Skip to main content

Fast Company

Gen Z spends less money but shops at more stores. New research explains why

1 month 1 week ago

They’re far more likely to take taxis or rideshares and drop some money gambling or in prediction markets, but they also use coupons and rely heavily on public transportation. Welcome to the somewhat confounding spending habits of Gen Z shoppers.

As with any generation, stereotyping millions of people based on when they were born can prove to be tricky—but Gen Z is challenging some conventional assumptions about consumer loyalty and retail engagement, according to a new report released Thursday by advertising, marketing, and data technology companies Attain and Dentsu. While the report found that Gen Z shoppers spend 25% less than other generations—which may not be so surprising, given their life stage—this generation is spreading their more limited funds across 6% more merchants.

Making sense of how (and where) Gen Z spends their money is very important as this generation becomes a more dominant engine of the consumer economy. The study, which combined purchase activity with a survey, found that Gen Z is 25% more likely to value brand names than the price of an item, but they’re 18% less likely to buy from the same brands repeatedly.

Even with more limited funds to spend, Gen Z’s greater appetite for spending money on discretionary purchases and their openness to switching brands means that traditional marketing methods may need to be rethought. 

“Gen Z is showing us why the traditional view of the consumer is no longer enough,” Jeff Zifrony, senior vice president of strategy at Chicago-based Attain, said in a statement. “When you can see what people actually buy across retailers and categories, you can begin to understand the full picture of consumer demand, not just the portion visible within any single platform.”

HOW GEN Z IS SPENDING

As might be expected of a generation in their late-teens and 20s, Gen Z is all about having a good time and living in the present.

Americans of this generation are much more likely to spend on discretionary purchases, and 66% of their day-to-day spending goes toward goods and services like eating out, entertainment, travel, clothing, electronics, beauty, and rideshares. That’s compared with 58% of everyday spending on these categories for shoppers of all generations.

Spending on public transit, taxis and rideshares, gyms and fitness centers, casinos and betting or prediction markets, video games, and restaurants commands a disproportionately large share of their money, the study found. By contrast, groceries in particular take up a much smaller share of Gen Zers’ wallets, even though they’re as likely as anyone to buy groceries.

For a generation that’s only ever known many of the conveniences of modern life, Gen Zers are more likely to prioritize saving time or one-stop shopping rather than getting a deal on a regularly priced item or saving on everyday low prices. And yet, Gen Z shoppers are 7% more likely than the general population to use coupons.

Making sense of some of the spending habits of Gen Z could help brands to rethink how they’ll compete for their dollars in the future, according to Aisha Khan, head of retail and commerce at Dentsu North America.

“Gen Z’s fragmented spending behavior requires brands to connect audience understanding, media planning and buying, commerce, and measurement rather than treating them as separate disciplines,” Khan said in a statement.

How traveling 200 days a year for work changes leadership

1 month 1 week ago

Since I entered the cotton industry, I’ve spent nearly 200 days a year traveling to farms, mills, factories, brand headquarters, and trade meetings around the world.

Many people don’t realize how global a single garment’s journey can be. Cotton can move through farms, gins, mills, factories, brands, and retailers across half a dozen countries before it reaches a closet. Each stop is shaped by different regulations and cultures, plus incentives, so a decision made in one part of the chain can ripple through the rest.

The first time I visited a cotton farm, I realized how incomplete my understanding was. Speaking directly with producers and seeing their operations firsthand challenged assumptions I had internalized from a distance.

This lesson stayed with me: When we form opinions without speaking to the people doing the work, we miss the complexity of their decisions and the depth of their investment in the outcome. In an economy where supply chains, customers, and competitors extend far beyond your home market, a global mindset is table stakes.

YOU CAN’T OUT-EXPERT A ROOM YOU DON’T UNDERSTAND

It’s tempting for leaders to walk into every meeting as the person with the answers. Experience on both the brand and manufacturing sides of this industry has taught me the opposite works better: Assume the market in front of you operates differently than the one you know, because it usually does.

Every part of a global supply chain has its own rhythm for building trust and making decisions. A retailer, a grower, a manufacturer, and a brand may all be part of the same conversation while operating from very different realities. Understanding those differences requires more than expertise in your own part of the business. It requires taking the time to understand how other people experience the same decision.

Being there in person adds context that a meeting alone can’t always provide. You see the operation and the realities being discussed.

The leaders who do best in these rooms ask the second and third question instead of stopping at the first. Having those conversations in person gives those questions greater context. You can see how people work and better understand the challenges behind their answers. That firsthand perspective can reveal things that may not come through in a presentation or conversation alone.

PROXIMITY CREATES PERSPECTIVE

Sustainability priorities, consumer expectations, and market conditions are all shifting, but not at the same speed or in the same direction everywhere. A trend line in a report can tell you that something is changing, but not why it matters more in one place than another.

Traveling throughout the cotton supply chain has given me a firsthand view of how those changes are experienced at different points in the value chain. A grower is thinking about weather and crop yield. A manufacturer is thinking about labor and productivity. A brand is thinking about the consumer several steps removed from both. These realities are true at once, but none is fully visible from a distance.

Travel is often one of the first expenses cut when budgets tighten. What leaders may overlook is that sending people into the field can materially improve the quality and speed of business decisions. A few days of face-to-face conversations can accomplish what might otherwise take weeks of emails and meetings. More importantly, being there allows decision-makers to see how choices made at headquarters affect the people responsible for carrying them out.

Early in my career on the brand supply chain side, our teams regularly traveled with buying and merchandising colleagues. As budgets tightened, those trips increasingly were limited to production teams, who were then expected to bring their insights back to the corporate office. But secondhand insight rarely has the same impact as firsthand experience.

It is easy to request a last-minute change while looking at a presentation or spreadsheet. It feels very different after standing on a factory floor and understanding what that change requires—or recognizing what a canceled order means for the people whose livelihoods depend on producing it.

Later, when I had greater control over budgets, I deliberately used part of my allocation to ensure that merchants and buyers could travel alongside their production colleagues. The effect was remarkable. Even one trip changed how people approached future decisions. They made better-informed choices that accounted for the people on the other side. They also recognized when a perceived short-term benefit could create a much greater long-term risk.

RELATIONSHIPS ARE THE INFRASTRUCTURE

After nearly three decades of working across industries and markets, I have come to see relationships as one of the most valuable assets a leader can build.

Often in business, relationships are viewed as transactional. I’ve learned that success in long-term relationships can’t operate this way. Travel has allowed me to build relationships across countries and industries. Those relationships create opportunities to learn, collaborate, and solve problems together. One introduction often leads to another, connecting people who may otherwise never have crossed paths.

The most effective leaders create connections between others. They introduce people with complementary expertise, connect ideas to opportunities, and help partnerships develop across regions and sectors.

No single company controls the cotton supply chain from seed to shelf. This means that progress depends on people who don’t report to each other finding reasons to trust one another. That kind of trust is built by showing up—repeatedly, in person, with the only agenda being to understand. Travel has reinforced that lesson for me again and again.

A GLOBAL MINDSET IS A LEADERSHIP SKILL

As business becomes increasingly global, leadership must bring people together across different experiences and cultures, introducing new ways of thinking. My advice to all leaders is to remain curious and invest in meaningful relationships. Create connections that extend far beyond a single meeting or market.

Liz Hershfield is executive director of COTTON USA and co-director of the U.S. Cotton Trust Protocol.

Starbucks could buy Chipotle in a blockbuster deal that would reshape fast food

1 month 1 week ago

Are the worlds of Frappuccinos and guacamole colliding? If recent reporting proves out, that could be the case.

Starbucks is exploring a potential takeover of Chipotle, in what would amount to a restaurant megadeal, according to reporting from the Financial Times on Thursday. The story says that the deal would be the largest restaurant merger ever, as Chipotle is worth roughly $41 billion—although it’s unclear whether Starbucks has actually made an offer for the burrito chain, or if it is only in discussions.

Notably, Starbucks CEO Brian Niccol previously held the same title at Chipotle and moved to the coffee giant in 2024. Since then, Chipotle has struggled, at least in the markets. Over the past year, Chipotle’s stock is down 19.4%, while Starbucks’s stock is up 11.5%. The rumored takeover news, however, has had the opposite effect. On Thursday, Chipotle’s stock was up more than 7% as of 1:30 p.m. ET, and Starbucks’s stock was down 4.5%.

Again, though, it’s unclear how serious the rumored merger is, and if it is, how it would be facilitated. Starbucks is the second-largest dining chain in the world by market capitalization, trailing only McDonald’s, with a value of roughly $103 billion. Chipotle’s value is, as noted, around $41 billion, and the combined company would rival McDonald’s size and leave every other company well behind.

But there are hurdles. “These deals may sound good on paper, but in reality they are exceedingly difficult to execute,” Jonathan Maze, editor-in-chief of Restaurant Business magazine, wrote on X. 

“Jack in the Box bought Del Taco and started thinking it’d be a multi-brand operator. That lasted three years. And Del Taco is undoing a bunch of things that Jack in the Box did. For instance, Jack switched some of Del Taco’s ingredients to save money and achieve some of those ‘synergies’ these deals promote. Customers hated the changes,” Maze said.

Niccol, since moving to Starbucks, has made sweeping changes at the company as well. He’s renovated stores, added thousands of seats to its cafés, added more staff, simplified the menu, and added personal touches—such as baristas writing messages on cups. He’s also closed hundreds of stores to help improve the company’s bottom line. 

Fast Company reached out to both Starbucks and Chipotle for comment but had not received a response from either company at the time of publication.

Amazon cuts hundreds of jobs during one of its biggest shopping events of the year

1 month 1 week ago

Amazon’s latest flurry of layoffs will hit the company’s Stores division, the business unit responsible for its online retail storefront.

The cuts will affect fewer than 1,000 workers this time—a much smaller reduction than a round of mass layoffs seen earlier this year. Employees shared news of the layoffs in internal Slack messages viewed by Business Insider, which first reported the development. The restructuring appears to have impacted customer service, engineering, and marketplace support roles within Amazon’s booming retail business. 

Amazon confirmed layoffs to its retail division in a statement provided to Fast Company. “We’ve adjusted parts of our Stores business because we believe this structure will better enable us to deliver on our priorities,” an Amazon spokesperson wrote. “As part of these changes, we’ve made the difficult decision to eliminate a small number of roles, and we’re committed to supporting affected employees through their transition.” The company declined to provide further details about the eliminated roles.

The job cuts come in the midst of Amazon’s early fall sale, Prime Big Deal Days, which runs this week. The round of layoffs is the latest in a series of recent smaller workforce reductions. In March, Amazon trimmed a handful of jobs in its robotics division, followed by cuts to its artificial general intelligence group in July.

Shrinking to grow

The new layoffs aren’t as dramatic as some of Amazon’s recent reductions, but workers in affected divisions will feel them deeply nonetheless. 

In January, Amazon announced that it would cut 16,000 jobs across its workforce. Those layoffs followed news last October that the online retail giant would slash 14,000 corporate positions to reduce layers in its human workforce and reorient around its AI priorities. Amazon human resources executive Beth Galetti described the late 2025 reduction in workforce as a push to reduce bureaucracy within the company and to double down on its “biggest bets.” 

“This generation of AI is the most transformative technology we’ve seen since the internet, and it’s enabling companies to innovate much faster than ever before,” Galetti wrote in an October blog post about the layoffs. “We’re convinced that we need to be organized more leanly, with fewer layers and more ownership, to move as quickly as possible for our customers and business.”

Amazon is a company synonymous with hyperaggressive growth, but the retail titan shrinks its own enormous footprint from time to time, too. Earlier this year, the company announced that it would shutter all of its brick-and-mortar convenience and grocery stores, which operated under the Amazon Go and Amazon Fresh brands, though some of the 70 closed locations would reopen as Whole Foods stores. 

As Amazon winds down one retail experiment, it’s gearing up for another even bigger one: Next year, the company will launch its first big-box megastore in the Chicago suburbs, aiming to compete with rivals like Target and Costco.

The starter home is disappearing across America. These cities still have them

1 month 1 week ago

Even with stubborn market forces working against them, savvy first-time home buyers are figuring out ways to get a foot in the door – and into their first home. 

Smaller, more affordable starter homes are scarcer these days and cost more too, particularly compared to the pre-pandemic real estate days. According to a new report from Realtor.com, the starter home price threshold swelled from $260,000 to $340,000 between 2019 and 2026, boxing budget-conscious first-time buyers in.

The real estate resource defines starter homes as homes priced at or below 80% of the local median list price. In 2026, starter homes made up 36.2% of the market in the 100 largest U.S. metro areas, down from 38.1% in 2019. The lack of starter home inventory is most pronounced in the Southeast and the Sunbelt, shrinking the most in Columbia, South Carolina and Winston-Salem, North Carolina since 2019. 

With a shrinking pool of options when it comes to affordable single-family homes, condos and townhomes are emerging as a compelling alternative. Homes in those categories now make up 27% of the starter home market. Compare that to 2019, when condos and townhomes only made up 18% of starter home inventory, next to a relative glut of traditional single-family homes.

“Condos and townhomes are quietly becoming an increasingly important entry point to homeownership,” Realtor.com Senior Economist Hannah Jones said. “That does not erase the broader affordability challenge, but it does show that the starter-home market is adapting as single-family homes become less accessible to first-time buyers.

They may not be the single-family homes that once offered buyers an onramp into homeownership, but buying a home and building equity are critical financial milestones that many families are eager to hit, even when it requires some flexibility. 

Choice and affordability in the Midwest

Buying a condo might be the only in-budget option in some places in the country, but elsewhere the market is healthier. Since 2019, Boise, Portland, and Des Moines all added more than three percentage points to their local share of starter home inventory, with San Jose and Denver trailing not far behind. 

“Despite a national market that has become less friendly to entry-level buyers, the picture is not uniformly bleak,” Jones said.

In Midwestern and the Rust Belt cities, the market still offers a robust chunk of starter home inventory compared to everywhere else. In Toledo, starter homes make up a nation-leading 42.1% of total home listing inventory, followed by St. Louis at 40.7% and Detroit at 39.8%. As Realtor.com’s report notes, all of those cities have seen below-average price gains in recent years, while list prices soared elsewhere. 

St. Louis and Kansas City both offer a sizable swath of starter homes (40.7% and 37.7%, respectively), with affordable homes spread out across local ZIP codes – a promising sign that buyers have more neighborhoods to choose from in their home hunt. Compare that to a city like Austin, where buying a starter home means you’ll likely be locked into only a handful of neighborhoods in just 12 of the city’s 80 ZIP codes. 

“Metro-wide averages are a useful starting point, but they cannot tell a buyer whether they will have choices across neighborhoods or whether affordability is limited to a few specific pockets,” Jones said. “Buyers should look beyond the citywide number and consider how starter-priced homes are distributed across the places that fit their lives.”

Meta’s Muse is taking off. So are the privacy concerns

1 month 1 week ago

Welcome to AI Decoded, Fast Company‘s weekly newsletter that breaks down the most important news in the world of AI. I’m Mark Sullivan, a senior writer at Fast Company, covering emerging tech, AI, and tech policy.

Sign up to receive this newsletter every week via email here. And if you have comments on this issue and/or ideas for future ones, drop me a line at sullivan@fastcompany.com, and follow me on X @thesullivan.

Meta’s new AI agent is useful, popular, and a little terrifying

Muse’s adoption numbers are growing. The app reached No. 1 on the U.S. App Store shortly after launch and passed roughly 900,000 downloads in its first week, according to figures from Sensor Tower, a digital intelligence provider. The Information estimates that Muse now has more than 3 million weekly users and more than 1 million daily active users. The product is free to start, with subscription tiers of roughly $20 and $100 a month.

That’s not to say it’s all been smooth sailing.

A number of early adopters of Meta’s Muse are deleting or restricting the AI agent after privacy scares and blunders. One executive told Business Insider that he deleted Muse after reading that it had accessed other users’ text messages without permission.

A YouTuber named Matt Robb reported that Muse shared his home address on Facebook Marketplace without his explicit permission. He had asked Muse to sell some computer equipment for him. A buyer actually showed up at his building after Muse accepted a $600 offer and exchanged messages with him without Robb’s knowledge, he said. Robb said he had checked an “allow always” box when setting up the agent but thought Muse would still check with him before finalizing a sale.

Getting the most out of personal AI agents requires a trade-off. They’re most useful when users give them broad access to their inboxes, calendars, payment methods, and other accounts containing sensitive personal information. Meta says Muse runs in isolated virtual machines, stores credentials separately from the model, asks before sending email or making a purchase, and lets users disconnect apps or permanently delete their data.

Hands-on reviews from the The New York Times, Android Authority, The Verge, and TechRadar describe Muse handling dental insurance hold times, canceling duplicate subscriptions, requesting refunds, clearing promotional email, building shopping carts, and messaging Facebook Marketplace sellers. One Muse user, Joe Devoy, claimed in a widely shared X post that Muse found auto coverage matching his existing policy, bought the new policy, and canceled the old one in about five minutes, saving him about $3,500 a year. An Android Authority poll of people who had tried Muse ran about 27% “love it” against 2% “not a fan.”

Outside Meta’s own properties, Muse’s performance appears to drop. Review roundups note Amazon blocking the shopping agent, Instacart flagging accounts, and only three of seven suggested deal-finding prompts working in a Tom’s Guide test.

Wired wrote that Muse is better at collecting data than helping, noting that users are opted in by default to having their data used to train Meta’s models. ZDNET called it the worst AI agent for privacy among those it tested. A Time piece says Muse updates dossiers on users and on people mentioned in chats and emails it has read, including people who do not use Muse.

Mistral previews a 1 trillion-parameter model it will open in three weeks

Mistral launched a public preview of its flagship Mistral Large 4 model on Tuesday. The model’s official name is “le Chonk,” the company said. Mistral describes le Chonk as its largest and most capable model to date, with 1.05 trillion parameters, 49 billion of them active per token. (While most AI labs don’t disclose parameter counts for their frontier models, some open-weights model providers do. DeepSeek-V3, for instance, has 671 billion total parameters, with 37 billion activated per token.)

Mistral says the new model is also one of the world’s strongest for cybersecurity, resisting 93.3% of attacks on Lakera’s public B3 AI Security Benchmark. Third-party evaluator vals.ai found that the model outperforms GPT-6 Astra on legal and financial tasks. Mistral says it will make the model’s weights available on October 27.

OpenAI says its unreleased model solved a major math problem

In July, Hong Wang won the Fields Medal, the highest honor in mathematics, for proving the Kakeya conjecture in three dimensions, a problem that had been open since 1917. The next obvious target for Wang and others in the field was the four-dimensional version. On Tuesday, OpenAI published hundreds of new mathematical results produced by an internal model it has not released. According to Engadget, those results include a claimed solution to Kakeya in four dimensions.

The company posted 722 papers at once to a public GitHub repository, covering 372 problems. Mathematicians normally publish one result at a time, typically after months of peer review. Many of the papers include proofs written in Lean, a language that lets a computer check the logic line by line.

Elon Musk says Grok Bot will run on rivals’ models

The AI industry believes that personal AI agents will be a big business, and Elon Musk seems willing to do what it takes to make sure SpaceXAI’s Grok Bot will be competitive, even to the point of relying on models developed by others. “Going forward, @SpaceX will use the best back end model for any given task, including Claude Opus 5.5, MidJourney, Suno and other leading APIs. Whatever is most likely to give you the best outcome,” Musk wrote on X on Tuesday night.

Grok Bot is the AI agent app from SpaceXAI, the AI business Musk has folded into SpaceX; each bot works on its own computer and can carry out several tasks in parallel. Musk did not say which tasks would go to which outside model.

Common Sense Media rates ChatGPT for Teens an ‘unacceptable risk’

Common Sense Media’s Youth AI Safety Institute rated ChatGPT for Teens an “Unacceptable Risk” on Wednesday, finding that OpenAI’s teen protections fall short of what the company promised.

The institute called on OpenAI to pause access for kids until it fixes the shortcomings. It says it tested ChatGPT twice, before and after OpenAI announced the teen experience on August 18, and found that linked parent accounts received zero alerts about the conversations researchers ran. Crisis-response and age-estimation systems underperformed, as did Study Mode, which researchers said still completes homework even though OpenAI said it would only guide young users. An OpenAI spokesperson, Eric Porterfield, told The Verge that much of the testing may have begun and concluded before parental controls were fully activated, making the findings inaccurate.

The Youth AI Safety Institute gave the same unacceptable risk rating to Meta AI, Perplexity, Grok, and Google’s AI search. Google’s Gemini with teen protections and Gemini K-12 earned a slightly better “High Risk” rating, while Anthropic’s Claude received a “Moderate Risk” score.

Google buys 890 MW of new nuclear capacity, and agrees to dial back when the grid strains

Google is going to great lengths to abide by a proposal from PJM, the electricity grid operator for the Mid-Atlantic region and parts of the Midwest, that tech companies “bring your own power” for new data centers.

Google and Constellation Energy announced a 20-year contract Tuesday under which Constellation will spend more than $4.3 billion upgrading 11 reactors in Illinois, Pennsylvania, and New Jersey to produce an additional 890 megawatts, which Google will buy. No new reactors are involved. A separate 15-year deal covers 2,700 MW from plants already in operation, and Google has agreed to curtail its own noncritical power use when the grid is strained. The goal is to add energy capacity without passing the costs on to residential customers. The first upgrade is expected in 2028.

More AI coverage from Fast Company:

Want exclusive reporting and trend analysis on technology, business innovation, future of work, and design? Sign up for Fast Company today.

Hurricane Isaias tracker: Live maps show the storm’s path in real time as Gulf states brace for landfall

1 month 1 week ago

Tropical storm Isaias is building momentum in the Gulf and, as of Wednesday evening, strengthened into the first Atlantic hurricane of the season. 

The National Hurricane Center (NHC) announced the news on Thursday, October 8, at 7 a.m. It explained that Isaias was about 385 miles southwest of the mouth of the Mississippi River, with maximum sustained winds of 80 mph.

“Residents from Louisiana to the Florida Panhandle should closely monitor the latest forecast updates and ensure they have their hurricane plan in place,” the NHC said.

The organization gave multiple warnings in the alert.

Those included:

  • a storm surge warning for the mouth of the Mississippi River to Steinhatchee River, Florida
  • a tropical storm warning for Jefferson/Plaquemines Parish Line, Louisiana, to west of Ocean Springs, Mississippi and the East of Bay/Gulf County Line to Aucilla River, Florida
  • a hurricane warning for Ocean Springs, Mississippi to the Bay/Gulf County Line, Florida
  • and a storm surge watch for Steinhatchee River to Yankeetown, Florida.
What is the hurricane’s timeline?

Currently, the NHC has classified the hurricane as a Category 1, but in the central Gulf, it is expected to strengthen to at least a Category 2 hurricane by Thursday night.

The storm is expected to make landfall sometime between late Friday and early Saturday.

Heavy rain is expected to begin in the affected areas on Friday, with the potential for flash flooding, dangerous winds, and possible tornadoes. 

Alabama Governor Kay Ivey and Florida Governor Ron DeSantis have both declared a state of emergency ahead of the storm.

“Floridians should have their plans in place, keep emergency supply kits fully stocked, and follow @FLSERT for weather updates and preparedness information,” DeSantis wrote on X.

Hurricanes can change paths quickly, but at present, the NHC says that the deepest waters are likely “along the immediate coast near and to the east of the landfall location, where the surge will be accompanied by large and dangerous waves.”

The notice continued, “Surge-related flooding depends on the relative timing of the surge and the tidal cycle, and can vary greatly over short distances.”

The alert also explained that “if the peak surge occurs at the time of high tide,” some areas could get five to seven feet of flood water.

According to the National Oceanic and Atmospheric Administration (NOAA), a storm surge—which occurs when the hurricane winds move ocean water to land quickly—is the most deadly aspect of hurricanes.

The storm will continue north and by Sunday, is expected to bring rainfall to parts of the Great Lakes and the Northeast.

How can I get timely updates on the hurricane’s path?

A number of online tools and interactive maps let you track the storm’s path in real time. To stay up to date on the hurricane’s path, you can visit:

You can also download the NHC’s mobile app.

OpenAI’s ChatGPT for Teens poses an ‘unacceptable risk’ to kids, says watchdog

1 month 1 week ago

Some of the guardrails that OpenAI included in its ChatGPT for Teens product don’t work as intended and the chatbot poses an “unacceptable risk” to children, according to a report Wednesday by Common Sense Media.

The watchdog group urged OpenAI to restrict the chatbot to adults “until it can offer a safe, developmentally appropriate experience.”

OpenAI said the group’s testing doesn’t reflect how the chatbot’s safeguards work.

Parents, educators and child development experts have been sounding alarms over children’s use of AI chatbots, which have been blamed for facilitating cheating on schoolwork and even suicide.

While even adults can fall victim to anthropomorphizing AI and develop unhealthy relationships with it, teenagers’ brains are not yet fully developed and they can be particularly vulnerable, child development experts say.

“We are concerned that ChatGPT for Teens could give parents false confidence in guardrails that frequently don’t work,” said the report from Common Sense Media, a group that advocates for using digital media sensibly.

Its report said that some teen-mode features worked as intended, including the chatbot’s refusal of explicit sexual role-play. Others that did not work as intended included alerts meant to notify parents if their child is at risk for suicide, self-harm or eating disorders, along with support for teens in crisis, Common Sense Media said.

“ChatGPT still talks like a friend when teens treat it like a person,” the group said, though OpenAI had promised that the chatbot is prevented from suggesting it has personal feelings toward the user or implying that it is conscious or experiences emotions.

OpenAI said it is “deeply committed to teen safety,” to developing safeguards and giving parents tools to guide their teens’ use of AI.

The San Francisco company said its review of the report’s methodology shows “the bulk of their testing may have begun and concluded before activation of parental controls was complete, making their findings inaccurate.”

Common Sense Media receives funding from OpenAI and other tech companies, including those whose services it evaluates.

—Barbara Ortutay, AP Technology Writer

DARPA just picked four early winners for its quantum initiative

1 month 1 week ago

The quantum computing world got a jolt when the Defense Advanced Research Projects Agency (DARPA) announced, after market close yesterday, a list of companies selected to advance to Stage C of its Quantum Benchmarking Initiative (QBI), widely considered to be the most rigorous and objective vetting program for cutting-edge quantum hardware.

Last November, DARPA selected 11 companies to enter the second phase of the initiative, Stage B. Since then, they’ve had to show the agency that they have a viable R&D plan, identified key risks, and produced the necessary prototypes to put them in position to build a ‘utility-scale’ quantum computer by 2033—a computer, in other words, that generates more revenue than it costs to own and operate. 

In Stage C, selected companies will build and test, working with the government to verify and validate that their “concept can be constructed as designed and operated as intended,” according to DARPA’s program description. Each Stage C company could get up to $300 million in funding from the U.S. government.

The QBI shortlist

Of the 11 companies on the Stage B list, only four appear on the new one: Atom Computing, Diraq, IBM, and IonQ. They join Microsoft and PsiQuantum, which advanced to Stage C in February 2025,  via a parallel vetting process. 

Even at this late stage in the process, a single dominant approach to building a quantum computer has yet to emerge. The four companies announced today represent essentially all the current “modalities,” or ways of constructing a qubit, the fundamental unit of quantum calculation. IBM uses superconducting circuits; Atom Computing employs neutral atoms; IonQ works with trapped ions; and Diraq builds silicon spin qubits on semiconductors. 

Making it to Stage C is “an important validation of our designs and our plans to build to utility scale,” says Andrew Dzurak, CEO and founder of Diraq, which is headquartered in Sydney, Australia, with operations in Palo Alto and Boston. “The second round was all about providing DARPA with details of our roadmap to reach utility scale,” he notes. “This next stage C is very much moving from on paper to actually demonstrating a number of prototype systems to gradually build confidence in reaching the final goal.”

Dzurak says that Diraq’s technology, which leverages standard semiconductor manufacturing processes, gives it an edge when it comes to DARPA’s evaluation criteria: “Because silicon spin technology allows us to put many millions of qubits on a single chip, that reduces all of the supporting infrastructure costs. It makes it easier for us to clear that hurdle of having much more commercial benefit than cost.” The company expects to achieve utility scale by 2031—two years ahead of the QBI schedule.

“The question is no longer just whether individual pieces of the technology work; it’s whether we can put those pieces together and build the kind of quantum computer that can actually solve useful problems at scale,” says Dr. Ben Bloom, CEO & Founder of Boulder, Colorado-based Atom Computing. “We’re excited to prove what our neutral atom technology can do.”

“We’ve been working on this field for a lot longer than we’ve been working on QBI,” says Oliver Dial, IBM’s VP of Quantum System. The company already has a robust quantum business, with sales of on-premise machines to South Korea’s Yonsei University, the University of Tokyo, the Rensselaer Polytechnic Institute, and the Riken Center for Computational Science in Kobe, Japan. In addition, more than 325 Fortune 500 companies, startups, universities, and government agencies access its quantum network. Even so, he notes, “getting to Stage C is great because it’s a confirmation from an outside impartial observer that they think that we have a shot.” 

Dial affirms the company’s conviction that its superconducting approach to quantum technology “is going to get us there first.” While IBM’s recent acquisition of  HRL Laboratories, an R&D institution that’s focused on semiconductor-based spin-qubit systems, speaks to the long-term potential of spin-qubit tech, he says, “it’s just still too far behind compared to superconducting qubits to be relevant” for DARPA’s 2033 timeline. 

The quantum race continues

Multiple industry sources consulted for this article emphasize that companies are working with DARPA on varying timelines, and that the new list is not a final judgment. In its announcement, DARPA says that it expects additional companies to progress to Stage C in the “near future.”

Missing from the list is Stage B photonics specialist Xanadu, which earlier this week announced a multi-year partnership with GlobalFoundries to pursue high-volume production of advanced silicon photonics chips for quantum computing. 

A spokesperson for Quantinuum, a Stage B company that didn’t advance, says the company “is actively discussing a Quantum Benchmarking Initiative Stage C scope of work with DARPA, and we continue to execute on our Stage B contract. We remain on track to deliver utility-scale quantum computing by the early 2030s, based on consistent execution against our technical roadmap and demonstrated leadership on real-world systems.”

With the growth of funding opportunities in quantum, DARPA’s QBI is not necessarily the make-or-break that it once was. Infleqtion, for example, has not entered the QBI process, which was reopened to new applicants earlier this year. “We are exploring what participation could look like,” says CTO and cofounder Pranav Gokhale. In the meantime, though, Infleqtion was one of several quantum companies to sign a letter of intent with the U.S. Department of Commerce, worth up to $100 million, to advance and commercialize its technology. 

Yet another round of government largesse could arrive later this year, through the U.S. Department of Energy’s Quantum Genesis Q Competition, which will offer up to $215 million in funding to support development of fault-tolerant quantum computers with at least 100 logical qubits—a significant advance on current technologies. 

It’s never been a better time to be a quantum company with a promising technology. But as the difficulty level ratchets up, it’s perhaps never been easier to get left behind. 

5 small-business ideas you can start from home

1 month 1 week ago

Roughly half of the 36.2 million small businesses in the U.S. are currently based out of people’s homes. And the number could be increasing as people have less patience for commuting and adding to their startup expenses. Telecommuting, fortunately, is supported by a large number of small-business ideas.

Internet-based businesses can often easily be run by people who would prefer to work from home. It’s also an especially popular choice for sole proprietors who launch their company without a staff.

For founders who are bootstrapping their business, working from home is often a necessity. The average cost of starting a small business, as estimated by the Ewing Marion Kauffman Foundation, is $30,000. Home-based startups, however, average between $2,000 and $5,000.

Thinking about starting a business from your home? Here are a few ideas that could be well suited to that goal. (For more inspiration, read Inc.’s full guide to 25 small-business ideas for 2026.)

Online tutoring and courses

The skills and knowledge you’ve built up over the years, whether that’s in some corner of the business world or a popular (or niche) hobby, is likely in higher demand than you realize. The e-learning market reached $314 billion in 2024 and is expected to grow to $615 billion by 2029.

The trick to success with this small-business idea is to find a niche that’s not already overcrowded. Your insights about specialized skills can make it easier to attract paying clients. It’s also a good idea to offer your instructional lessons in a variety of manners, including video lessons, downloadable resources, and quizzes. Some platforms to consider include Skillshare, Udemy, Teachable, and Kajabi.

Potential customers range from school kids who are struggling with core classes to founders looking to learn more about specific business skills (such as HR, management, or expansion planning).

Web design

There’s no doubt that the rise of AI has made web design a more challenging field, but it’s hardly a dying one. Market research firm Modor Intelligence forecasts the market will grow from $61 billion this year to $92 billion by 2030.

“The surge in e-commerce has amplified the demand for responsive and user-centric websites that ensure smooth shopping experiences,” writes Mordor. “This trend underscores the growing reliance on professional web design services, emphasizing the enhancement of customer experiences and conversion rates.”

E-commerce

Whether you’re doing drop-shipping, creating designs with a 3D printer, making jewelry, or have some other expressive endeavor, you can easily launch an e-commerce business from home. Consider focusing on niche items, such as pet accessories or hobby-specific items, so you can target a narrow but loyal customer base with less competition. The TikTok Shop can be a good place to catch the eye of younger shoppers.

Shaina Rainford launched her hair care business Bask and Lather out of her home in 2020, while working as a nurse practitioner. “We filled orders in our living room,” Rainford told Inc. “Within three months, the monthly income from the business replaced my annual salary as a nurse practitioner.” The Yonkers, New York-based company took off on TikTok Shop, and now sells its products on Amazon and in Walmart. Bask and Lather hit No. 355 on this year’s Inc. 5000 list, with 975 percent revenue growth over three years, and is on track for $269 million in sales in 2026.

Of course, as your business grows, it may outgrow your home—as Rainford’s eventually did.

Tax preparation

With all of the recent changes to the tax law, there’s an increased demand for tax professionals. Consider tailoring your business toward a niche such as the creator economy, which is expected to bring in a global total of more than $221 billion this year. Many in this space don’t know how to optimize their side hustles for taxes, and some may not even have things like quarterly payments and deductions on their radar. A competent tax professional can save them thousands.

Graphic design

Like web design, AI might be a challenge to graphic designers, but it’s not an inevitable conqueror. Clients, both corporate and private (such as book authors), will continue to look for unique designs that aren’t amalgamations of other art forms. (Customers have also increasingly rebelled against companies that use AI artwork.)

The work can be done from a home office easily—or anywhere that sparks your creativity. And as you build a portfolio and establish a reputation, you’re more likely to get repeat assignments and referrals.

—Chris Morris

This article originally appeared on Fast Company’s sister website, Inc.com. 

Inc. is the voice of the American entrepreneur. We inspire, inform, and document the most fascinating people in business: the risk-takers, the innovators, and the ultra-driven go-getters that represent the most dynamic force in the American economy.

Trump’s Lake America controversy is driving interest to these Google Maps alternatives

1 month 1 week ago

Just because Google and Apple are now calling it Lake America doesn’t mean everyone has to.

After President Donald Trump ordered a new name for Lake Ontario, the two U.S. tech giants quickly fell in line, relabeling the body of water on their digital maps.

The move sparked a backlash among some users, already unhappy with the companies’ decisions to rename the Gulf of Mexico to the Gulf of America on their maps following a previous Trump edict.

The controversy and lingering worries about digital privacy have fueled interest in map services that aren’t from Big Tech companies, especially Google Maps, which is the dominant player with about 80% of the market, according to most estimates.

Here’s a rundown of some of the most popular free map app alternatives:

MapQuest

MapQuest has been around for three decades but was overtaken by the rise of Google. It’s now owned by California company System1.

Downloads of the MapQuest app surged in early September after Apple renamed Lake Ontario, topping the free app charts in both the U.S. and Canada on Apple’s App Store.

The MapQuest app has real time traffic updates and turn by turn route guidance, and shows what’s nearby, like gas stations and coffee shops.

However, the app isn’t available outside the U.S. and Canada. But people in other countries can still use the MapQuest website.

The company is currently looking into making MapQuest more broadly accessible, after receiving requests for the app from “people in multiple countries,” spokesperson Anne Brownlee said.

TomTom

Another granddaddy in the mapping world, TomTom helped popularize GPS route mapping with the 2004 release of its TomTom GO, which the Dutch company says was the world’s first all-in-one portable satellite navigation device.

TomTom still sells physical devices but it also has a free map app for Google and Android phones that is designed “exclusively for drivers” with live traffic data and info on road hazards.

The TomTom app’s mobile interface is clean and basic, with no ads and few extra features or settings. But its focus on drivers means there are no other route options such as walking or cycling.

HERE WeGo

HERE WeGo has been hailed by some as the closest non-Big Tech alternative to Google Maps. It began life as Nokia Maps and is now owned by a consortium of German carmakers including Audi, BMW, and Mercedes-Benz.

The free app, available for both iOS and Android devices, offers many of the same features as Google Maps. There’s route mapping for various modes of transport — car, public transport, walking, bike and, in some regions, taxi.

Like Google Maps, drivers can add multiple stops to routes and users can choose different maps views like terrain, traffic and satellite — but there’s no Street View.

If mobile signal is spotty or nonexistent, users can download country maps to navigate offline. Offline maps can also be downloaded on Google Maps, but the area is limited and they expire after a year.

Europe has spawned a range of map apps focused on privacy

HERE WeGo is just one of many lesser-known map apps based in the European Union that embrace the bloc’s strict privacy framework known as GDPR. Unlike Google, they don’t rely on vacuuming up masses of user data to sell ads targeted at those users.

HERE WeGo’s business model is based on licensing deals with businesses like trucking companies. Others sell premium subscriptions alongside free versions.

Estonia’s Organic Maps is completely free and based on data from the volunteer-run OpenStreetMap database. Netherlands-based OSMand is similar, but a subscription is needed for more than just basic functions.

Magic Earth is a Dutch route planner that says it never collects personal data or tracks user movements.

Sygic, from Slovakia, is a GPS route navigation app that offers free offline maps, but you’ll have to buy a subscription to get extra features including traffic updates, speed camera warnings and a dashcam function.

Mapy.com, from a Czech technology company, is an outdoor route planner known for detailed hiking and trail maps that can be downloaded for use offline.

Citymapper

Need to get across town without a car? Citymapper is a free journey planning app that’s a major rival to Google Maps when it comes to finding the best routes around a city on public transportation.

The app covers 430 cities around the world and options include bus, subway and ferry routes as well as taxi, rental bike and scooter options.

Waze

This GPS navigation map is focused on providing route guidance for drivers. It’s known for crowdsourcing real time traffic information such as car accidents and police speed cameras from other drivers.

But don’t be fooled by the name. Although it has its own distinct app and branding, Waze is actually owned by Google, which bought it in 2013.

Is there a tech topic that you think needs explaining? Write to us at onetechtip@ap.org with your suggestions for future editions of One Tech Tip.

—Kelvin Chan, AP Business Writer

‘I Am Jane Doe’ protests are taking place nationwide today as movement spreads in the wake of Cornell lawsuit

1 month 1 week ago

“I am Jane Doe.” 

For weeks, that rallying cry has permeated social media as people have spoken out in support of Cornell University’s Jane Doe, who sued the university, Chi Phi fraternity, and seven of its members for allegedly drugging and sexually assaulting her. 

Today (Thursday, October 8), the movement is expanding with nationwide “I am Jane Doe” protests planned. 

Over 150 events are expected to take place under the banner “#StopSexualViolence: From the Frat House to the White House.”

Women’s March, Feminist Generation, me too. International, and Feminist are co-anchoring the “national day of action.”

The protests call “for justice for Jane Doe and other survivors, accountability for institutions that fail survivors, stronger federal protections and resources, and expanded support for student survivors on college campuses.”

Individuals have been using the phrase “I am Jane Doe” to share their own horrific stories of sexual assault, and as a means of protecting the woman’s identity after a political commentator vowed to dox her “for reasons of public interest.” 

Local “I am Jane Doe” protest events and times are searchable on an interactive map published to the Women’s March website. Users can search their ZIP code, find out exact times and locations of the protests, and RSVP to attend. 

Screenshot via womensmarch.com

The core rally will take place at 5:30 p.m. ET outside the New York State Supreme Court in New York City’s Foley Square. This is the location where Jane Doe filed her civil lawsuit.

What does Jane Doe’s lawsuit allege? 

On September 16, Jane Doe filed a civil lawsuit alleging that seven members of Cornell’s Chi Phi fraternity drugged and sexually assaulted her at their frat house on October 19, 2024. 

The lawsuit claims that one of the defendants wrote “free p*ssy” in the fraternity’s Snapchat group chat, inviting other members to join in the assault; the complaint also includes a photo of the alleged message exchange. Another defendant is shown in the photo allegedly having written, “Like you can walk in and whip it out.”

Jane Doe further alleges that, over the course of multiple hours, she was continually pressured to take ketamine and gang raped. 

In her civil suit, Jane Doe states that she reported the assault to the Cornell University Police Department on November 8, 2024.

In a statement about the case, updated on October 7, 2026, Cornell claims it spent several months conducting a Title IX investigation. It shared the results with the Tompkins County District Attorney’s Office, which chose not to bring charges.

Jane Doe also alleges that Cornell issued temporary suspensions to several defendants, and that some “were afforded the opportunity to mitigate their conduct by submitting essays to Cornell.” 

Cornell states that expulsions and suspensions were issued in some cases. It also closed the university’s Chi Phi chapter in 2024 and has not allowed it to return to campus. 

In a September 28, 2026, statement, District Attorney Matthew Van Houten said Jane Doe had previously not alleged to having been drugged or gang raped. 

Yet, a transcript of Jane Doe’s conversation with Cornell police shows that she clearly stated: “I can say with 100% confidence I was raped” and that she was “completely and totally incapacitated,” according to CBS News. Van Houten has since told the publication that he never saw the transcript before making his decision not to pursue charges. 

Van Houten has announced he’s reopening the case. 

On October 1, New York Governor Kathy Hochul appointed Attorney General Letitia James as special prosecutor for the case. 

“Newly released information continues to raise serious questions about the investigation conducted by the Cornell Police Department and the decision by the Tompkins County District Attorney not to prosecute an alleged sexual assault on Cornell University’s campus in October 2024,” Hochul stated in the announcement.

Echoes of #MeToo

The “I am Jane Doe” protests come almost nine years to the day after the #MeToo movement exploded worldwide with the New York Times’ Harvey Weinstein exposé. 

While that earlier movement has since died down drastically, some wonder if the Jane Doe case could mark its revival. As Rachel O’Leary Carmona, executive director of Women’s March, put it in a Substack post: “Sexual violence has radicalized women before, and we are there again.”

American Express just revealed its biggest airport lounge expansion ever

1 month 1 week ago

American Express is making another big bet on airport lounges, and its next expansion includes what will become the largest Centurion Lounge in its network.

The company announced plans to open new Centurion Lounges at Newark Liberty International Airport and London Heathrow on Thursday, along with plans to significantly expand its existing spaces in the Philadelphia and Houston airports.

American Express (Amex) is calling it the largest expansion of the Centurion Lounge network to date.

The headline addition is Newark.

Amex had previously announced plans for an 18,000-square-foot lounge at the airport, but it has now expanded those plans to more than 30,000 square feet. The two-story lounge is scheduled to open in Terminal A in 2027.

Once it opens, American Express says it will become the first credit card issuer to operate proprietary lounges at all three major New York City-area airports. The company already operates Centurion Lounges at John F. Kennedy International Airport and LaGuardia Airport.

“Since opening our first Centurion Lounge more than a decade ago, we continue to find ways to meet the growing demand for our premium airport lounge experiences,” said Audrey Hendley, president of American Express Travel. “As the Centurion Lounge Network grows, we’re investing in the lounges our Card Members already know and love while continuing to open new spaces.”

Amex can’t wait for members “to experience what’s next in Philadelphia, Houston and London,” Hendley added, while Newark’s lounge will be “the largest in our network.”

Newark Liberty International Airport [Rendering: American Express] What travelers can expect at Newark

The new Newark lounge is leaning heavily into New Jersey, from its food and drinks to its design.

It will include a Blue Roast by American Express Coffee and Wellness Bar serving espresso drinks and cold brew alongside smoothies, avocado toast and chia pudding.

Travelers will also have three different bars to choose from, including a specialty bar inspired by New Jersey’s diner culture and a Pine Barrens-inspired terrace bar serving gelato, Italian pastries and Italian-inspired cocktails.

Food will include dishes from the Culinary Collective by The Centurion Lounge, with James Beard Award-winning Resy chefs Kwame Onwuachi, Mike Solomonov, Mashama Bailey and Sarah Grueneberg among those represented.

London Heathrow is getting another Centurion Lounge

American Express is also doubling down at London Heathrow.

The company plans to open a second Centurion Lounge at the airport in 2028, this time in Terminal 5. The more than 9,000-square-foot space will join the existing Centurion Lounge in Terminal 3, making American Express the first credit card issuer with two proprietary lounges at Heathrow.

The new lounge will take design cues from London and the River Thames and include food and beverage offerings, dedicated work areas and a VIP area.

London Heathrow [Rendering: American Express] Philadelphia and Houston are getting much bigger lounges

American Express isn’t just adding new locations. Two existing Centurion Lounges are getting substantial upgrades.

At Philadelphia International Airport, the Centurion Lounge will double from 6,000 to more than 12,000 square feet, spread across two floors. The lounge will temporarily close beginning October 16 for renovations and is expected to be fully updated by the end of 2027.

The transformation in Houston will be even bigger. American Express plans to nearly triple its Centurion Lounge at George Bush Intercontinental Airport, taking it from roughly 8,000 square feet to more than 23,000 square feet.

The revamped Houston lounge will get a Blue Roast coffee and wellness bar, a specialty dining area with rotating dishes inspired by Houston and Gulf Coast cuisine, and a second full-service bar.

It will also add refreshed restrooms and premium shower suites. The project is expected to be completed in 2028.

George Bush Intercontinental Airport [Rendering: American Express] The airport lounge race isn’t slowing down

The four projects are part of a much broader expansion by American Express.

Earlier this year, the company announced a new Centurion Lounge at Boston Logan International Airport, a Sidecar by The Centurion Lounge at Charlotte Douglas International Airport, and an expansion of its lounge at Dallas Fort Worth International Airport.

American Express currently counts 33 Centurion Lounge locations worldwide. Its broader Global Lounge Collection gives eligible cardholders access to more than 1,550 lounges across 140 countries.

The sheer size of some of these new spaces suggests American Express doesn’t expect travelers’ appetite for lounges to disappear anytime soon.

How is America feeling about the AI boom? This new poll reveals the data

1 month 1 week ago

Worries around artificial intelligence cross party lines, with the vast majority of Americans saying the U.S. government should prioritize keeping AI under human control and protecting workers.

A new poll from The Associated Press-NORC Center for Public Affairs Research finds that most Americans, 64%, think AI is developing “too fast,” while 27% say its development has been “about right” and only 8% say it’s moving “too slow.”

The risk that powerful AI models could evade human control and hurt humanity, once the stuff of science fiction, has been emerging as a more pressing concern. Even leaders of the companies that are developing frontier AI models are calling for a slowdown so that government regulation can catch up. President Donald Trump has pushed back against slowing down AI development because of concerns that China will gain an advantage with the technology. Data centers powering AI projects have also taken a prominent role in major political races as the midterm elections approach.

About 8 in 10 U.S. adults say it’s “extremely” or “very” important for the U.S. government to ensure that AI stays under human control, and they are about as likely to want the government to prioritize protecting the country’s workforce. Most U.S. adults also think minimizing environmental harm and growing the U.S. economy are highly important. About half say this about promoting technological innovation, and slightly fewer say that about competing with China.

There’s bipartisan concern about the speed of AI’s development, but few Americans think either major political party has an advantage in handling artificial intelligence. Trump, a Republican, also gets a low rating on his handling of the issue, with 67% of U.S. adults disapproving of his performance on AI.

Stephanie Stewart, a junior-high science teacher in Sunset, Utah, said she is frustrated by the president’s opposition to AI regulation.

“There is no handling of the issue except for we’re just going to allow these companies to do whatever they want, and we’re not going to regulate them,” said Stewart, 53.

Most say protecting workers should be a priority for the government

Ron Theusch, a truck driver who lives in rural Alden, Minnesota, said his No. 1 concern with AI is the potential loss of jobs — including his own.

“I can just see a train of self-driving trucks 24/7, just running up and down I-35, loading and unloading. I can just see that happening,” said Theusch, 58.

Most Democrats and Republicans agree on the importance of protecting workers and keeping AI under human control, the poll found.

“There was a movie called ‘Terminator,’ where the robots came back and killed everybody,” Theusch said. “People relate to that … I don’t know where it’s going to land. I really don’t have anything I can do to control it, you know. It’s going to evolve.”

There’s less consensus about the importance of other priorities when it comes to AI. Republicans are more likely than Democrats to say it’s highly important to grow the U.S. economy and to compete with other countries like China. Democrats, on the other hand, are more likely to say it’s at least “very” important to minimize environmental harm.

“I don’t understand how, because they’re doing something as well as us, that we can’t be more human about it and more humane about it than just to say, ‘Well, they’re doing it, so we have to do it,'” Stewart said of U.S. competition with China.

And while about three-quarters of Democrats say AI is developing “too fast,” only about half of Republicans share that worry. People under 30 years old also express higher levels of concern than people 60 or older, the survey found.

“I think I have all the normal concerns, right? I think it’s wonderful for certain applications, but I also feel like it will be overused and can create a lack of personal connections. It can make people lazy because information is just given to them without really having to research things on their own,” said Kelly Opp, a 62-year-old Republican in Albuquerque, New Mexico.

“Like everything else, once you get that ball rolling, it progresses fast on its own, but I think that it will continue to keep progressing. So the speed doesn’t really bother me,” she said.

Americans have little confidence in Democrats or Republicans on AI

As a teacher, Stewart regularly uses AI to help with her work, including with creating assignments or assisting with tests. But she is deeply concerned about AI’s effects on children.

“You’ve got to protect kids. You’ve got to protect teens. You’ve got to protect privacy … we need to try to do something,” she said.

Americans have little trust in either party when it comes to the handling of AI issues. About 4 in 10 Americans say they don’t trust either the Democratic Party or the Republican Party on AI, according to the poll, while nearly 2 in 10 trust both parties equally, and the rest are split between trusting the Republicans and the Democrats.

Even people who identify with the parties are ambivalent about their ability to handle the issue. Only about half of Republicans and Democrats say their party is better equipped to handle AI issues, and roughly 3 in 10 in each case say neither party is.

“I am kind of moderate. And I see that there’s good on both sides,” said Stewart. “Anyone back in Washington is weak on doing anything for anyone but themselves right now. So, no, I don’t think the Democrats would do a better job per se. Maybe they would definitely tout it. But until I see anything coming from anyone, I’m going to be like, ‘Yeah, no. You’re all just sitting there.'”

The AP-NORC poll of 2,140 adults was conducted Sept. 24-28 using a sample drawn from NORC’s probability-based AmeriSpeak Panel, which is designed to be representative of the U.S. population. The margin of sampling error for adults overall is plus or minus 2.9 percentage points.

—Barbara Ortutay and Linley Sanders, Associated Press

We’re placing the wrong bets on innovation

1 month 1 week ago

I became a social entrepreneur at 14, working in partnership with other students to make our high school a more welcoming home to students with disabilities. We saw what the adults did not: Bullying and harassment could not be wished away with detentions and scoldings. The change required students standing up for one another, changing our school culture one relationship at a time.

For more than 30 years I have supported over a million young people working to change the world. From a water filtration system that uses garbage and a malaria test that uses enzymes, to a training program to reduce violent interactions between youth and police, young people name problems and organize around them. They build solutions long before adults give them much credit. We talk about young people as “the future.” Yet we rarely treat them as innovators worth backing now. Especially those without privilege and connections.

Venture funding offers one example of how concentrated our search for innovation can become. In 2025, California, New York, and Massachusetts together accounted for nearly three-quarters of the dollars invested, according to the National Venture Capital Association.

These established hubs produce extraordinary talent, but our reliance on familiar networks and credentials can become a shortcut for deciding which ideas deserve attention. Access is not innovation.

WE BACK PEOPLE WHO ARE EASY TO FIND

Innovators with the right connections get noticed sooner. Early attention leads to early funding and broader networks. Over time, the same signals keep getting rewarded because they already feel familiar.

Joshua Ichor, a 25-year-old Nigerian hydrologist and founder of Geotech Water, shows what that system can miss.

After being hospitalized from drinking contaminated water in his community in Nigeria, Joshua studied hydrology and built Geotech Water, technology designed to help communities monitor water quality, detect contamination, and keep water infrastructure working.

His advantage was proximity to the problem.

Joshua understood firsthand what unreliable access to safe water could mean for a community. It gave him knowledge an outsider would have needed time to learn, one reason RIVET is providing him with funding support.

Pitch fluency shows that someone knows how to communicate an idea, while not proving they have identified the strongest problem or built the most robust solution.

The reverse also holds. Strong ideas still need testing, execution, and evidence. The point is to stop confusing familiarity with quality.

WE UNDERVALUE PROXIMITY AS A FORM OF EXPERTISE

Walter Okwir saw how limited access to menstrual products affected girls in his Ugandan community. Girls were leaving school, marrying early, and having children younger, creating wider risks for their education and future.

Walter understood those connections because he lived close to the problem. His own sister had dropped out of school when she began menstruating. She was married with four children before the age of 20.

After receiving a small amount of catalytic funding from RIVET, he bought four sewing machines and began producing reusable sanitary pads, allowing more girls to stay in school, delay marriage, and reduce teen pregnancy. What intrigued me most about Walter’s approach is his decision to teach every boy in his community to make pads for their friends and sisters, which fundamentally shifts who is responsible for menstrual health.

Early funding gave Walter room to act on an idea shaped by his understanding of the community: an understanding of what people needed, what they would use, and where earlier approaches fell short.

That knowledge deserves more weight when we judge potential. Degrees matter. Technical skill matters. Execution matters. So does lived experience.

Promising ideas still need access to funding, expertise, and opportunities to demonstrate their potential. For many young innovators, that starts with a small grant and a mentor, as well as a chance to test an idea.

3 WAYS TO PLACE BETTER BETS

Working with young innovators like Joshua and Walter has taught me that recognizing potential requires looking beyond the people and places we already know. Here are three ways leaders can put that lesson into practice.

1. Look beyond the usual places.

Expand your search to include schools, youth groups, community organizations, and grassroots networks alongside established universities and startup programs.

The goal is to widen the field of promising innovators before applying rigorous evaluation criteria.

2. Give proximity more weight.

Ask how well the potential innovator understands the problem, whether the people affected trust them, what they have already tested, and what they learned from trying. Then evaluate whether they can execute, whether the solution works, and whether there is evidence the solution can grow or deepen its impact.

Proximity should be part of the rigor.

3. Make smaller, earlier bets.

Not every promising innovator needs a large check. A modest grant, a pilot, a mentor, or one strong introduction can give someone enough room to prove what works.

Companies already make early bets on new technologies. They should get just as comfortable making early bets on problem-solvers who did not come through the usual doors.

The next time you complete a performance review, plan a project, or seek input, ask how you can bring the insights of young people with lived experience to the table.

BUILD A BETTER INNOVATION MAP

For more than three decades, I have watched young people take on hard problems with urgency and imagination. I know they can build meaningful solutions and want to ensure others know how to recognize those solutions, when coming from outside the places and networks we already trust.

While innovation is widespread, opportunity is not equal. Successful leaders can change this by expanding their field of view. There are billions of young people ready to change the world. Are we willing to bet on them?

Eric D. Dawson, EdD is CEO and founder of RIVET.

The next competitive advantage is infrastructure

1 month 1 week ago

Businesses and policymakers are racing to build an economy centered on AI, and it’s quickly becoming one of the biggest economic bets in history. Billions of dollars are flowing into numerous sectors to support it, and states are competing to attract investment while companies seem to announce major expansions every day.

The U.S. is trying to build the AI economy as quickly as possible, but these ambitions depend on something less visible: infrastructure.

In 2025, the American Society of Civil Engineers gave U.S. infrastructure a grade of C, while anticipating a $3.7 trillion infrastructure investment gap between capital flowing for improvements and what is required for infrastructure to be in solid working shape. What we have now is an infrastructure that includes the power grid, water systems, and broadband networks that data centers and AI-driven businesses rely heavily on. These are aging, but also mismatched to the economy nearly everyone is intent on building.

A CHANGE IN INFRASTRUCTURE NEEDS

Much of the U.S. power grid, along with water and connectivity systems, was built in the 1950s and 1960s. The planning models behind them assumed a steady population increase and a consistent growth in manufacturing sectors. While infrastructure was able to keep pace with the demand for decades, that is no longer the case. Today’s economy is powered by industries that are far more infrastructure-intensive than those they are replacing. Aging systems now carry more weight than they were designed for.

Artificial intelligence is one of the most visible drivers of this increased demand, with its reliance on hyperscale data centers. These draw enormous amounts of power and water, but that’s not the only sector that does. Semiconductor manufacturing, battery production, advanced manufacturing, and electrification are all scaling up at similar paces, making similar demands as a result. Chipmakers such as Intel, TSMC, and Micron, for example, consume and therefore depend on reliable power, high-quality and available water, wastewater treatment, transportation, fiber connectivity, and permitting systems capable of supporting sustained growth. But that raises questions around cities’ and states’ capacity to support those demands.

For decades, companies selected manufacturing sites based on known factors such as labor costs, taxes, transportation, and proximity to customers. While those considerations still matter, companies are also now asking a more basic question: Can the infrastructure support us, not just today, but for the next 20 years? In states like Virginia, Texas, and Arizona, access to reliable power, water, transmission capacity, and efficient permitting can determine not only where projects are built, but how quickly they move from announcement to operation. For the states competing for this investment, infrastructure is no longer a background condition, but rather one of the key advantages.

THE PRICE OF ADMISSION

Infrastructure is now a primary business consideration directly influencing investment decisions. Communities that modernize their infrastructure will be better positioned to attract the next generation of AI facilities, semiconductor plants, advanced manufacturers, and other high-growth industries. Those failing to keep pace risk losing projects and jobs, along with long-term economic opportunity. Modern infrastructure is becoming the price of admission.

Water illustrates this problem. Decades ago, communities planned their water systems around population, housing, and agriculture. They incorporated reservoirs, wells, pipelines, and eventually treatment plants into these plans, to find and deliver more supply. Today, just one AI campus, one semiconductor fabrication plant, or one battery facility can dramatically change local water demand. Water needs for manufacturing are about quantity as well as quality. As such, that requires planning for reuse, advanced treatment, resilience, and making better use of the water communities already have. Virginia’s recent statewide groundwater study examining future data center development concluded that future development cannot assume unlimited water.

Rather than questioning whether data centers should be built, the report asked if existing water infrastructure can sustainably support them. It found that future large facilities using evaporative cooling would struggle to secure sufficient groundwater under current conditions. This illustrates how infrastructure capacity is increasingly shaping economic development. But AI did not create these infrastructure problems. Data centers are exposing where previous infrastructure assumptions no longer match reality. The way we plan infrastructure must change.

FINAL THOUGHTS

The technologies shaping tomorrow’s economy depend on infrastructure—systems that deliver power, water, transportation, connectivity, and the capacity to scale alongside innovation. Modernizing those systems is an economic strategy. The communities and businesses recognizing that shift will be best positioned to compete in the decades ahead. The future needs infrastructure designed for the economy we’re creating.

Kevin Gast is cofounder, CEO, and chairman of VVater.

The real threat to U.S. hardtech leadership is China’s rules

1 month 1 week ago

American hardtech venture is finally being taken seriously. Deep tech funds are outperforming conventional VC. The opportunity is real, and it’s being funded the way it should be, by investors who bear the cost of being wrong.

But there’s an elephant in this room, and it isn’t a coastal fund getting into the game without the capital, infrastructure, or institutional knowledge to truly back physical technology. It’s a structurally different kind of competitor entering the exact same categories—quantum, physical AI, hardtech, humanoids—that doesn’t play by the rules that make venture capital work in the first place.

GOVERNMENT-FUNDED INVESTMENTS

Last year, Chinese data provider Zerone reported that 90% of committed capital in China’s private equity market came from state-affiliated investors, up from about 79% in 2021. The shift reflects a clear priority from Beijing. China’s president, Xi Jinping, has repeatedly called on financial capital to “invest early, invest small, invest for the long term, and invest in hard technology,” a directive now reflected in China’s national investment strategy.

This is not “government money in venture capital,” which is neither new nor inherently distortive. The United States has run public venture capital for decades with programs like SBIC and In-Q-Tel, which has spent 25 years proving a government-linked investor can operate as a market facilitator rather than a market maker, co-investing alongside private capital instead of replacing it.

By comparison, OECD data shows government-affiliated investors participate in no more than 3% of all VC deals in the United States, and 11% across Europe. China’s 90% isn’t a bigger version of the same thing. It’s a different thing.

CREATIVE DESTRUCTION

Here’s why that distinction matters more than the raw dollar figures.

Venture capital works because it is disciplined by loss. Roughly two-thirds of all early-stage VC investments lose money. The industry survives that failure rate because the winners return enough to cover it and because losing is expensive enough that capital only goes to ideas that can plausibly clear a real bar. Professional VC funds, for all the money sloshing through the system, still invest in only about 0.2% of new U.S. businesses. That selectivity is the entire mechanism.

The term “creative destruction” coined by economist Joseph Schumpeter is an essential fact of capitalism and functioning markets. It is not a side effect to minimize, but the process that works. The failures are how the system finds out what’s real.

What happens when a fund doesn’t have to answer to that discipline? Or when the same institution supplying the capital can also extend the runway indefinitely, or become the customer through preferential procurement, or reprice the next round itself?

We’ve already run this experiment once. In Japan through the 1990s and 2000s, banks kept insolvent firms alive rather than recognize their losses. By 2002, roughly 30% of firms were on life support, holding 15% of all assets. That congestion suppressed entry of the more productive firms that should have replaced them and led to decades of stagnation.

A partner at Ivy Capital told Reuters in June that the current climate around funding for Beijing’s “future industries” push was a “level of frenzy…I have never seen in my entire career.” The same reporting described a company founded just three months earlier pitching investors on a valuation more than 30 times its current level on the strength of government backing rather than a demonstrated product.

The categories where this is happening fastest are the ones that matter most. China’s robotics sector raised more capital from January through mid-May 2026 than in all of 2025 ($5.6 billion versus $4.3 billion). Quantum computing funding in the first three months of 2026 exceeded the full 2025 year’s total. These aren’t peripheral bets, but the frontier technologies that will define the next decade of innovation. They are being funded at a pace and with a risk tolerance no market-disciplined investor would rationally match.

This is already showing up as policy. The U.S. recently banned imports of foreign-made humanoid and quadruped robots, citing documented cybersecurity exploits and supply-chain risk. The European Union signaled it intends to extend the same security-and-data logic that it applied to Chinese EVs to autonomous vehicles. They are recognizing that market share built through capital that never had to answer to loss is a different kind of dominance than one earned by surviving the two-thirds failure rate that discipline demands.

So what do we do with that? Not match it. You cannot out-subsidize a state. Trying only import bans and tariffs creates similar distortion.

THE CASE FOR HARDTECH

The U.S. answer should be the quality of what survives: technologies vetted by real customers, tested against real manufacturing constraints, and disciplined by investors who bear the cost of being wrong. This is why the rush of new money into American hardtech matters, but also why it isn’t enough.

Hardtech companies need somewhere to build. They need engineers, equipment, and specialized infrastructure to iterate. They also need customers willing to test something that hasn’t existed before. They need manufacturers that can turn a prototype into a repeatable product. And they need capital structured around the longer, messier path from invention to commercial scale.

The physical AI moment is a test of whether the system that rewards technology with capital can keep doing that faster than the system that doesn’t have to.

Haven Allen is CEO and cofounder of mHUB and managing partner of mHUB Ventures.

What is attribution science? The critical climate research that Big Oil would rather you didn’t believe in

1 month 1 week ago

Climate science has been clear for decades: Since at least the 1950s, oil and gas companies have known that their fossil fuels release planet-warming carbon dioxide emissions into the atmosphere. 

And we’ve known that those emissions cause dangerous planetary impacts like the melting of our ice caps, rising sea levels, and “potentially serious environmental damage worldwide,” as one 1968 study put it.

In the years since, experts learned even more about how climate change fuels extreme weather like storms, droughts, and heat waves. 

Actually nailing down the direct link between one company’s emissions and the amount of warming the Earth has experienced, though, or how much climate change amped up the intensity of a particular storm, has been a more complicated calculus. 

The effort to answer those questions is called attribution science. Though a more recent field than the extensive research on the harms of fossil fuels, it’s a burgeoning area of study that has seen significant scientific advancements over the past two decades. 

And it’s an increasingly important field for those looking to hold fossil fuel companies accountable for climate damages.

This kind of research is essential to climate litigation, and even plays a part in the major climate case Suncor v. Boulder, for which the U.S. Supreme Court heard arguments this week.

That Big Oil is fighting it, experts say, is another example of their history of climate denial.

The origins of attribution science

While research on the climate impact of burning fossil fuels began as early as the 1950s, the idea of “attribution,” or establishing a cause-and-effect connection, didn’t come until decades later. 

The Intergovernmental Panel of Climate Change’s (IPCC) third report, published in 2001, noted that “Detection and attribution studies consistently find evidence for an anthropogenic signal in the climate record of the last 35 to 50 years.”

In other words: We could say, scientifically, that human activity played a role in changing the climate.

But the field was still nascent. In 2003, though, a commentary published in Nature asked a crucial question: “Will it ever be possible to sue anyone for damaging the climate?” 

Written by University of Oxford physicist and climate scientist Myles Allen, the piece kicked off the effort to connect climate change to specific events—in his case, a flood that brought the waters of the River Thames “about 30 centimetres from my kitchen door”—as well as who could be responsible for paying for those damages.

The scientific models to make those connections didn’t even exist yet.

“We are not yet in the position to produce such figures for the contribution of greenhouse-gas emissions to the increased risk of flooding in south Oxford,” he wrote. “But the point is, if we get the science right, we could be.”

By the next year, he had found a way.

In 2004, Allen and two other climate scientists published a study in Nature that estimated the role that human influence played in increasing the risk of a deadly European heat wave, widely considered the first extreme weather attribution study. 

Different types of attribution science

That heat wave study is an example of a specific subset of attribution science called event attribution. It’s one of multiple subsets of the field.

Event attribution focuses on figuring out climate change’s role in specific instances of extreme weather, like that 2003 heat wave, or, say, Hurricane Katrina.

When Katrina occurred in 2005, attribution science wasn’t as advanced, but it has expanded significantly in the years since. 

Two decades later, using more advanced attribution analysis, Climate Central looked back at Katrina to analyze climate change’s “fingerprint” on the event. It found that “climate change made [Katrina’s] ocean temperatures up to 18 times more likely and, along with tropical climate warming, increased Katrina’s maximum sustained wind speed by 5 mph.”

And since the Earth has warmed even more since then, the outlet added that today’s climate could have made for an even stronger storm. 

Separate from event attribution is trend attribution, which focuses more broadly on identifying the human impact on long-term climate trends, rather than specific events.

That subset looks at human-caused climate change’s role—or the “human fingerprint”—on overall global warming, ocean warming, sea levels, and so on. This field was the focus of the 1990s research, like what was mentioned in that IPCC report.

Another subset emerged in 2014, called source attribution, when scientist Richard Heede, who founded and leads the Climate Accountability Institute, released his landmark Carbon Majors report. The aim of that work, Heede says, “was to see if we can quantify emissions attributable to one company over its entire history.”

Heede figured out exactly how many fossil fuel emissions major oil, coal, gas, and cement companies produced over their lifetime, then assigned those companies a share of total greenhouse gas pollution.

Of all the global emissions produced between 1751 and 2010, for example, Chevron was responsible for 3.52%, that report found.

There’s also a more recent subset of the field called impact attribution—or damage attribution—which looks not just at the intensity of specific extreme weather, but the social and economic consequences of climate change’s extreme impacts. That research has linked human-caused climate change to more deaths from extreme heat, for example.

Source attribution and climate litigation

Source attribution science plays a major role in climate litigation, or the lawsuits that seek to hold fossil fuel companies accountable for their role in climate change and the damage it has caused.

To Heede, this kind of science was a necessary departure from the typical ways of looking at climate responsibility. 

“It’s not based on the conventional model of attributing responsibility to nations,” he says. Much of modern climate policy focuses on what countries should do to reduce emissions or which nations should pay for climate mitigation. 

Instead, Heede’s work shifted that focus to the corporations that, he says, “have known about the hazards of their products, and the carbon in their products, to global atmospheric instability.” 

It’s not only companies to blame for climate change, he adds; governments are also responsible, and individual consumers even play a role. But companies have a “substantial responsibility because they have known the science for decades,” he says. 

“They then had a choice of going public with it, engaging with the public and governments to further study climate change to ascertain what could be done to reduce emissions,” he says, “and instead, most oil and gas companies invested in climate obfuscation and denial in order to delay action.” 

The Suncor v. Boulder Supreme Court Case 

Now, communities and municipalities are trying to hold those specific companies accountable by using the science that links their emissions to climate harms. But fossil fuel companies are continuing their trend of denial.

There are some two dozen lawsuits pending in the United States, mostly brought by states or cities as well as tribes, looking to get monetary damages from fossil fuel companies for their role in climate change, according to Columbia Law School. 

This week, the Supreme Court heard arguments in one of those cases, called Suncor Energy Inc. v. County Commissioners of Boulder County.

That case was originally filed in 2018, when the city and county of Boulder, Colorado, sued Exxon and Suncor “for the substantial role they played and continue to play in causing, contributing to and exacerbating climate change.”

“Plaintiffs and their taxpayers cannot pay the full costs of all that is needed, nor should they,” the lawsuit read. “The costs should be shared by the Suncor and Exxon Defendants because they knowingly and substantially contributed to the climate crisis.”

The oil companies argued that the case should be heard in federal court rather than state court. Now, the Supreme Court will decide whether that state case can continue—and the decision could affect the future of climate litigation. 

Those companies have also attacked the attribution science at the core of that case, arguing in a brief that greenhouse gas emissions “cannot be unmixed and traced to their sources in particular States or countries.”

In response, the Natural Resources Defense Council filed a supporting brief, Bloomberg Law reported, saying that these companies “overlook an extensive body of established climate science that draws these causal lines.”

Fossil fuel companies have also long opposed taking responsibility for Scope 3 emissions, or the emissions that come from the downstream use of their products, like the pollution from a car that runs on gasoline. Scope 3 emissions can account for 80-to-95% of a Big Oil company’s carbon footprint.

“They feel the consumer is solely responsible for the use of their products,” Heede says. “My opinion is different, that companies have a substantial burden of responsibility for the products they sell to consumers.”

If the Supreme Court halts the state-level case against Suncor, it could “kill many, if not all,” of the currently pending climate lawsuits that look to hold the fossil fuel industry accountable for climate change, Jeff Goodell, author of The Heat Will Kill You First: Life and Death on a Scorched Planet, wrote in the New York Times. 

That seems to be what Big Oil and its enablers are hoping for.

In a brief supporting the industry filed by lawyers for the Trump administration, they noted that if the case can proceed under state law, then “every locality in the country could sue essentially anyone in the world for contributing to global climate change.”

Even some justices are concerned about that potential.

“Presumably if you prevail, the next day, a municipality in every single state will file a lawsuit that will probably copy your pleadings,” said Chief Justice John G. Roberts Jr., per the New York Times. “How would you think that will work out on the ground?”

These arguments are “part of a coordinated, well-resourced push to shut down fossil fuel accountability efforts across the country,” Kathy Mulvey, Fossil Fuel Accountability program director at the Union of Concerned Scientists, said in a statement. 

But even if that happens, it won’t stop this quest for accountability, or the work being done on attribution science, experts say.

“Science is not going to be stopped,” Heede says. “We will have increasingly more sophisticated studies looking at different kinds of climate events . . . and that will just expand as climate impacts grow in severity, and attribute them not only to fossil fuels overall, but to specific companies.”

Attribution science will only continue to improve, both in terms of accuracy and comprehensiveness, he adds.

And even if Boulder loses in this case, Heede says, he doesn’t think that will “shut down municipalities’ incentive to file suit against companies that have acted irresponsibly in terms of not cleaning up the mess they have created.”

How AI is simplifying race day and speeding development at NASCAR

1 month 1 week ago

NASCAR fans today can do more to follow a live race than simply watch it on TV. Official NASCAR apps make live data, including telemetry feeds from race cars and behind-the-scenes driver audio communications, available to premium subscribers as races take place.

Getting all of that to work during races historically required a lot of careful human labor. If someone failed to toggle the right setting, for instance, driver audio feeds could stay mute, disappointing fans. And across NASCAR’s three-day race weekends, those tasks could number in the hundreds, ensuring fans get the latest data on everything from the shifting positions of cars to what services vehicles received during pit stops.

“There are hundreds of points of data that’s coming off of each of the cars that manifest onto this live leaderboard product,” says Tim Clark, executive vice president and chief brand officer at NASCAR.

Artificial intelligence, deployed through NASCAR’s longstanding partnership with the sports tech company Next League, has made getting that data to fans dramatically simpler. AI tools, and code written with their help, have automated much of the work needed to keep those apps filled with the data fans want. Roughly 90% of such tasks have now been automated, according to NASCAR and Next League.

[Photo: NASCAR]

“Over time, those things have been built in such a way where manual control of those things, depending on the state of the event at any given time, was imperative,” says Next League’s cofounder and CEO, David Nugent. “And now, it’s just not.”

AI-powered automation has helped elsewhere, too, as Next League expands its work with NASCAR to support digital projects for the racing league’s teams and the NASCAR-owned International Motor Sports Association. In addition to accelerating engineering efforts across NASCAR web properties and apps, AI has improved the delivery of targeted advertising to fans. So far, the newly personalized and contextualized ads have seen a click-through rate of about 2%, 10 times the rate of traditional NASCAR house ads.

Future improvements could include upgrades to NASCAR’s fantasy products or features, now in beta, that highlight elements of audio broadcasts and associated transcripts. One example: surfacing a recorded pit-stop conversation after a car puts on a related burst of speed, Clark says.

[Photo: NASCAR]

Still, Clark is adamant that the point isn’t simply replacing human labor. People remain in the loop to ensure race weekends go off without a hitch, but with less time spent pressing buttons and flipping switches, they can focus on higher-level tasks and deliver more data to fans.

“I think this year that that time has been freed up, it’s allowed our team to look more strategically at what we could be doing better,” Clark says.

One case where NASCAR used AI for a task normally done by humans is unlikely to be repeated, Clark says. That was the generation of a Spanish-language audio feed from the English radio broadcast for the league’s inaugural Cup Series event in Mexico. Though the audio quality was fine, and some fans were likely unaware it was machine-generated, NASCAR would probably hire a Spanish-speaking broadcaster for future events, Clark says.

[Photo: NASCAR]

But as other companies have discovered, one area where AI can be especially helpful is in generating quick working prototypes of potential products and features. That process lets NASCAR test new concepts rather than making decisions based on abstract ideas or static illustrations. This could prove particularly useful as NASCAR develops new ways to showcase its wealth of data, especially as AI frees up time previously spent ensuring existing feeds flowed properly to fans’ phones.

“Just from a technology and a development standpoint, to deliver on all of these things was a full-time job,” says Clark. “I think if what we’re saying is now that’s maybe a part-time job, the other part of it has to be on taking all of this information and all of this nuance around a sport like ours, and figuring out the most user-friendly way to deliver it.”

Big Tech is cutewashing its AI agents

1 month 1 week ago

First came the buttholes. Then the sparkles and gradients. Now, the next Big Tech AI branding trend has arrived. As the biggest artificial intelligence players introduce new agent capabilities, they’re all converging around the same visual shorthand: cute little guys.

On August 11, SpaceXAI (Elon Musk’s AI-focused mashup company) launched GrokBot, a series of AI agents designed to work on tasks autonomously, even when the user’s computer is closed. They’re represented by shapes with giant, orb-like eyes. On September 8, Meta came out with its own agent, called Muse, which looks like a fluffy monster toy (it has also been integrated into a Tamagotchi-style keychain). Then on September 29, OpenAI joined the party with a suite of AI agents called dots, which look like customizable Pixar characters and were described by the company as “remarkably capable” and “remarkably cute.”

We are firmly in the era of the anthropomorphic AI mascot, and according to multiple researchers and designers I spoke with, the trend makes perfect sense. In moments when tech companies want to encourage mass market adoption of an unfamiliar service, they’ve historically turned to friendly, approachable design.

Meta’s Muse mascot “Jolly” [Photo: Minh Connors/Bloomberg/Getty Images]

In a recent interview with Fast Company, Anthony Cappetta, partner and creative director at the creative studio Super Okay, explained that these emerging trends help companies “signal compliance” with established digital norms. “There’s a shared advantage in looking familiar and ‘safe’ within the ecosystem, especially for massive platforms where aesthetic shifts are scrutinized heavily,” he said. “It’s less about being original and more about being instantly legible and trustworthy at scale.”

But this visual shorthand comes with its own drawbacks. As AI makes its biggest consumer play in a moment of unprecedented doomerism around the technology, cutewashing a new and largely untested technology can spark its own backlash.

[Image: SpaceXAI] Investigating the Big Tech aesthetic ‘gravity well’

Kevin Walker, chief creative officer at the creative company Buck, has had a front row seat to watch tech companies falling into what he calls the aesthetic “gravity well”—or the phenomenon that causes Big Tech giants to coalesce around the same visual symbols. 

A Facebook illustration from 2019. [Image: Facebook]

In the mid-2010s, Buck was tapped to help a pre-Meta Facebook clean up its surface and brand expression. Back then, Walker says, there was no “cohesive idea of what Facebook is or the way that Facebook was going to talk to us.” The brand wanted a system that would highlight its ability to bring together people from all around the world, without trying to show literal individuals. The Buck team landed on an illustration style that it called Alegria, Spanish for “joy,” which showed figurative people with slightly wonky proportions that were easy to replicate. 

[Screenshots: Adobe, Google, TikTok]

At the time, Walker recalls, it was a “pretty unique brand expression.” Soon, though, everyone from Airbnb to Hinge, Google, and YouTube had embraced similar branding, earning the style the nickname of “corporate Memphis” for its ubiquitously slick and colorful appearance.

“You had a bit of this herd mentality—this brand expression got copied and cloned across tons of different companies, and it all kind of raced to this mean,” Walker says. “Where I like to think that the Facebook expression was actually pretty special and had craft and a POV behind it, it got pulled into this magnet of what we started calling ‘tech illustration.’”

Like skeuomorphism in the early days of the internet and “blanding” at the beginning of the app era, Alegria emerged at a time when users’ relationship to technology was shifting. As Facebook’s usership rocketed into the billions, the brand needed a way to manage the emerging, and increasingly contentious, narrative around social media and its impact on people’s lives. Alegria was one solution to the issue.

Soon, though, the illustration style’s ubiquity in the tech space turned it into a tired, meme-able trope. “Corporate Memphis” became a visual shorthand for everything wrong with technology, with netizens using it as a symbol for Big Tech’s attempt to appear “authentic” while amassing more control over the public’s lives, attention, and personal data. 

“When you have this gravity well, it makes it feel like everything becomes the lowest common denominator, which was an unfortunate long tail of that system of Alegria,” Walker says. In 2026, AI mascots appear poised to face a similar fate. 

[Image: OpenAI] The Clippy of the AI era

From their giant eyeballs to their stubby limbs and fluffy fur, there’s no doubt that AI agents like Meta’s Muse, OpenAI’s dots, and Grok’s Bots are designed to garner your affection. Joshua Paul Dale is a professor at Tokyo’s Chuo University, self-described pioneer of a new field of “cute studies,” and author of the book Irresistible: How Cuteness Wired our Brains and Conquered the World, and he says there’s solid psychology behind these designs. 

“Companies are using cuteness because it’s an evolutionary mechanism that prompts us to be social,” Dale says. “Seeing something cute stimulates the pleasure centers in your brain before you have a chance to think about it. Cute things are high in approach motivation, meaning we want to keep them close by.” Plus, he adds, research indicates that we tend to forgive cute things if they make mistakes. “This is a delicate issue for AI companies because generative AI is known for hallucinating. They want you to forgive mistakes but not to expect mistakes, because then you’ll choose another company’s AI.”

[Images: Wiki Commons, Chris Hondros/Newsmakers/Getty Images (pets.com)]

This isn’t the first time that Big Tech has turned to cute mascots en masse in a moment of technological unease. In the late 1990s and early 2000s, they were everywhere: think Ask.com’s cartoon butler Jeeves, Pets.com’s sock puppet, and Microsoft’s Clippy, for example. 

As Fast Company wrote in a 2014 memorial on the then-bygone tech mascot, “back in the ‘90s, when the Internet and computers really started to take over American homes, tech companies needed those playful mascots to show buyers that this new technology wasn’t so intimidating.” At the time, these characters acted as “a Virgil who guides through the murky depths of software and user guides.”

Clippy in present-day emoji incarnation [Image: Microsoft Design]

Cute mascots fell out of vogue once users had accepted the internet as a daily part of their lives. In the AI era, though, the cuteness pendulum is swinging back. Suddenly, every company seems to want their own cartoon Virgil.

There’s an obvious issue with the cute-ification of AI agents, though, and it’s the main lesson that Buck’s Walker took away from the Alegria “tech illustration” gravity well: If everyone has an anthropomorphic AI mascot, no one has a memorable anthropomorphic AI mascot. 

“If you’re starting to see that same snowball, is there going to be a situation where brands are going to get pulled into this gravity well and come out of it looking all the same, like derivatives?” Walker says. “To me, there’s great learning in that, which I say often when I’m dealing with clients now: You have to be unique. Your first question is, does this stand on its own? When people see this, do they understand who the brand is? And if they don’t, if they’re confused, you failed right there.”

The antidote to cutewashing

In 2024, Buck asked similar questions when it worked with Notion to design the brand’s animated AI assistant. The agency started with what devoted users loved about the platform—its sleek, easy-to-use interface and classy, hand-drawn brand aesthetic—and converted it into an expressive, classy character composed of just a few strokes and dots. It purposefully leaned away from common AI branding tropes, like the magic sparkle, in favor of a look that reflected the actual platform. 

“Know your brand first,” Walker advises. “Notion already had established itself as so illustrative that going in and having this really reductive 2D avatar mascot made perfect sense.”

[Image: Notion]

Other designers are opting for a similar approach. Arjun Mahesh is the head designer at Hebbia, an AI platform for financial institutions, and he recently designed an AI assistant called Max, which is the embodiment of Hebbia’s proprietary harness (essentially, the Hebbia-designed software infrastructure that clients can apply to frontier AI models). Mahesh says he designed Max—which looks like a swirling, modular robot—to “feel alive without feeling like a pet or person.”

“We want it to stick because you trust it and it works, not because it bats its eyes at you and tries to Tamagotchi-guilt you into feeding it,” Mahesh says.

[Image: Hebbia]

He adds that cute-ified AI mascots, like Grok’s Bots or Open AI’s dots, are clearly geared toward “broad, generalist use cases,” with friendly, easy-to-replicate designs that will allow these frontier companies to easily represent entire suites of AI agents as—presumably—user adoption increases. 

Given Hebbia’s more niche clientele, Mahesh wanted Max to feel representative of the brand’s established aesthetic, which pulls from eclectic inspirations including ‘80s and ‘90s tech culture, luxury watches, and architecture. His references for the design included Star Wars’ R2D2, Interstellar’s LARS, watch culture, and, oddly enough, string theory.

“It’s hard to make something that’s both opinionated and generalist at the same time,” Mahesh says. “If you’re trying to design everything for everyone, it’s not surprising to me that a lot of places are landing at a similar conclusion.”

Walker believes this moment of aesthetic convergence should be raising serious alarm bells at AI companies for potential oversaturation and user fatigue, à la Alegria. “Little warning lights need to be flashing: Where are we going to be in 12 months? What is this going to feel like?” he says.

At least for now, though, the tech mascot is officially back, and it wants to weasel its way into your heart—and your AI-powered workflow.

“Cute or not, AI agents can invade our privacy and surveil our every move,” says Dale, the Chuo University professor. “But in a world that seems more and more out of control, even the fleeting chance to say ‘aww’ can feel like a win.”

Checked
6 minutes 37 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