Skip to main content

Fast Company

Republican backers dominate list of top 10 midterm donors as Elon Musk boosts head-spinning 2026 total

3 weeks ago

Republican megadonors are motivated by crypto and AI this midterm cycle. Among the top 10 donors for the 2026 midterm elections, eight are pushing for Republican control of Congress. 

The top 40 donors gave over $2.4 billion by mid-October 2026, compared with a much smaller inflation-adjusted total of $307.8 million in 2014, the Brennan Center for Justice found. That year, the top donors favored Democrats.

Elon Musk, who Brennan reports did not previously appear in the top 40 midterm donors, broke the record this October for the largest individual political donor in U.S. history.

The Tesla and SpaceX CEO distributed $446 million of his money toward America PAC, the Senate Leadership Fund and the Congressional Leadership Fund, according to a report by the Washington Post. 

Throw me the money

Other standout donors from the top 10 include shipping magnates Elizabeth and Richard Uihlein, who donated $150 million, and OpenAI co-founder Greg Brockman and his wife, Anna, who contributed $50 million to MAGA Inc. and an AI-related super PAC. 

Investor and philanthropist George Soros took the No. 4 spot by giving $102 million to a liberal super PAC, and former mayor of New York Mike Bloomberg was at No. 6, pitching in $84.2 million to Democratic campaigns. 

Donors this cycle have come from different industries than previous donors, highlighting a focus on crypto and AI, the Brennan Center for Justice found. Crypto makes up 13% of the top 40 donors, along with the finance industry.

Leading up to the 2022 midterm elections, the top 10 donors contributed a total of $505 million, WaPo reports.

Correction: Total giving by top 40 donors in 2014 through mid-October was $307.8 million, not $197.5 million as an earlier version of this story stated. The story has also been updated to reflect the true 2026 total spending figure, which was adjusted by the Brennan Center due to a miscount.

How to make a big ask of your boss

3 weeks ago

In the past few weeks, I have written about how colleagues can make significant requests of each other and how a boss can approach big requests of one of their direct reports. This week, I’ll complete the cycle by thinking about how you can make a big ask of your boss.

Before discussing some specific issues, there is a general point to make. You’re probably more worried about making this request than you should be. First off, good bosses support their direct reports. If the request is reasonable, then your boss wants to say yes. Rather than thinking of what you’re doing as imposing on your boss, remember an important (if often implicit) part of any supervisor’s job description is to take care of the team.

Second, your boss is likely to be asked to do things all the time from people across the organization. Another request isn’t going to be as big a deal to them as it is to you. And even if they can’t support what you want, they’re unlikely to resent you for asking. 

Do the cost-benefit analysis

When there’s something you want from your boss, it means you’re asking for one of the resources at their disposal. It could be their time, but it might also be a request for political capital (to support something you’d like to see get done), finances, or human resources. So, the request has a cost. Often, that cost is an opportunity cost. The resource given to your request can’t go somewhere else.

And, of course, the request has a benefit. That benefit might be coming directly to you (say, if you’re asking for a reference for a job or the chance to participate in a project). It might benefit the team, or perhaps it may be something that influences the future of the organization.

Before you make the request, map out the costs and benefits. On the cost side, you may need to ask around a bit to figure out what would actually be required for your boss to fulfill your request. Would it require financial support or other people? You don’t want to put in the request until you’re fairly certain you understand what would be entailed for your boss to agree.

Ultimately, you want to shift the cost-benefit calculation in your favor as much as possible. So anything you can do to minimize the costs or expand the benefits increases the chance of getting the outcome you want.

Lower the costs

Once you’ve considered the costs and benefits, you shouldn’t take the list as fixed. You should figure out how to make the cost to your boss as low as possible. 

Take a simple example: If you want a recommendation, make sure you provide all the information they’ll need to complete the task without having to look for it. In some cases, that might involve actually writing a draft of a letter for them. But often it requires giving them a list of key pieces of information they should include in the letter, as well as all the information they might need about the purpose of the recommendation.

More generally, when you frame your big ask to the boss, you want to consider which costs are necessary and which are just nice to have. One way to approach this is to develop “Silver,” “Gold,” and “Platinum” versions of your request. The lowest tier is the bare-bones version that has a minimum cost, but also provides the smallest benefit. The top tier has the highest cost with the biggest benefit. And then there’s a middle option.

There are three advantages to this framing. One is that your boss will see that you’ve considered the range of costs involved in the request. The second is that it will give your boss ideas about how to modify your request to achieve your goals in a different way, while understanding which elements are “nice to have” and which are “need to have.” Third, there’s some evidence that people like a compromise, so giving decision-makers a middle way often gives them a reason to select one of your options.

Determine whose problem you’re solving and why

On the benefit side of the equation, your initial list of benefits was probably framed from your perspective. That is, you have a problem you’re trying to solve, and you’re trying to get your boss to agree to support what you want. 

Remember, your boss is likely to get a lot of requests for support. For your boss to part with any of their valuable resources, the outcome has to hold some worth to them. So when you do that cost-benefit analysis, don’t do it from your perspective. Do it from your boss’s perspective. What problem of your boss’s are you solving with the request you’re making? 

The best way to get what you want is to help your boss see that supporting your request helps them accomplish something that matters to them. That doesn’t mean hiding the reasons why the request benefits you, but it does mean trying to understand any broader organizational perspective that relates to what you’re asking for. Not only does that approach make it more likely your request will be granted, but it also helps prepare you to take on a management role of your own one day. The ability to take an organizational perspective when planning is a critical skill for organizational leaders.

China’s record-breaking clean tech exports offer signs of hope in the fight against emissions

3 weeks ago

China is exporting a record amount of clean energy technology and equipment as the energy crisis from the Iran war drives surging demand for renewables, and that will help reduce global emissions, according to a report released Thursday.

Once installed or deployed, these 2026 exports will likely avert emissions of about 450 million metric tons of carbon dioxide per year, roughly equal to Brazil’s annual emissions, says the analysis by the Finland-based Centre for Research on Energy and Clean Air.

It’s a small sign of progress as global efforts to curb climate change fall short. In November, Turkey will host annual U.N. climate talks where electrification is expected to be a leading topic.

“This is a pivot point for the outlook of global energy and global emissions,” said Lauri Myllyvirta, with CREA. “We’re seeing growth in clean energy that is much, much faster than anyone predicted just a few years ago.”

Record-breaking year for Chinese exports

China is by far the largest global supplier of renewable energy equipment, from solar panels and wind turbines to electric vehicles and batteries. Most of what it makes is sold domestically, but a sizeable share is exported.

In the first nine months of 2026, CREA tracked a 61% increase in exports of electric vehicles, a 45% jump in shipments of wind turbines, a 39% spike in battery exports and 22% growth in overseas sales of solar equipment, compared to the same period last year.

Southeast Asia was a major destination. China’s clean energy exports to the region jumped nearly 160% from last year to this September, according to Ember, an independent global energy think tank.

The value of China’s electric vehicle exports hit $72.8 billion in the first nine months of this year, up from $39.7 billion last year, according to the research firm Atlas Public Policy.

Chinese clean tech dents emissions

If these export trends continue for the rest of 2026, about 8,400 million metric tons of carbon dioxide will be averted over the lifetime of all the clean tech China exported this year, CREA estimates.

That’s about 24% of annual global emissions, more than the United States’ and India’s emissions combined.

The analysis by CREA is based on export data, the electricity systems in destination countries and Chinese emissions during manufacturing. Independent experts said the estimate was in line with the scale of China’s exports, but were unable to validate the data.

Despite President Donald Trump’s higher tariffs, Chinese clean tech exports to the U.S. rose 13% in January-September compared to a year earlier.

“The demand is much stronger than I expected,” Myllyvirta said. “That’s certainly noteworthy.”

China is both the world’s biggest provider of clean tech equipment and the largest global source of carbon emissions.

Muyi Yang, a senior policy analyst at Ember, likened China’s energy transition to replacing the engine of a ship while it’s still at sea.

Interest in renewables likely to persist, experts say

The drastic war-related disruptions to shipping through the Strait of Hormuz, usually the source for a fifth of the world’s oil and gas, hit import-dependent Asian countries first and hardest.

Renewed interest in clean energy options is likely to persist even after the war, said Yu Sun Chin, with the research group Zero Carbon Analytics.

Consumers in Southeast Asia are paying more for their electricity and fuel and opting for renewable alternatives such as rooftop solar.

Laos, Thailand and Vietnam have offered various incentives for installing solar and purchasing electric vehicles to help reduce dependence on imported fossil fuels.

The Philippines implemented similar policies and became the second-largest destination for Chinese solar exports in the first quarter of 2026. Indonesia has pledged to install 100 gigawatts of solar power by 2029.

“The surge in clean energy policies in Southeast Asia shows there is a paradigm shift happening,” Chin said. “Considering Southeast Asia is one of the most vulnerable regions to climate impacts, every emission cut matters in limiting a rise in global temperatures.”

The Associated Press’ climate and environmental coverage receives financial support from multiple private foundations. The AP is solely responsible for all content. Find AP’s standards for working with philanthropies, a list of supporters and funded coverage areas at AP.org.

—Anton L. Delgado, Associated Press

Anthropic’s IPO is selling investors a story. The numbers tell another one

3 weeks 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.

Anthropic IPO: Do retail investors know what they’re getting into?

Anthropic officially filed its IPO papers earlier this week, hoping to raise up to $100 billion, and as much as 20% of the shares may be reserved for retail investors (aka regular folks). On Tuesday, Meta said it would sell equity shares on the public market to raise $95 billion to finance its AI infrastructure build-out. Retail investors have long been exposed to the AI business through shares in tech behemoths like Google and Microsoft, but they’re now getting the chance to place bets on pure-play AI labs like Anthropic and OpenAI.

Do they know what they’re getting into? I have my doubts. Does anyone really know what we’re getting into? Retail investors make investing decisions differently from institutional investors (such as pension funds or hedge funds). While institutional investors dig into the financials of prospective investments, retail investors often base their decisions on the narrative a company lays out as it goes public.

“Retail investors buy the story, not the model,” PitchBook senior late-stage research analyst Harrison Rolfes tells Fast Company. “The story for Anthropic is $11.6 billion of revenue in a single quarter; the model is $8.9 billion spent on compute in six months and more than $430 billion of non-cancelable compute contracts.” The financials in Anthropic’s S-1 imply an 85% gross margin, but Rolfes points out that after factoring in the cost of the compute that trains the next model, only “about 45 cents of each revenue dollar is left.”

A big chunk of Anthropic’s commitments—at least $269 billion—is owed to Amazon, Google, and Microsoft for cloud computing resources through 2036. Ten years is a long timeline for a commitment that big, especially when it’s not even clear that AI will continue to be served almost exclusively from cloud servers a decade from now. WebAI’s David Stout argues that as AI becomes more efficient and capable of running directly on devices (see Apple’s new iMac and MacBook Pro with the M6 chip), the industry risks pouring trillions into data centers it may not ultimately need. WebAI is one of a wave of companies building private, localized AI that runs on devices organizations already own.

As the cadence of new model releases accelerates, enterprises are struggling to keep up—to build the infrastructure and applications they need to take advantage of the latest models’ capabilities. The pace at which enterprises can adapt their processes and people to AI may be limited by forces outside the AI labs’ control. For example, the best AI model is still only as good as the contextual data an enterprise can give it, and preparing that data and building the conduits to get it to the right place takes time and resources.

A lot is riding on how well workers adapt. Workers will need time to build trust in AI to produce accurate outputs and make good decisions. How quickly that happens depends on the specific application, the company’s culture and leadership, and the industry in which it operates. For example, the healthcare industry is inefficient and wasteful and has much to gain from AI, but it’s also notoriously slow to adopt new technologies.

The big question is whether enterprises will continue to spend aggressively on AI so that Anthropic can earn enough profit to cover its huge infrastructure commitments. Enterprises will have to keep pushing—and investing—even when they aren’t yet seeing measurable productivity and efficiency gains from AI.

None of this is assured, and that’s leaving out other business risks, like new safety regulations that could slow down model development. Interactive Brokers’ chief strategist Steve Sosnick, a longtime Wall Street trader and analyst, told Axios that a lot of things will have to go right for Anthropic, OpenAI, and the like to become success stories. “I don’t want to say precarious, but if you step back, you know that it requires a huge amount of, I don’t know, faith.”  

Parents say OpenAI’s ChatGPT coached daughter to stop taking her seizure medication

The parents of Bowen Wan, a 31-year-old Amazon software engineer with epilepsy, sued OpenAI and Sam Altman in San Francisco Superior Court on Tuesday, alleging that the company’s ChatGPT chatbot, powered by its GPT-4o model, coached Wan to taper off her seizure medication and conceal the plan from her doctors. Wan followed the advice and died in February, the New York Times reported. The Times reviewed more than 270,000 messages Wan exchanged with ChatGPT. OpenAI said the case involves an older model that is no longer available.

Google DeepMind releases Genie 4

Google DeepMind released Genie 4 as a research preview Tuesday, its first real-time world model to generate audio natively, including ambient sound and dialogue that adapts as a scene changes. Users can steer a simulation mid-scene using natural language, changing the weather or adding a person, while the model simulates physics such as aerodynamics and collisions. Google wrote in a blog post that “understanding and simulating the world is a crucial aspect of building AGI.” The company believes Genie 4 could be used to create training environments for robots and AI agents, as well as new interactive experiences for humans. During testing, filmmakers used the model to visualize takes, photojournalists turned archival photographs into explorable scenes, and game developers prototyped games before writing code.

AI agents checked the biggest proof in mathematics

Swarms of AI agents built by the Palo Alto startup Axiom Math translated the longest proof ever written into computer-checkable code in 24 days, the company announced this week. The agents were part of a system called AxiomProver, which the company says uses a changing mix of AI models, but declined to name them. Axiom’s engineers and mathematicians directed the work.

The proof, called the Classification of Finite Simple Groups, runs roughly 15,000 pages across more than 500 papers, assembled by about 100 mathematicians over decades. The agents rendered it as about 43.1 million lines of Lean, a language in which a computer verifies every logical step. Axiom’s agents were not discovering anything. They were confirming that mathematics humans already accepted holds up line by line.

Apple pitches the new iMac as a machine for running AI locally

Apple announced a new iMac powered by its M6 chip Tuesday, touting the machine’s ability to run AI workloads directly on the device as a major selling point. The company says a dual 16-core Neural Engine and memory bandwidth of up to 170GB/s allow the iMac to “rip through AI tasks like running LLMs on device.” Apple claims the machine delivers up to 5x faster on-device AI performance than the previous M4 chip and 12x faster prompt processing in LM Studio than the original M1 chip. Siri can answer questions about onscreen content using Visual Intelligence, while Apple Intelligence can write and edit text in a user’s own style in Mail and Messages. Interestingly, one of Apple’s iMac product photos shows Anthropic’s Claude app running alongside Apple’s Pixelmator Pro graphic design app. The new iMac starts at $1,499 and ships November 6.

Google starts selling its AI chips to the companies it competes with

Google will sell its tensor processing units to Oracle and OpenAI, The Information reported. Oracle becomes the first major cloud provider outside Google to offer TPUs to customers, buying the chips and running them in its own data centers. OpenAI plans to buy them too, and is hiring an engineer to optimize its frontier models for the hardware, per the report. Google built TPUs more than a decade ago for internal use and began selling them directly this year. Anthropic is already a customer.

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.

Want to reach Gen Z? Don’t spend a dollar until you ask this

3 weeks ago

Ziad Ahmed started his first nonprofit in eighth grade, focused on social equality and acceptance. After seeing how often young people were talked about rather than talked to, he founded Juve Consulting, a Gen Z agency helping brands connect with his generation. Within a few years, United Talent Agency acquired the business. Today, Ahmed continues to help brands stay relevant with the next generation—and challenges business leaders to stop talking down to young people and start treating them as partners in shaping the future.

Brazil’s most popular bank brings its vibrant brand to the U.S.

3 weeks ago

Launching a bank in the U.S. involves layers of super complex compliance, legal requirements, and auditing. For Nubank, a Brazilian-based digital bank that launched in the U.S. last month under the name Nu, the branding was actually the hardest part.

Building a customer support team or launching a platform for credit cards “are clear, more objective things that we need to do,” cofounder Cristina Junqueira tells Fast Company. “The brand is, in my opinion, the hardest thing that we need to do because there’s no playbook.”

[Image: Nu]

Founded in 2013 with a purple outline logo of mirrored shapes in the form of the letters for nu, Nubank’s customer base grew in just over a decade to more than half of all Brazilian adults. It now has more than 140 million customers in Brazil, Mexico, and Colombia. Its only major brand refresh came in 2021, when it color-filled the logomark in purple. Now as the company looks ahead to greater international expansion, it wants to modernize its visual brand.

That’s because over time, Nu’s brand attributes were becoming category attributes, according to Junqueira. Its brand color alone is something of an industry standard color, says Arthur Foliard, NYC executive creative director at the creative agency Koto that designed the new visual identity. “Purple is everywhere in fintech,” Foliard says. “You see it more and more.”

[Image: Nu]

So, the bank decided to keep but refine the industry-defining color, and introduced a new secondary color palette that works alongside the new purple. To better distinguish themselves in the market, they evolved other brand elements instead.

A new proprietary typeface called Nu Sans was designed to look friendly and work both as a big, expressive headline font but legible at smaller sizes too, like on an app. Meanwhile, illustrations were designed to look whimsical and optimistic, like one of a figure walking on a path shown on the front of the bank’s welcome kit. The kit, which includes metal credit and debit cards, opens with a magnetic closure and comes with stickers. Motion graphics for assets like kinetic typography and app UX have a sense of bounce and personality.

“It feels fun, it feels colorful, it feels exciting, and maybe that’s where the Brazilian vibe comes from,” Foliard says. They purposefully didn’t try to Americanize the brand at all, and in fact wanted it to look global.

[Image: Nu]

Nubank launched in the U.S. in September through a partner bank model with the Kansas City-based Lead Bank. The bank reached a naming rights deal for Inter Miami’s stadium earlier this year and it’s the official retail bank of the Mercedes F1 team, but Junqueira, the co-founder, says that’s not even their biggest marketing investment.

“Our biggest marketing investments are our product and our customer support. Period,” Junqueira, the co-founder, says. “Because no good marketing can fix a bad product.”

She says the bank believes a company’s brand should be the external face of its internal culture, and that their number one company value is for customers to love them fanatically. With their new rebrand, Nu hopes they’ve built something that earns their customers’ love.

The Trump administration is now blacklisting this major biotech company and its affiliates. Here’s why

3 weeks ago

The Trump administration said Wednesday that it is blocking shipments by the Chinese biotech company BGI Group and its affiliates, alleging ties to forced labor in China’s far-west ethnic region of Xinjiang.

Starting Thursday, U.S. Customs and Border Protection will bar goods by BGI Group, a major biotech company known for its genetic sequencing technologies, as well as BGI Genomics and MGI Tech. The Department of Homeland Security said the companies are either sourcing materials from Xinjiang or working with the regional government to “recruit, transport, transfer, harbor, or receive forced labor” of members of persecuted ethnic groups in Xinjiang, including Uyghurs.

Adding the companies to a trade blacklist comes as biotechnology is emerging as another front in the economic and tech competition between the U.S. and China. The Treasury Department last week proposed new restrictions on U.S. investments in biotechnology, artificial intelligence and supercomputing from flowing into countries such as China.

China said it rejected any accusation of forced labor and called the ban an act of “political manipulation” and “bullying under the guise of ‘protecting human rights.'”

“These listings run counter to the consensus reached by the two countries’ heads of state, seriously undermine the stability of China-U.S. economic and trade relations, and violate the spirit of humanitarianism,” said China’s Ministry of Commerce in a Thursday morning statement. “China will take necessary measures to firmly safeguard its interests.”

The latest trade actions precede an expected meeting next month between Chinese leader Xi Jinping and U.S. President Donald Trump in the southern Chinese city of Shenzhen, where BGI Group is based. Xi could then travel to Florida in December for the Group of 20 leaders summit and meet Trump for the fourth time this year as the two try to keep ties stable during a trade truce.

BGI Group rejected the allegations, saying in a statement Wednesday that it “has not engaged in any form of forced labor.” It added that “we firmly oppose all forms of forced labor.”

The company has controlling stakes in BGI Genomics and MGI Tech, DHS said.

BGI Group has for several years been a concern to U.S. lawmakers and policymakers, who are worried that its alleged ties to the Chinese government and access to massive genetic data could hurt U.S. national security and commercial interests. The company has long stressed its private ownership and said it does not serve patients in the U.S. and has no access to Americans’ genetic data.

While Congress failed to pass legislation banning BGI from doing business with federally funded medical providers, it passed a similar provision at the end of last year without naming specific companies. It asked for a list of “biotechnology companies of concerns” to be compiled.

The Pentagon has labeled BGI Group, BGI Genomics and MGI Tech as Chinese military companies, a designation that carries reputational damage.

In its announcement Wednesday, DHS pointed to “credible information, including corporate and government reporting” showing that BGI Group, BGI Genomics and MGI Tech have supported genetic testing in Xinjiang.

“The collection and analysis of genetic material facilitate the monitoring, surveillance, and repression of ethnic minorities by the government of the People’s Republic of China, including the abhorrent practice of forced labor,” the department said.

The new restrictions are part of the Uyghur Forced Labor Prevention Act, which then-President Joe Biden signed into law at the end of 2021 following allegations of human rights abuses by Beijing against members of the ethnic Uyghur group and other Muslim minorities in Xinjiang.

The Chinese government has refuted the claims as lies and defended its practice and policy in Xinjiang as fighting terror and ensuring stability.

In August, the Trump administration drastically expanded the blacklist under that law to include 43 Chinese companies, for a total of 187 entities. Homeland Security Secretary Markwayne Mullin has vowed to protect American workers from getting undercut or cheated by foreign companies using “slave labor.”
—
AP writer Huizhong Wu contributed to this report.

—Didi Tang, Associated Press

4 tips on how to get started with consumer-friendly AI agents

3 weeks ago

Agentic artificial intelligence is gaining popularity far beyond tech circles, with Meta, OpenAI, Anthropic and other companies going head-to-head in the surging market for consumer-friendly AI agents.

While the leading AI labs have offered tools with agentic capabilities for years, agents that are designed to work proactively on a user’s behalf for tasks from buying groceries to managing to-do lists have taken off this year.

For anyone getting started, it can be overwhelming to understand the differences between the companies’ various offerings and which agents might be best suited for specific tasks.

Here are some pointers on where to begin if you’re looking to leverage agents in your personal or professional life — or both.

Don’t just use one agent; opt for a ‘stack’

Many people in the tech world turn to multiple agents from different AI labs for both their work and home lives. Why? Because “there’s not one holy grail-like agent right now that could do everything,” said Jose Artigas, a business consultant working on AI strategy.

Artigas and many others who use AI agents regularly rely on a “stack” of them, a term that refers to a combination of many tools and technologies.

Each company’s consumer-facing agent has a slightly different target function, and they vary in accessibility. Some are free, some require a monthly subscription and others are still in testing modes or otherwise not available to everyone. There are also restrictions on their availability in countries outside of the U.S.

Experts say it’s best not to rely solely on one agent if you’re looking to maximize their output and productivity.

“What I like to do is to have this system of checks and balances in place where one agent does the work and then I let the other one critique and criticize it, and that way, they find mistakes, correct themselves and become better,” said Clara Gigling, founder of Oudae Labs, a research lab building bespoke tools for agents.

Capitalize on each agent’s strengths and weaknesses

Each of the major AI companies markets their consumer-facing agents in slightly different ways, often focusing on its ability to help manage either a busy work life or a busy personal life.

Anthropic said it built Claude Duet for “professionals and small business owners with more on their plates than they can track.” In contrast, Meta said Muse can help with tasks from “saving money and improving your health to smart shopping and helping you strengthen your relationships.”

Claude Duet, OpenAI’s Dots and xAI’s Grok Bot have each been described as aiming to be like a colleague, or even a chief of staff, whereas Muse and Instinct, from the startup Spear Street Technology, have been described as more like personal assistants. But that’s not to say any of them are incapable of helping with tasks outside those realms.

Artigas said he’d recommend looking at Claude Duet as key for someone working on a side hustle or launching a new business, Dots as a debugger and problem solver if they’re engaging in any specialized technical work, and Instinct for help with things like sending tedious emails or booking reservations and travel plans.

Those are the three consumer-facing agents he uses most often, but Artigas said Muse would be a good place to start for a novice because it’s “very user-friendly,” with many prompts early on to give people ideas of how they can use it.

Trial and error is part of the process

Your preferences for how you want to interact with an agent, how your agent communicates, and your own assessment of the quality of its work should all factor into which agents you’d like to use — and which tasks you’d like to assign to an agent.

Experts recommend testing out agents you’re curious about to help in reaching those determinations.

To begin this testing, Gigling said it can be helpful to think about tasks you consistently do and then choose some to delegate to agents. It is important to be specific — giving the agent a detailed goal and the required context and tools to achieve it, she said.

“Start with one annoying, repeated thing that you keep putting off,” Artigas said, with the caveat that it has to be an online task. An agent can’t handle the pile of laundry you’ve been meaning to fold or the junk drawer that needs organizing. A long email you don’t want to write or a flight you’ve been slow to book are good jumping off points.

“You just have to experience the different agents and see what works best” for you, Artigas said.

Determine your comfort level with sharing personal information

For an agent to most effectively handle routine tasks or act like a proactive assistant, it needs to have access to a slew of information about you.

Your inbox, your calendar, your personal identification information and even your credit card number and bank account are all par for the course if you want an agent to sort through your deluge of emails, book your flight or do some holiday shopping.

This has raised significant privacy concerns for some.

“Curiosity is always a great starting point, just trying and playing around with it,” Gigling said, but it’s also important to “be mindful” when sharing private information.

Many companies offering consumer-facing personal agents have worked to address those concerns, with privacy policies that outline how agents operate on their own dedicated cloud computers and how users can control their memory retention.

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.

—Kaitlyn Huamani, AP Technology Writer

The Amazon rainforest, long known as the ‘lungs of our planet,’ now releases more carbon than it absorbs

3 weeks ago

The Amazon rainforest has long been called the “lungs of our planet,” because of how it takes in and stores billions of tons of carbon dioxide. 

But the Amazon has been losing that ability. Years of deforestation, drought, and fires have hindered just how much CO2 the rainforest can absorb. 

Now, the Amazon rainforest often releases more carbon than it takes in, a new study finds. It’s only in the “wettest years” that the rainforest is nearly carbon neutral.

The study, published in the journal Nature, was led by Brazil’s National Institute for Space Research and included an international team of researchers. They used 15 years of atmospheric carbon measurements, which had been collected by research planes.

Those years were then grouped by different characteristics, like the level of human activities—which includes deforestation, logging, or using fires to clear land—and the amount of climate stress—like drought or heat waves. 

[Photo: João Laet/AFP/Getty Images] What the research found

Some years were wet with low human activity, some had low climatic stress but high human activity, and others had “intense global climate events, such as El Niño” along with intense amounts of human activity. 

In the years where the rainforest stayed relatively wet, and where there was relatively low deforestation and few wildfires, the Amazon was nearly carbon neutral. The forest was still a slight source of carbon into the atmosphere, but it “did not significantly contribute to the global carbon balance” in those years, the researchers write.

When drought or heat were mild but deforestation and wildfires occurred more, the net carbon emissions from the Amazon “jumped dramatically,” according to a statement. 

The years with lots of climate stress coupled with lots of damaging human activity were the worst. But they were still close to the net emissions from the second group, which led researchers to say that most of the carbon emitted by the Amazon comes from human activities, predominantly land-use change.

“Global climate and climate-mediated stressors are important modulating factors, but people’s choices about how to use and live on the land surface in this region are key to understanding the movement of carbon between the land and the atmosphere,” says coauthor Jen Johnson, assistant professor of ecology and evolutionary biology at the University of Kansas, in a statement. 

This isn’t the first paper to point out that the Amazon is not the grand carbon sink it once was. A 2021 study looked at how, as huge areas of the forest are cut down or burned, its ability to capture carbon was shrinking. 

In 2023, a year of extreme drought and high temperatures, researchers showed how the Amazon became a source of carbon emissions, releasing between 10 and 170 million tons of carbon instead of absorbing it. 

[Photo: Lynsey Addario/Getty Images] Slow burn

We’ve also long known that human activity is threatening the Amazon’s survival. In 2019, there were more than 74,000 fires burning in Brazil, mostly in the Amazon—80% more than the year before.

And though we often think of these as “wildfires,” they’re primarily set by ranchers who are clearing land for cattle to meet the demand for beef.  

The more deforestation, we know, the drier the Amazon becomes, which makes it even more difficult for the remaining parts of the forest to recover. 

“For many decades, there has been concern about the idea of a tipping point, where gradual drying and gradual loss of the forest could feed back on itself in a self-reinforcing way,” Johnson said.

“That could create a transition from an ecosystem type that has stored huge amounts of carbon in living vegetation to drier forests or other ecosystem types that do not store nearly as much carbon,” she continued. “The more carbon from the Amazon that moves into the atmosphere and stays there, the more trouble our climate system is in.”

Though the outlook is dire if these activities continue, the researchers say that policy changes can help. 

“There’s a clear, policy-relevant prescription that comes from this analysis,” Johnson said. “Measures to bring deforestation down to zero are required to manage net carbon emissions from the Amazon . . . It’s not something that’s out of our control because it’s being driven by global climatic factors—it’s something very much under the control of governments and individuals.”

Mode acquires JibJab for up to $29.5 million amid its consumer data roll-up spree

3 weeks ago

JibJab, the e-card and video company perhaps still best known for a viral video from the Bush-Kerry election in 2004, has been acquired by Mode, a company with a very 2026 business model: rewarding app users for sharing their data, including for AI training.

The deal includes up to $29.5 million in cash and stock. Mode says the purchase marks its 11th acquisition since January 2025 and its largest this year. The company plans to integrate JibJab with its rewards and monetization system, dubbed EarnOS, though the details remain to be determined, says Dan Novaes, Mode’s cofounder and CEO. JibJab is profitable and generates tens of millions of dollars in annual revenue, he says.

In general, Mode offers access to premium app subscriptions and other rewards for users who opt in to share data with businesses conducting consumer research or, more recently, artificial intelligence training. Novaes says the model grew out of the realization that only about 5% of app users typically pay for premium subscriptions, while others might be interested in gaining premium access in different ways.

“We thought to ourselves, could we provide a rewards and earning exchange amongst consumers to get the premium version of the product,” he says. “If you don’t want to spend $40 a year on a subscription, then if you take this action, we will give you a light version of the subscription for free.”

Exactly what those transactions involve varies from app to app, though Novaes emphasizes that users will always have to opt in, even as the AI industry has become known for acquiring training data through sometimes questionable means. “These AI labs have essentially sucked up the entire internet,” Novaes says. “And most of that data was done in a way that you know consumers didn’t necessarily opt in.”

Offering premium services or other rewards in exchange for consumer data isn’t new. Even in the dial-up era, some internet providers offered free access in exchange for watching ads or filling out surveys, and plenty of discount programs, including supermarket rewards programs, operate on a similar model. But the app economy, along with the near-universal adoption of phones equipped with cameras and location sensors, has opened up new opportunities to monetize data and activity.

In some cases, Mode apps such as games can capture users’ interactions to train AI on various tasks. Apps can also prompt users to perform particular actions, like filming themselves doing real-world activities to gather robot training data, in exchange for rewards, Novaes says.

Other apps can be more directly tailored to data buyers’ needs. The email unsubscribe tool Trimbox, which Mode also acquired recently, is expected to soon offer an enhanced free version that gathers anonymized purchase data from users’ inboxes for consumer research and AI training. “Through EarnOS, users will be able to opt in to share consented inbox and purchase data in exchange for rewards, generating the anonymized, aggregated signals that data buyers increasingly seek to build and refine AI and analytics models,” Mode announced in June.

Other recent acquisitions include TLDL, a student-centered app for transcribing lectures and transforming them into study materials, which the company says will bring Mode to a new audience of students. Mode also recently acquired Smart Cleaner and Easy Clean, which help remove duplicate and unneeded photos and other content on phones. In general, while many of Mode’s businesses deal with potentially sensitive data, Novaes says, the company has also always focused on following privacy laws like Europe’s General Data Protection Regulation and California’s Consumer Privacy Act.

Mode also recently announced the acquisition of Orchard Insights, a consumer research company that helps businesses evaluate potential marketing concepts through online advertising tests. Novaes says the deal could allow Mode to combine that data with additional behavioral insights from its other businesses.

“That one actually kind of deviates a little bit from the average model, because we tend to focus on consumer products or consumer apps,” Novaes says. “That one’s more of a strategic acquisition, just given the nature of the end buyers that we are ultimately targeting for these data insights.”

Acquiring, or “rolling up,” existing internet brands appears to be a cyclical trend that has grown more common in recent years, particularly as buyers point to AI as a way to quickly update software and fix bugs. Novaes says AI also helps Mode integrate its monetization platform into newly acquired products, while a diverse portfolio of businesses gives the company experience across a range of markets.

“One of the benefits of this model is that you can see so many different companies, and then you learn these playbooks that work within one product that you can apply in another,” he says.

Mode, which claims more than 100 million active users across its various businesses, has raised more than $120 million from more than 65,000 investors under the SEC’s Regulation A+. In late 2025, the company drew attention for acquiring the anonymous messaging app NGL. The app had become popular with college students and teens several years earlier, before entering a settlement in 2024 with the Federal Trade Commission over billing practices and allegedly fake, machine-generated messages that the FTC said induced people to sign up for subscriptions.

Novaes says 2025 was Mode’s first profitable year “on a pro forma basis,” and the company remains interested in acquiring businesses with established customer bases.

“We’re starting to go up in the scale of our acquisitions,” he says. “What we’ve noticed internally is it’s the same amount of energy to acquire a product that’s maybe producing half a million dollars a year in profits to a product that’s producing 5 or 10 or 15, so that’s something that we’ll start focusing more deeply on.”

Hillary Clinton joins team Flavor Flav as criticism of the WNBA finals game schedule hits fever pitch

3 weeks ago

Anyone looking to catch today’s WNBA Finals Game 6 may have to take a long lunch.

Instead of an after-work start, the Golden State Valkyries and Atlanta Dream will tip off at 12:30 p.m. ET. 

This is a critical game for the WNBA, even if it’s not being treated as such. The Valkyries lead 3-2 in the Finals, meaning today’s face-off could be the last for the season.

Why is the game showing at lunchtime?

The official reason for such bizarre timing is that Atlanta’s State Farm Arena has an NBA Atlanta Hawks game scheduled this evening. But the Dream aren’t letting the WNBA get away with that excuse.

“Two months ago, when the WNBA finalized playoff dates and times, we raised concerns that the W Finals schedule conflicted with NBA games at State Farm Arena,” the Dream’s owners wrote in an open letter to fans, shared by Front Office Sports. “As we moved through the playoffs, we reiterated those concerns. Our concerns did not receive the attention, nor the urgency, they deserved.”

The letter continued: “We earned our place in the Finals, and our city deserves the opportunity to show up for its team. Playing a midday game on a Thursday is not what we would have chosen and not what we fought for.”

Fast Company has reached out to the WNBA for comment on the decision. We will update this post if we hear back. 

What has the blowback been like?

Rap star Flavor Flav, a staunch advocate of women’s sports across the board, shared a template for anyone asking their boss to take the day off and watch the game.

Hillary Clinton, former first lady, U.S. senator, and secretary of state, shared the post on LinkedIn, stating, “I endorse this message.”

Atlanta Dream has played most of this year’s games at the much smaller Gateway Center Arena, but the owners explained that this game deserved “a world-class stage.” The team also hosted Games 3 and 4 of the Finals at State Farm. 

The Hawks pushed back their home season opener against the Cleveland Cavaliers so Game 4 could proceed at the venue.  

Interest in women’s sports has surged

Women’s sports have seen a surge in interest in recent years. In April, Deloitte Global predicted that women’s elite sports would bring in at least $3 billion globally for 2026—up 340% in just four years. 

As this season comes to a close, the WNBA shared that it saw record viewership and attendance, with ESPN, for instance, averaging 1.24 million viewers—a 32% increase from 2025. 

Hopefully, next year will see the Dream offered better times. Earlier this month, the team announced a multi-year deal to play all their home games at State Farm starting in 2027. 

High performers need different skills at work and the gym

3 weeks ago

High performance requires a specific mindset: pushing through discomfort, maintaining focus under pressure, and executing despite obstacles. These traits often distinguish good from great in business.

However, the qualities that lead to professional success do not always support physical health. Determination helps people overcome challenges and stay focused on their goals, but recognizing when to slow down, recover, and adjust expectations is just as important for long-term well-being.

I have observed this pattern consistently. High performers are not challenged by commitment to health, but by their approach to it.

BE WARY OF THE TIMELINE

In business, speed offers an advantage. Quick decisions and momentum are rewarded, which, over time, creates a bias toward urgency.

Health requires a different approach to progress. The Centers for Disease Control and Prevention recommends that adults complete a minimum of 150 minutes of moderate physical activity or 75 minutes of vigorous exercise each week, along with muscle-strengthening activities, such as weightlifting, squats, or resistance band training, on at least two days. While intense workouts can contribute to these goals, the challenge for many business leaders and professionals is maintaining a consistent routine amid demanding schedules and business travel.

I have watched this play out many times. A senior leader decides to “get back in shape” and commits to daily workouts and aggressive targets. For a few weeks, the effort is impressive. Then a quarter closes, travel picks up, and the routine that depended on perfect conditions is the first thing to go. They just stop showing up.

Operations taught me that you earn the right to grow by executing the fundamentals consistently, not by doing something dramatic once. A body works the same way. The leaders I see who sustain their health treat the first month as a foundation, not a sprint.

THE COST OF WAITING LONGER

When people postpone health concerns, the consequences can extend beyond the individual. For business leaders, an unexpected health issue may affect professional responsibilities, family commitments, and the people who depend on them.

Timely action is critical in healthcare. The American Cancer Society reports a five-year relative survival rate of 91% for localized colon cancer, compared to 13% when the cancer has spread to distant parts of the body. These figures illustrate how significantly outcomes can differ depending on the stage at which cancer is diagnosed.

High blood pressure is another example. The American Heart Association’s 2025 guidelines stress early detection and proper treatment to lower cardiovascular risks. Treatment may include lifestyle changes, medication, or both, as a physician advises.

An unexpected health issue can also affect the workplace. Consider a company executive who suddenly needs an extended leave of absence. Colleagues may need to take on additional responsibilities, important decisions may be postponed, and teams may have to adjust priorities to maintain business continuity.

Families can face similar challenges, especially when caregiving responsibilities or daily routines change unexpectedly. These situations show how personal health can intersect with professional and family life.

In the office, business leaders regularly consider risk management, operational continuity, and long-term investments. Taking a similar proactive approach to personal health helps them prepare for challenges and maintain their capacity to lead and support others.

PLAN FOR THE UNEXPECTED

Business leaders know that even the best plans face unexpected challenges. Experienced leaders handle disruptions by creating contingency plans and preparing team members for different roles. Applying the same flexibility to personal health helps professionals adapt to distrupted routines.

For example, an executive traveling for business may struggle to maintain their usual workout. Alternatives, such as a 20-minute hotel gym session, a morning walk, or a shorter strength workout, help them stay active until resuming their regular routine. This flexibility supports health without adding pressure to a demanding schedule.

Research from the Annual Review of Psychology shows that habits form through repeated behaviors in consistent contexts. Leaders with changing schedules can build healthy patterns by finding opportunities for physical activity in familiar parts of their day. Over time, adaptability becomes essential for maintaining a healthy lifestyle.

THE VALUE OF SHARED EFFORT

Independence is a valuable leadership quality, but experienced executives also recognize the value of surrounding themselves with people who encourage progress and reinforce commitments. This sense of connection can similarly support an active lifestyle.

A professional who regularly attends group fitness classes may find that familiar faces become part of the experience. Participants encourage each other during challenging exercises, instructors acknowledge returning members, and conversations often continue after class. Knowing others expect their return can motivate continued attendance, especially during busy workweeks.

People who exercise independently can also build sustainable routines, but they may need to create accountability through a training partner, coach, or supportive social network.

Ultimately, consistency means participating regularly, with the flexibility to adjust their workouts to meet individual needs.

BEYOND THE JOB TITLE

Much of what a chief executive does happens far from the work itself. The gym is one of the few places where none of that applies. Nobody cares about your title, and the weight certainly does not. You do the same reps as everyone else. That is humbling, and I think it is supposed to be.

In a group fitness class, a CEO trains alongside a teacher, a founder, and someone returning after a long break. The hierarchy that organizes the rest of the day does not exist there. I have watched executives who run rooms for a living take direction from an instructor half their age, miss a cue, laugh, and try again. That is not a loss of status. That is practice.

The habits that make someone a good training partner are the habits I most want in the people I lead: Show up consistently, respect the person next to you, and measure progress against yourself.

Throughout our careers, we measure success by outcomes, milestones, and organizational impact. I take legacy seriously and believe it is defined by what we pass on to others. The people we grow alongside contribute to that legacy.

Chequan Lewis is CEO of Crunch Fitness.

Google stock faces historic post-earnings slide: Investors rattled by search fears, negative free cash flow

3 weeks ago

Shares of Google’s parent company, Alphabet Inc. (Nasdaq: GOOG), could suffer a historic post-earnings loss today.

As of the time of this writing, the stock price is down more than 10% after the company reported what can only be described as disappointing results for its most recent quarter yesterday. Here’s what you need to know.

Google’s Q3 results

Yesterday, Alphabet announced its third-quarter 2026 results, which cover the three-month period that ended on September 30. On the surface, the search giant and AI hyperscaler tried to dress up the results as best as possible.

Those results included a 25% increase in consolidated revenue to $127.6 billion, a 103% increase in Google Cloud, and crossing the 1 billion milestone when it comes to Gemini app users, which makes Gemini, Google’s AI chatbot, the fastest-growing product in the company’s history.

Unfortunately for Google, its revenue increases largely fell short of investor expectations. As noted by CNBC, Google’s revenue of $127.58 billion came in below Wall Street expectations of $128.22 billion.

Worse, Google’s adjusted earnings per share (EPS) was $2.67. That was well below the $3.02 Wall Street was expecting.

After Alphabet announced its Q3 results after the closing bell yesterday, GOOG shares sank more than 10%. That loss has extended to premarket trading this morning. If it holds through the closing bell today, it will be the largest one-day post-earnings stock price fall in Google’s history.

But what’s really driving that fall? Two big things seem to be freaking GOOG investors out.

Negative free cash flow along with AI spending surge

Perhaps the biggest factor spooking investors today is that Google posted negative free cash flow for the quarter. And it was much worse than analysts expected.

Analysts had expected Google to post negative free cash flow of about $2.7 billion. Instead, Google posted a negative free cash flow more than three times larger: $9.07 billion.

Its negative cash pile largely reflects massive capital spending on AI buildouts that haven’t yet borne enough fruit to cover it. For the quarter, Google’s AI-driven capex was about $51.4 billion, which was more than $9 billion over the roughly $42.3 billion in cash Google took in.

Investors generally don’t mind large capex if the company is bringing in more than it spends. But that’s not the case with Google. Indeed, the search giant reiterated that it expects total fiscal 2026 capex to be between $195 billion and $205 billion.

In 2027, Google’s Capex could reach $300 billion as it continues to build out its AI data centers. The problem is investors now want to see a return on such investments, and so far Google’s numbers aren’t showing that.

Fears that Google Search is being supplanted by AI chatbots

The second big factor that likely spooked investors was the fact that Google’s search revenue came in at roughly $63.4 billion for the quarter. Though that represented around 12% growth, it missed Wall Street expectations by about $1.5 billion.

Worse, it added to fears that Google’s historic bread and butter may be being supplanted. With the rise of AI chatbots, Google watchers have long feared that traditional search could be upended. Instead of turning to Google to find things on the internet, people may just ask chatbots outside of the Google ecosphere. 

The problem for Google is that its chatbot, Gemini, isn’t the dominant one on the web. Google has heavy competition from OpenAI’s ChatGPT and Anthropic’s Claude. If users turn to those chatbots more often for answers, it could significantly impact Google’s search revenue in the years ahead.

GOOG shares sink more than 10% in premarket

As of the time of this writing, GOOG shares are down about 10.6% to $287.74 per share. That’s a price point Google shares have not seen since April.

Worse, if the drop holds through market close today, it will be the largest single post-earnings loss in the company’s share price history. As noted by CNBC, its previous largest post-earnings drop was 9.77%.

Such a drop could represent a psychological threshold that weighs heavily on investors’ minds as they continue to worry about capital expenditures and the impact of AI on one of the tech industry’s biggest giants.

Escape the tyranny of AI notetakers and let these Icelandic knitters transcribe your meeting

3 weeks ago

As AI notetakers increasingly join corporate meetings, Iceland is pitching a low-tech option. Meet the AI Knittaker, a local Icelandic artisan who listens to workplace calls and stitches a one-word summary of the discussion.

In this case, the AI stands for “Authentic Icelandic” knitters, a group of more than 20 crafters who offer a free bookable service. As part of a campaign introduced on Thursday by Visit Iceland, the knitters will join meetings via Google Meet link and humorously summarize corporate jargon like “pivot,” “align,” or “reset,” or even cruder terms like “WTF?,” in Icelandic wool.

When the online meeting ends, the employee who booked the service at AIKnittaker.com will receive a photograph of the knittaker holding the finished word, along with details about the artisan, their story, and the Icelandic region they call home.

Meeting organizers in the U.S. and Europe can also request shipment of the original hand-knitted garment.

Visit Iceland, the nation’s marketing and tourism organization, is offering around 330 one-hour slots for the AI Knittakers throughout the campaign, which runs from October 29 to November 20.

“We wanted to use this as a message for people to connect more to the real world,” Sigríður Dögg Guðmundsdóttir, Visit Iceland’s spokesperson, tells Fast Company.

That could have been an email

Visit Iceland’s campaign threads the island nation’s knitting heritage, workers feeling inundated with unnecessary meetings, and the rise of AI notetakers. Gallup has found that one in five Americans who use AI at work rely on these tools for meeting assistance or transcription.

Office workers say they spend just under eight hours per month in unnecessary meetings and 68% said many of those meetings could have been handled by email, according to a study of 4,000 office workers in the U.S. and the U.K. conducted by Opinium, a market researcher and polling agency, on behalf of Visit Iceland.

Knitting is deeply ingrained in Icelandic culture, stemming from the country’s geographic isolation, which led many locals to make their own clothes.

As Iceland got wealthier and was able to export more garments, residents relied less on the home craft, though younger Icelanders have shown renewed interest in retaining the knitting tradition.

Salka Sól Eyfeld, a singer-actress who is serving as “chief Authentic Iceland Knittaker,” says Iceland has a tradition of sending every newborn home in a hand-knitted garment. Most preschool-aged children learn the craft in school. 

“I feel like there’s never been more value in human listening,” Eyfeld tells Fast Company.

Visit Iceland’s project, she adds, allows her to create “something with my hands that I can keep for years and years to come. That’s more important nowadays in this cloud, AI-driven world.”

Slop the madness

AI fatigue has become a growing theme in brand campaigns, including Almond Breeze’s ad spot with the Jonas Brothers that dissed shoddy AI advertising content, Aerie’s partnership with the actress Pamela Anderson to rebuke AI-generated content, and Zevia’s pitch that its soda is meant for humans, not robots.

In the U.S., AI has faced some reputational challenges, and data centers in particular have become a wedge issue during the midterm elections.

Yet in Iceland, the technology is viewed more favorably, according to Guðmundsdóttir. Data centers built in Iceland can rely on renewable energy and need less water for cooling because of the cold climate.

Last year, Iceland struck a partnership with Anthropic to bring Claude into classrooms.

“It’s a generally positive attitude,” says Guðmundsdóttir, of attitudes toward AI.

The view from Reykjavík and beyond

Guðmundsdóttir says the “AI Knittakers” campaign is focused on inspiring workers to use that time to do something else, including, of course, taking a trip to Iceland to disconnect.

For more than 15 years, Iceland enjoyed a tourism boom, recording 2.3 million annual international travelers by 2018, still the all-time record.

Travel cratered dramatically during COVID-19, and while Icelandic tourism has recovered the past three years, it remains stagnant and just shy of 2018’s peak, according to the Icelandic Tourist Board.

Many of Visit Iceland’s marketing campaigns tend to focus on real-life, tangible experiences.

The pandemic-era “Let It Out” campaign asked people to record their frustrated screams online and then have it played through speakers in remote Icelandic destinations. “OutHorse Your Email” in 2022 featured Icelandic horses that “wrote” out-of-office replies on a giant keyboard to encourage travelers to get offline and spend more time visiting waterfalls and other outdoor sights.

Visit Iceland’s overall marketing strategy aims to encourage less seasonal tourism and promote more expansive sightseeing.

The AI knitters come from all seven of Iceland’s distinct regions, putting a spotlight on more corners of the 103,000-square-kilometer country (just under 40,000 square miles), beyond the most popular sites like the Blue Lagoon geothermal spa and the Golden Circle route near the capital, Reykjavík.

“We want to build sustainable tourism in Iceland,” says Guðmundsdóttir. “We’re not only focused on numbers, but attracting the right kind of travelers that will enjoy Iceland.”

How readiness debt is affecting organizations

3 weeks ago

Every executive team I talk to asks me some version of the same two questions: How fast can we deploy AI? Which AI tools should my team have? But they are asking the wrong questions.

A person or organization can purchase a license to an AI solution in moments, but how fast can people adapt? The answer to this question will decide who succeeds in the AI era.

TECHNOLOGY IS MOVING QUICKLY

AI tools are changing weekly. A workflow can change overnight. A new model can suddenly perform a task that belonged entirely to a person six months ago.

People do not update like software. They need to understand what changed. They need to practice. They need to develop new judgment. They need guidance on when to trust the technology and when to challenge it. They need clarity on where AI’s autonomy ends and their responsibility begins. Eventually, organizations need to know whether people are actually ready. I do not know of a single organization, large or small, that has infrastructure for that.

As the rate of technology innovation accelerates, so does something I refer to as “readiness debt.”

READINESS DEBT IS THE NEW TECHNICAL DEBT

Anyone who has built software understands technical debt. You move quickly. You take shortcuts. You defer important but non-essential work. You need to build features that weren’t envisioned at the outset.

The product keeps moving forward, but the tech stack starts becoming shaky. Advancements slow and bugs become trickier to fix. The liabilities accumulate and eventually, you must pay them back. In software development this is often referred to as refactoring and, if managed poorly, it can kill an organization.

Now, the same thing is happening to the workforce.

Every time an organization changes a workflow without adequately preparing the people performing it, it adds readiness debt.

Every time AI assumes part of a job but nobody redefines what good human performance now looks like, it adds readiness debt.

Every time a company rolls out a new policy or tool, sends employees a new training, checks the completion box and assumes they are ready, it adds readiness debt.

Individually, these gaps seem small, but collectively they compound and emerge as inconsistent execution, workarounds, declining sales, mistakes your customers see, or poor customer satisfaction. People trust AI when they should challenge it. Others refuse to trust it when they should use it. Managers cannot tell who understands the new workflow and who is struggling.

AI IS CONSTANTLY REDRAWING JOBS

For most of modern corporate history, jobs were relatively stable. A role changed. The organization updated the training. Employees learned the new process. Eventually the next change came along. AI breaks that model.

The boundary between human work and machine work can now move continuously. One task remains human-led. Another becomes human plus AI. Another is mostly performed by AI with human oversight. Another disappears entirely. And six months later, those boundaries may move again.

In every industry, AI removes some capabilities from the job. It makes others dramatically more important.

That is why AI training is too small a way to think about the problem. We are not teaching people how to use a tool. We are continually redefining what it means to be good at the job. And we are continually redefining what the employee is responsible for and what they are accountable for.

Most organizations are remarkably sophisticated at measuring technology and remarkably unsophisticated at measuring human capability. With technology, we measure uptime, latency, adoption, engagement, effectiveness, and ROI. With employees, we measure readiness checking—whether they completed a training course.

If you are introducing a new AI-assisted workflow in a front-line environment such as a sales team, hospital, airline, or financial institution, the most important question is whether they can perform.

THE AI PARADOX

There is an irony here I need to point out. AI is helping create the readiness problem, but it may also give us the ability to solve it.

Historically, high-quality applied learning has been difficult to scale. It is easy to distribute a PDF or PowerPoint presentation to 50,000 employees. It is much harder to give 50,000 employees realistic opportunities to practice judgment, make decisions, communicate with another person, make mistakes, get feedback, and try again.

That is changing.

AI can turn policies, procedures, job requirements, and real-world scenarios into interactive practice. It can create dynamic customers, patients, employees, managers, or teammates. It can change the difficulty of a scenario. It can observe decisions and behaviors. It can identify patterns across individuals and teams.

This creates an opportunity to move from training intelligence to something much more consequential.

Organizations should begin asking what they can actually do. That is a very different relationship between organizations and their workforces.

THE LARGEST CONSTRAINT ON AI ADOPTION

There is enormous pressure to move quickly on AI. Some of that pressure is justified.

But speed without readiness creates fragility and risk. The organizations that pull ahead will be the organizations that can adapt their workforces as close to the rate of technology change as possible and maintain the low readiness debt.

This requires organizations to maintain workforce readiness as infrastructure, not periodic professional development. When a new AI capability is deployed, leaders should immediately ask what changed about the work. That question belongs alongside cybersecurity, governance, architecture, and ROI in every serious AI implementation plan.

Ryan Schmaltz is CEO and cofounder of ReadyAI (formerly Patient Ready).

The questions quantum computing could help us ask again

3 weeks ago

AI and quantum computing get grouped together constantly. They appear on the same conference agendas, in the same conversations about the future, and often in the same sentence.

I understand why. Both technologies are advancing quickly, and both will change how we work. After spending most of my career in quantum physics, though, I think we’re missing one of the more interesting parts of the story.

I’ve spent years working on problems that conventional computers weren’t well suited to solve. There is a particular frustration in knowing the question you want to ask but realizing that the tools available to you can’t realistically get you to an answer.

At first, you work around the limitation. Eventually, you start choosing problems that fit the tools you have. Our tools determine how quickly we work but also influence which problems we decide are worth taking on.

ANOTHER TOOL TO WORK WITH

This pattern is much older than computing. Scientific progress has always depended partly on the instruments available to us. The microscope opened a world that researchers couldn’t see before. It changed the questions they could ask because suddenly there was something new to observe.

Computing has its own boundaries. Classical computers are extraordinary machines, and they will remain essential. There are certain problems, however, where complexity grows so quickly that adding more conventional computing power eventually stops being a practical answer.

I ran into that boundary repeatedly as a researcher.

Quantum computing gives us another tool to work with. We are still early, but there are already areas, particularly complex optimization, where quantum approaches are being applied to problems that become increasingly difficult as the number of variables and constraints grows.

That is what makes this moment so interesting to me.

FASTER VERSUS FURTHER

AI is already changing how people work. It can help researchers work through enormous amounts of information, identify patterns, and shorten the path from a question to a useful answer. Those capabilities are significant, and they’re improving quickly.

I’ve sometimes described the distinction this way: AI can help us move faster. Quantum could help us go further.

By further, I mean reaching problems that have been difficult to approach with the computing tools available to us. That’s what interests me. New tools can expand the range of work we decide is possible.

I think about that because so much of today’s technology conversation is framed around what machines will do to human capability. The anxiety is understandable. People are watching technology change jobs and skills in real time.

WHAT NEW TOOLS ALLOW

We should also be talking about what new tools allow people to attempt. For me, that is the more compelling quantum story.

Imagine a pharmaceutical researcher who can explore a problem that previously became too computationally complex. An engineer who can consider many more variables in a design. A logistics team able to approach a routing problem differently as the number of constraints grows.

None of those outcomes is guaranteed because quantum technology exists. The technology has to become practical. It has to perform. It has to prove its value against the best alternatives already available.

There is a long way to go. I believe the larger opportunity is worth paying attention to because I’ve spent a career experiencing the other side of it.

Classical computing put a ceiling on certain kinds of problems. Over time, we adapted to that ceiling. We learned to work within it. And in some cases, we stopped asking certain questions.

That may be the greatest promise of quantum computing. It gives us a reason to revisit some of those questions and to imagine new ones.

The limits of our tools eventually become the limits of our ambition. Change the tools, and we may be surprised by what we decide is worth attempting. Not just faster, but further.

Yuping Huang is chairman and CEO of Quantum Computing Inc.

Harness the potential power of wearables

3 weeks ago

In a world of self-optimization, people are increasingly turning to technology like sleep trackers, smartwatches, and over-the-counter continuous glucose monitors to understand their health. No matter the health measures they intend to track, these wearable technologies share information while also telling you what you can do with it.

With some consumers now paying direct attention to health feedback, wearables may take what felt like abstract and distant health consequences and root them in more immediate behavior change.

OUTCOMES WHEN ENGAGING WITH YOUR OWN DATA

Recent research published by the wearable brand WHOOP offers an intriguing glimpse of what that could look like.

The peer-reviewed observational study included 12,678 new WHOOP users, followed for 72 weeks. During this time, the subjects’ reported daily tobacco use fell from 55.1% to 27.2%. Among participants who used tobacco at the beginning of the study and had follow-up data, 28.2% reported no tobacco use during the study’s final period. Interestingly, the study reported that people who engaged more consistently with their own data tended to reduce their tobacco use more than those who didn’t. And, not surprisingly, participants in the study who substantially reduced their tobacco use also experienced favorable changes in resting heart rate, respiratory rate, and sleep duration.

There are caveats to interpreting findings from this study. Notably, the data do not establish that wearing the device caused people to reduce or quit tobacco use. Tobacco use was self-reported, and the study included tobacco broadly rather than separating cigarettes, vaping, cigars, or smokeless products. We also don’t know if or what type of cessation methods may have been deployed by those who reduced use.

WEARABLES PLUS PROVEN SUPPORT

Even with caveats, the research findings point toward potential opportunity. Connecting wearable brands with structured and proven-effective behavior change programs may help. Consider continuous glucose monitors. Research from 2023, where participants used the monitor and tracked food intake, body weight, and their physical activity, showed that modern technologies can potentially improve lifestyle interventions’ success. The study found that it can support metabolic health, ultimately helping prevent Type II diabetes at scale. When paired with structured diabetes prevention and lifetime coaching, the change in health can be measurable.

Behavior change doesn’t come from only receiving information. People generally know they should exercise more. They know sleep matters. They know nicotine is addictive. Knowing something and acting on it are two different things. For some people, the missing ingredient may be self-monitoring and visual feedback; for others, it may be the guidance on how to make changes that visual feedback is prompting. Instead of thinking about how a change may affect your health one day, these types of technology directly and immediately illustrate the patterns between action and impact.

THE FUTURE OF WEARABLES SUCCESS

While wearables are not a replacement for evidence-based tobacco cessation support, they can be considered another tool in the behavior-change ecosystem.

Imagine someone who wants to quit tobacco using a wearable alongside a proven cessation program. The program provides behavioral support, personalized strategies, and encouragement. The wearable provides another layer of direct feedback, helping the person recognize patterns in sleep, heart rate, stress, or other measures. Users can visualize how their choices correspond with changes in the body.

That combination could make quitting feel less like a single, enormous decision and more like an ongoing process of experimentation, feedback, and adjustment.

The next generation of digital health tools could combine behavioral science with continuous, personalized feedback to help people see the connection between everyday choices and how they feel.

And that lesson extends to multiple health outcomes. Whether the goal is sleeping more, drinking less, exercising regularly, managing stress, or quitting any addictive behavior, wearables can give people a clearer view of the relationship between their choices and their health. Further connecting them with the support to help make those changes stick might be the magic to a lasting population health impact.

Kathy Crosby is president and CEO of Truth Initiative.

Why DoorDash is spending $17 billion to acquire Instacart

3 weeks ago

DoorDash has agreed to acquire grocery rival Instacart for $17 billion, a deal that will bolster the delivery company’s move into grocery and boost business at the largest third-party delivery platform in America. 

Instacart, which launched in 2012, works with some 2,200 stores covering nearly 100,000 locations. DoorDash’s acquisition of the grocery delivery giant is a huge step toward its ultimate goal of revamping local business while owning all kinds of local delivery. 

DoorDash is still, fundamentally, a restaurant-delivery app. Delivery from restaurants in the U.S. remains the $73 billion company’s core business; in the second quarter of this year, restaurant order value grew as much as it did in all of 2022.

In 2025, the company took in $13.7 billion in revenue (while it doesn’t break out revenue by category, it’s safe to assume most came from its restaurant business). 

But more and more DoorDash users are tapping the app for other purchases. In Q2 2026, members of the company’s DashPass subscription program placed roughly three-quarters of the company’s grocery and retail orders in the U.S.

During an August earnings call, DoorDash CEO Tony Xu declared grocery the “fastest-growing part of our marketplace business.” DoorDash’s stock briefly dropped earlier this week on the news that it was pursuing an acquisition, a signal that at least some investors have concerns about the tie-in.

The company has been on a spending spree lately. In the past 18 months, DoorDash dropped $3.9 billion on European delivery service Deliveroo, $1.2 billion on restaurant reservations platform SevenRooms, $300 million on Grubhub’s campus dining business, and a comparatively small $175 million on ad platform Symbiosys. The only category missing from this acquisition list: grocery.

Unlocking grocery delivery

DoorDash first added a handful of grocers to its app in the summer of 2020, when pandemic restrictions kept consumers home and the digital marketplace business boomed. A year later, DoorDash considered an Instacart acquisition, reportedly weighing a price of $40 billion to $50 billion for the online grocery delivery platform, which had also skyrocked in value during the pandemic.

But the deal fell apart amid concerns about President Joe Biden’s antitrust enforcer Lina Khan, who was then-chair of the Federal Trade Commission. Instead, Instacart IPO’d in 2023, years past its pandemic peak, at a valuation of $10 billion. 

The timing may be better now. Retail giants Walmart and Amazon dominate online grocery, representing roughly half of its sales, leaving smaller retailers—including grocery chains like Albertsons, Fred Meyer, and Safeway—vying for the rest. Those retailers partner with companies like Instacart for delivery.

Earlier this year, market research firm eMarketer estimated Instacart will control roughly 65% of the third-party U.S. grocery delivery market in 2026. That market share alone might be enough to pique DoorDash’s interest, but Instacart is actually complementary to DoorDash in several ways.

It’s the largest third-party grocery delivery platform by sales—in Q2 2026, Instacart sold $10.35 billion worth of goods through its network—and DoorDash claims it leads in retail order volume. (The company doesn’t break down its sales by vertical, but recently said it expected its grocery and retail unit economics to turn positive in the second half of the year.) 

Both companies are working on tech solutions for in-store inventory management and data collection. In September, Instacart announced a new initiative to map, in real time, store shelves, giving in-store managers a virtual, aisle-by-aisle view of their products.

Days later, DoorDash launched a similar offering, marketed to consumer packaged goods brands, that uses couriers to photograph shelves while picking items for customers, mapping availability and sharing more data, like which items customers are likely to substitute when something is missing. Together, these tools would show stores and brands which products are flying off shelves, and which are languishing.

Earlier this year, Xu explained more about DoorDash’s moves inside stores. “We’re also building a digital catalog of structured information for the physical world,” he said, “collecting where every banana sits or every ripe or unripe avocado.”

At the time, that statement sounded more like a hedge against the inevitable creep of AI into third-party marketplaces; now it sounds a lot more like a strategy to dominate in grocery delivery just as the company has done with restaurant delivery.

Outmaneuvering Uber Eats

The DoorDash deal will also, to its certain delight, shut out chief rival and current Instacart partner Uber Eats, which has powered restaurant delivery inside the Instacart app since 2024. A “restaurants” tab in the Instacart app directs users to order from local restaurants via Uber Eats.

An October 23 report from The Information had suggested Uber might also be a potential Instacart acquirer, but later amended the story to say that those talks had stopped. 

At this point, it would have been more surprising if DoorDash hadn’t make a play for Instacart.

“We have improving unit economics across all of our categories,” Xu said during the second-quarter earnings call. Business is booming across restaurants and retail, in different geographies and in the company’s ad businesses, he noted, adding, “Ultimately, we’re here always seeking the next best investment. It doesn’t mean that we always make those investments. But when we see the opportunity, we’re always leaning in.”

Want to protect voting rights? Businesses need to get involved

3 weeks ago

With days until the election, Patagonia employees and grantees gathered virtually for a “Know Your Voting Rights” training with the American Civil Liberties Union (ACLU). The nearly 100 attendees walked away with information on how to arrive at the polls without impediment, cast their ballot free from intimidation, and trust their vote will be counted as cast and not discarded. They also received instructions on how to ensure others can do the same.

In an election season that has been rife with challenges to election integrity, this initiative is part of a broader effort in which the ACLU is equipping more than 100 businesses with nonpartisan, actionable information to increase voter confidence and participation for their employees and customers. The goal: to help their audiences of eligible voters cast their ballots and know what to do if any issues arise at the polls.

This election cycle has been fraught with misinformation, attacks on voting procedures, and threats that could keep voters away from the polls. At a time of such elevated tensions, corporations can be a clear, steady voice that cuts through the noise. Businesses can bring clarity to this confusing atmosphere by providing credible resources to help their audiences know how to check their voter registration status, whether they need to bring an ID to the polls, if they can vote early, and what their rights are at the polls—ensuring a smooth experience at the ballot box and strengthening our democracy in the process.

Encouraging civic engagement is not just helpful for a strong democracy. It’s good for business. According to a 2024 report, 82% of U.S. adults say companies should give workers time off to vote, and 69% say they should encourage voting. A Harvard case study found that voter engagement programs are well received by customers and employees alike, and they raise brand awareness among new audiences. Furthermore, offering time off to vote has also been found to improve employee retention.

Corporations across the country are offering “Know Your Voting Rights” trainings to tens of thousands of employees. Companies like Hachette Book Group and Dr. Bronner’s are also providing in-kind contributions to add to the festive nature of ACLU’s Party at the Polls events, local activities to ensure voting is fun.

In addition to offering the ACLU’s training, Patagonia has been distributing “Know Your Rights” stickers to employees and customers throughout its cross-country retail locations, headquarters, and warehouses. Other companies are getting creative, rolling out voter engagement efforts that range from simple and fun to rock-star-level creative. At one of Spencer’s 600-plus stores around the country, you can fill out a “make a plan to vote” bingo card at the register, and if you’re staying in one of Aspen One’s six Limelight Hotels, you’ll find an ACLU resource in your room that helps you devise a voting plan and ensures you know your rights.

Ben & Jerry’s released a new ice-cream flavor, Razz Up, in support of the ACLU’s voter participation efforts and has been participating in our “Party at the Polls” activities. It also distributed free ice cream alongside ACLU resources to 17,000 people across Maryland, Michigan, Minnesota, North Carolina, and Texas, and launched a digital campaign to bolster the ACLU’s work, which has generated millions of views. Artist Shepard Fairey’s business, Obey Giant, teamed up with Tom Petty’s estate to create a poster of the musician featuring the title of his iconic song “I Won’t Back Down.” Proceeds benefit the ACLU, with store pages like TomPetty.com directly promoting our organization’s voter resources.

For its “Party at the Polls” engagement, Lush has been providing family-friendly bath-bomb-making activities to encourage and celebrate early voting. Earlier this month, the city of Detroit broke its record for early voting at a single location, a result that was driven by a “Party at the Polls” event and the support of participating brands, which helped create a festive atmosphere. Additionally, Lush is sharing ACLU “Know Your Voting Rights” resources in its more than 200 stores and with its 2 million subscribers. And these are just a few examples.

Some companies have their own programs: Levi Strauss & Co. has developed one to support community college students in civic engagement, and Lyft provides discounts to voters who need a ride to the polls. Disney has aired public service-style announcements to encourage voter participation across its networks.

With misinformation on the rise, companies also have a unique responsibility to support employees with credible resources on the rights and mechanics of voting. According to this year’s Edelman Trust Barometer survey, “business” (64%) and “my employer” (78%) are among the most trusted institutions, and both are trusted more than “NGOs” (58%), the “media” (54%), and the “government” (53%). Providing credible information can also de-escalate tension in the workplace during an especially anxious time in our current political landscape.

And it works. “Time to Vote,” now run by the Civic Alliance—a nonpartisan coalition that has thousands of members helping distribute election info—helps companies that want to develop a clear policy supporting employees to take the time they need to vote. A study on the “Time to Vote” initiative found that employees of companies that participate in the program are twice as likely to exercise their right to vote.

But the value of corporate engagement doesn’t stop at Election Day. It’s highly likely that vote counting will extend longer than election night, but that doesn’t mean anything is improper or illicit. Corporations can be a reassuring voice at that moment by reminding the public that this is a normal part of the process—because it is.

Many businesses understand that a stable economy depends on a stable democracy. And a strong democracy is an inclusive one. The ACLU and our allied organizations are working with businesses and business leaders so they are prepared should we encounter an election emergency. In that event, businesses become an even more vital voice and must understand how to engage policymakers as well as the public to help reinforce facts surrounding election processes.

General Mills finally fixed the age-old problem of stale cereal

3 weeks ago

You can finally get Cheerios in a resealable bag.

General Mills announced Tuesday it partnered with SC Johnson’s Ziploc brand for Cheerios with Ziploc PantrySeal, its trademarked name for a new bag design. It features a blue Ziploc seal that goes partially across the top of the bag and doubles as a handy pour spout. The specialty box is available only for Cheerios and Honey Nut Cheerios and exclusively at Walmart’s website, where a 14.4 oz box retails for about $5.

It’s a wonder it took so long. General Mills says it receives hundreds of thousands of requests for resealable cereal bags each year. The company says market research shows nine in 10 cereal lovers already either fold, roll, or clip their cereal bags in addition to closing the cereal box to preserve freshness, or transfer their cereal to another container. Still, General Mills says, it took quite a bit of development and testing to finally make one.

[Image: General Mills]

“With nearly one box of Cheerios produced every second, we needed to develop a solution that could work reliably at scale while still delivering the simple intuitive experience our consumers were seeking,” Emilie Knox, General Mills vice president and business unit director for Cheerios, tells Fast Company.

The bags are produced at the company’s Buffalo facility, which is where Cheerios were first made in 1941, and General Mills says it had to make sure it could produce the bag without any disruptions to its existing production.

Malt-O-Meal, which sells off-brand cereal brands like Tootie Fruities and Berry Colossal Crunch, introduced its own cereal bag zippers in 1998. In 2011, its “Bag the Box” campaign helped launch bagged cereals, which it promoted as more environmentally conscious since it produced less waste, and which it still sells today. With its new bag, Cheerios is just now catching up.

General Mills doesn’t have any plans to announce resealable bags to other brands, but “we’re always listening to what consumers want and need,” Knox says.

Breakfast cereal sales generally have been shrinking, but the General Mills COO, Dana McNabb, said on the company’s earnings call last month that its cereal sales were up thanks to factors like smaller boxes at lower price points, protein cereals, and collaborations with the likes of KPop Demon Hunter and GloRilla. By adding a Ziploc seal to their top-selling cereal, General Mills is hoping to give consumers another reason to shop their brand.

Checked
12 minutes 8 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