeant went into the bathroom, put his arm around the boy, and started talking in a way I’d never heard from him before: like a human.

              “The Marines isn’t for everybody. Let me take care of you and get you out of here. It isn’t worth killing yourself over.” And that recruit got to go on with his life.

              Taking the lesson to heart

              My first reaction, apart from relief, was: My DI has been faking it. He’s been putting on an act. The guy I’ve known is tough, loud, and, shall we say, colorful. But my second thought was: Maybe, when the situation called for it, he was more human than I thought. Maybe you can be tough and disciplined, while also being compassionate.

              I can’t tell you when the real lesson from that night dawned on me, but it has guided me ever since. In hindsight, both faces were real. And the things you do under the guise of each persona are valuable. So today, in business, you’ll see me strive to show the control and stick-to-itiveness you’d expect from a Marine vet. But you’ll also see an echo of that fatherly DI who put his arm around a distraught kid. The circumstances are seldom that dramatic. But the value of human understanding, patience, and seeing things through another person’s eyes cannot be overstated.

              The long road since

              After Parris Island and after Desert Storm, I was out of the military before many of the people I work with today were even born. I worked for a cable TV company, a delivery company, and, finally, an insurance company. I started my career working the phones in a call center and earned 16 promotions in 12 years.

              In the call center, helping stressed and disgruntled policyholders wasn’t like a life-or-death experience you can have in a war zone, but my drill instructor’s example still spoke to me. I had rules to follow, and I couldn’t wave away people’s problems. But what people needed to hear was that someone understood how they felt and was trying to help the best way they could.

              People, process, performance. That is still The Formula I apply to leading our company. And I still oversee call center operations—albeit in a much different role. Some of the people who reach out to our teams have routine questions about how to fill out a form. Some have been in an accident or had a loved one involved in one. And some are cancer patients trying to navigate the scariest moments of their lives.

              Those callers, just like every Aflac employee, sometimes must hear some tough news. Other times they just need a compassionate arm around the shoulder. I try never to forget that we’re all human.

              My First Job is a recurring series in which prominent business leaders share what their first job was and what they learned from it.

              The new ‘mommy track’: Working moms are founding startups to solve problems the government and Corporate America won’t

              4 weeks 2 days ago

              When Stephanie Fornaro was 5 and her younger brother was 6 months old, her mother left and did not come back. When Fornaro had her own children, she vowed she would do better. She left her sales job in preparation for the birth of her second child in 2017.

              But after the birth, Fornaro struggled. Her husband, an entrepreneur, was on the road half the time. And she was lost in a fog of postpartum depression. Fornaro didn’t want to hire help. “It felt like I was failing as a mother because I was allowing someone else to raise my son,” she says. “I spent my entire life wondering how does a mother leave her children? It wasn’t until I went through a bout of postpartum depression myself that I understood. It wasn’t a lack of love. It was a lack of resources and support.”

              America tells women they can have it all: children, a career—and their sanity. But the numbers don’t bear this out.

              Mothers in the United States are not OK. The percentage of mothers with excellent mental health dropped from nearly 40% in 2016 to 26% in 2023. During the same period, mothers who rated their mental health as fair or poor jumped by more than 60%, according to a 2025 study published in JAMA Internal Medicine.

              Some mothers are voting with their feet. An analysis by KPMG found that over the past three years the number of college-educated women with children under 5 in the workforce has declined, while employment rates for women without children and most groups of men have increased.

              Many are opting out of motherhood altogether. Over the course of this century, the U.S. fertility rate has steadily declined. In 2025, the rate hit a low of 53 births per 1,000 women, a 23% drop since 2007.

              Fornaro took a different approach. She’s dedicated herself to improving the system. Employing a nanny had made motherhood doable for her, so in 2022, Fornaro started Hello Nanny!, a platform that simplifies hiring childcare, from recruiting and vetting candidates to making sure nannies are paid legally, as opposed to under the table.

              As she built Hello Nanny!, Fornaro realized that even the tax code around childcare is broken, making it even more difficult for families to get help. While employers can deduct payroll from taxes, families must pay nannies with posttax dollars. In 2026, Fornaro started Workforce Infrastructure Institute, a nonpartisan research foundation that advocates for tax code reform for childcare.

              Fast Company put out a call asking mothers who’ve started businesses aimed at addressing these problems what they were building and why. More than 130 responded, including a few from countries like Canada, the U.K., and Germany, which offer more extensive government support for working parents.

              Their companies range from one selling nontoxic formula for treating lice, to numerous AI apps for reducing the mental load, to lovingly crafted coaching and support group programs covering everything from sick children to the sandwich generation. (We’ve listed them at the end of the article. While a number were designed to employ working parents, we omitted companies that did not specifically have a product or service that helps parents.)

              Individually, each business tells the story of a talented mother who grew so frustrated with the system she decided to build her own solution. “If you look at other cultures, nobody raises a child alone,” Fornaro says. “Only in America do you have this cultural expectation that if we’re self-sufficient then we’re successful. If we need help, then we’re seen as weak and failing.”

              Taken together, the companies create a detailed map of the fault lines in American motherhood from workplace penalties to childcare gaps. They also offer a glimpse of what a better system could look like. 

              2 hinge points in American history that created today’s problem for working mothers

              American society has always put enormous social pressure for married women to abstain from working. In 1930, a series of laws made it illegal for the government to employ more than one person in a married couple. Frances Perkins, who became the first female secretary of labor in 1933, called it “selfish” for married women to work when their husbands could support them. Before World War II, only 15% of married women had a job compared with about 50% of unmarried women.

              This paradigm shifted when the U.S. entered the war, men went abroad to fight, and women were called into workforce to replace them. As the war dragged on, the government asked mothers with young children to join as well.

              To make this possible, the government funded day care centers. By the end of the war, almost 25% of married women worked outside the house. But in 1946, with the war over, the government shut down these day care facilities despite parents’ lobbying for more childcare.

              The next major moment came 25 years later, in 1971. Decades of lobbying led to the passage by Congress of the Comprehensive Child Development Act, which would have established a national day care system. However, President Nixon vetoed it because it would “commit the vast moral authority of the National Government to the side of communal approaches to child rearing over against the family-centered approach.”

              Public policy has never again come close to catching up to the culture as the number of women in the workforce steadily rose, crossing 50% in the late 1970s. In 1988, Jennifer A. Kingson, a journalist with The New York Times, described a new phenomenon: the mommy track. As prestigious law firms started offering maternity leave and flexible hours to working mothers, they also took them less seriously and put them on the “mommy track” where there was no hope of promotion.

              In the absence of any larger policy solution, the best we have is the corporate benefit of flexible spending accounts (FSAs). These started in the 1980s and allow parents to pay for childcare with pretax dollars. At the time, childcare was roughly $2,000 a year and the FSA cap $5,000. But childcare for two children is now almost $30,000 annually on average, while the average income is about $66,000. In early 2026, the FSA cap was raised—to $7,500.

              “The cavalry’s not coming,” says Suzanne Slaughter, CEO of Groundwork, a company that helps mothers find fractional work. “In the decades I have left to work, I don’t think we’re going to have meaningful policy change. So then, what can we do to evoke change?”

              This is why Slaughter, Hello Nanny!’s Fornaro, and so many other women are inventing a brand new “mommy track”: promoting themselves to CEO of a new business designed to address some element of the problem that bedeviled them as working mothers.

              Happy and healthy pregnancy and postpartum with dignity

              Thirty-six hours after Anu Sharma was discharged from the hospital following the delivery of her child, she walked back into the ER on the verge of a stroke. Sharma considers herself lucky; she comes from a family of doctors and recognized the symptoms—but she was acutely aware another woman might have died.

              The U.S., has a dismal—and frankly, terrifying—maternal mortality rate compared to other countries. In 2022, 22.3 women out of 100,000 died giving birth. (The comparable figure for Japan, for example, was 3.4 women; France, 7.6; Canada, 8.4.) Almost 80% of these deaths were preventable, and two-thirds occur within six weeks. Yet, once a woman is discharged from the hospital after giving birth, her first gynecological visit is six weeks later.

              In 2021, Sharma started Millie, which brings midwives to a modern clinic; midwives provide postpartum care, including a home visit before the six-week mark. Sharma points out that ob-gyns are specialists who are trained to handle high-risk pregnancies. By 2035, one study estimates only six states will have enough ob-gyns to meet demand. Bringing midwives into the equation can make care more accessible.

              Physical health is only half of the equation. About 10% of mothers have postpartum depression, but only about half of them are diagnosed. Another 20% have postpartum anxiety. Louise Song created Momma Chill after paying $800 out of pocket for psychiatric care, and ultimately left her job at Amazon. While still nascent, Momma Chill provides group support and coaching for mothers dealing with postpartum anxiety or depression to help them navigate caregiving and their career.

              On a larger scale, Christy Davis and Caitlin Probst built VillageFor, an app including AI voice analysis that can provide a snapshot of a mother’s mood and stress levels, allowing mothers to track their well-being over time. The app can connect mothers to a maternal mental health hotline or text their care providers if there’s a crisis.

              Even if a mother is healthy, the physical reality of postpartum is messy and inconvenient, especially if you’re breast feeding. In 2016, Patrice Hayden Meagher, a mother of four, forgot to pack a pump part. At the time she was an executive in commercial real estate and had a meeting with a major client. Because she’d forgotten a part, she could not pump, and she began to leak, staining her blouse. Meagher was humiliated. In 2020, she started MilkMate, which builds lactation rooms for offices and provides disposable pump kits, so mothers don’t have to worry about bringing a pump to work. In addition to being convenient, MilkMate’s spaces are welcoming—a change from most lactation rooms, which can be barren.

              MilkMate’s chief marketing officer, Brie Tascione, remembers working in an office that was very supportive and designed a maternity leave policy around her. Even then, she was pumping on a toilet seat. (This was prior to 2022, when federal law mandated that offices create a space other than a bathroom for women to pump.) 

              “I didn’t tell anyone I needed something more,” she says, noting it’s difficult for mothers to speak up, particularly when their larger needs are already being met or the system is making an exception for them. “We’re giving women permission to do something basic and necessary in a way that feels dignified.”

              Not everyone needs to pump in an office, but Tascione’s larger point stands. Pregnancy and postpartum are akin to having your mind and body leveled and then reshaped. In a society that likes to pretend we don’t have bodies that can get messy, how do we create space for mothers to feel safe and dignified?

              Childcare and house help on tap

              The crux of the problem working parents face is: Who takes care of the children and manages the household when the parents work?

              Last year, Mitra Raman and Karthika Shankar, childhood best friends, started Rosie, a subscription service that lets parents sign up for a certain number of hours of house help a month. Raman got the idea after watching her nanny play with the kids while Raman cooked dinner. “I realized I wanted to switch places,” she says. Workers hired through Rosie can provide flexible childcare or house help. As an added bonus, Rosie offers mothers who need flexible hours and find that caregiving is a good fit an opportunity to earn some money.

              In order to build name recognition, Rosie has been partnering with local events and providing childcare for free. That offering has been so popular, Raman and Shankar are considering developing it into a different offering.

              In another novel marketplace seeking to create a win-win, Alison Donnally is building JoeyCo, a platform that matches college students looking for housing to families that need an extra pair of hands. While JoeyCo was originally built for older people who needed help with chores, Donnally has used it herself to find childcare.

              Recess, started by Amy Kiska and Molly Morse, attacks the problem from a different angle. They’re building a platform that helps parents find classes and camps for kids. Recess cuts out the process of wading through dozens of websites and activity provider pages and also allows parents to coordinate with their children’s friends.

              An unexpected added benefit: Kiska and Morse have found they’re able to talk with camp and activity providers to let them know what parents really need. Kiska gives an example: “We’re able to say, ‘Hey, your camp ends at 4 [p.m.], and the vast majority of consumers needs 5. Can we stretch it an hour?’”

              Working while parenting

              Having a child is probably the riskiest career move a woman can make. In 2023, researchers at Columbia University found that women’s salaries in the U.S. drop by 50% and remain that way for up to six years. A number of the founders I heard from built companies to protect women’s careers during this time, whether it was finding interim talent to cover parental leave or coaching businesses.

              Suzanne Slaughter’s Groundwork is a job placement platform for mothers who want to go fractional. She points out that most startups don’t need full-time talent for every position, so they look for part-time talent through their networks. Most of those jobs never make it to the public.

              When Slaughter was able to show founders the caliber of the talent on her platform, they were hooked. “I was obsessed with making this period of peak child-rearing more joyful,” Slaughter says. “I observed so many friends grind through it, treating it as a period to get through to what’s next, and that made me so sad. We have to give this population better resources to continue working, on slightly adjusted terms, so you can show up for your family in a way you’re proud of.”

              Even if one is able to get and keep a job that works with parenting, there’s still the matter of keeping up with colleagues who don’t have children. When artificial intelligence started rearing its head in the workplace, Raquel Cadourcy worried about mothers. While her childless colleagues were able to learn AI after work, parents who had to take care of kids were being left behind.

              In 2025, Cadourcy started ai.moms, a virtual boot camp that teaches mothers how to use AI. It enables moms to build an AI assistant that can help them with everything from meal planning to creating standard operating procedures for delegated chores. Demand has been so high, Cadourcy says, that she’s expanded ai.moms to in-person workshops and events.

              Much of surviving working while parenting comes from having the support of people who are facing the same issues. In 2017, Darci Groves cofounded the online community Moms in Tech while working at Facebook. “Back then, being a mom in tech meant you had to pretend you weren’t,” says Groves, who is also Moms in Tech’s CEO. “It was pretty widely understood that tech was a man’s space, but being a parent was also alienating and being both was a double whammy of marginalization that was super lonely.”

              Today, Moms in Tech has grown into a worldwide organization with more than 19,000 members and in-person workshops and events. “The impact we care about most shows up in the day-to-day interactions where a mom may post that they were just a part of a layoff and within the hour, dozens of other members that person has never met before in person (but have connected through conversations over the years) are sending links for open roles and offers to support them in their job search,” writes Annie Tsai, who serves as the organization’s chief operating officer. “This kind of activation happens every day in our community, and it’s incredibly powerful.”

              Patching up the system

              While many of these businesses offer good solutions, they are still nascent. Several of them are also expensive, affordable for the kind of family where one member of the family can quit their job to build a company. Few have reached the status of providing an affordable, nationwide on-demand service.

              Part of the problem is the nature of the business many of these companies are in. JoeyCo has an extensive process for matching students and families and then training them to communicate and resolve conflicts. The company trains students to run 30-, 60-, and 90-day check-ins. Donnally is wary of growing too fast. “We’re laying the groundwork for understanding where we can go fast and where we need to maintain the human messy slowness that this business requires,” she says.

              Funding is another sticking point, as women founders still receive just about 2% of venture funding. There’s no good data on how much of this goes towards mothers building for mothers. Many founders also told me venture funding wasn’t a good fit. Slaughter of Groundwork has seen 10-fold growth, but she is wary of taking VC money. She wants to be able to log off and take care of her own family instead of having investors set the pace for her.

              While I was writing this story, my group chat blew up. A friend of mine who works in tech in Boston has a son who will enter kindergarten in two years and she’s already planning for that. School ends at 2:30 p.m., but the soonest someone could pick him up is 5:00 p.m., and Wednesdays are half-days, not to mention summer vacation. Her son has severe allergies, and there’s no nurse on staff during extended care. A mishap could be fatal. While her husband is an involved parent, she is the primary caregiver.

              She pays $450 a week for day care. A part-time nanny during the school year would cost roughly $500 a week, and $1,200-$1,600 a week full-time in the summer, assuming she could find someone willing to work that schedule.

              I looked through the 130 odd emails I’d received. Currently, Rosie is available only in Seattle, JoeyCo in California, and Recess in Texas. Camply, a similar platform, is available only in Atlanta. My friend doesn’t want to go fractional, but even if she did, Groundwork isn’t in Boston.

              “Maybe I quit my job,” she wrote after hours of fruitless brainstorming. 

              The businesses spotlighted here are important solutions. Below, I am sharing the full list people shared with me because I hope they can help anyone with specific challenges they’re facing balancing motherhood and work.

              But let’s be blunt: Some of the solutions working mothers need aren’t things companies can provide. “This is not just a VC problem; this is not just a startup problem,” says Leslie Feinzaig, founder of the venture capital firm Graham & Walker and the Female Founders Alliance. “This is an America problem.” 

              75+ businesses moms are building to help working mothers—and themselves

              Jump to Benefits & navigating parental leave / Career / Coaching & Community / Childcare & home help / Health / Media & Research / Pumping and Postpartum / Products / Reducing the mental load

              Benefits & navigating parental leave

              Back to top

              Career

              Back to top

              Coaching & Community

              Back to top

              Childcare & home help

              Back to top

              Health

              Back to top

              Media & Research

              Back to top

              Pumping and Postpartum

              Back to top

              Products

              Back to top

              Reducing the mental load

              Back to top

              Checked
              11 minutes 38 seconds ago
              Fast Company
              Fast Company inspires a new breed of innovative and creative thought leaders who are actively inventing the future of business.
              Subscribe to Fast Company feed