• Elon Musk said a shortage of training data meant teaching Tesla's Optimus robot is like raising a toddler.
• Tesla has built an "Optimus Academy" where robots can learn through "self-play," Musk told investors.
• The EV giant said on Wednesday that it had begun production of Optimus.

Tesla is sending its Optimus humanoid robot back to school.

Speaking on the company's earnings call on Wednesday, CEO Elon Musk compared training the EV maker's humanoid robot to raising a toddler.

Musk said that while Tesla's autonomous driving models can learn from data collected by its massive fleet of vehicles, its Optimus robot has to learn by watching people do tasks and practicing them itself, a process similar to how humans learn.

"You can see this when you have a toddler, and they initially are sort of wobbly, and they don't know how to do things, and eventually, they learn to wipe their own ass, which is a big milestone," Musk told investors.

"They learn to read and do all sorts of things, and they learn this by playing. And we actually need Optimus to play at scale in order to have it be really good at tasks in the real world," Musk said.

The Tesla CEO added that the company was building out an "Optimus Academy" where prototypes can practice a wide range of tasks and generate training data to teach future versions of Optimus.

"For Optimus to get the training data we need, we actually need to make a lot of Optimi. So we need to make, I think, on the order of 10,000 or 20,000 bots and then have them do self-play," said Musk.

Tesla said on Wednesday that it had begun production of Optimus at its Fremont factory, with Musk saying the company expects to begin delivering Optimus to customers by the end of next year.

The near-trillionaire cautioned that production would be "somewhat bumpy" and that development of the humanoid robot still has some way to go.

He listed the challenges of designing durable yet responsive robotic hands, as well as AI capable of handling open-ended tasks, among the main hurdles Tesla faces.

Tesla is not alone in setting up robot academies to provide training data for humanoid robots.

Unlike large language models like ChatGPT, which were trained on the vast corpus of text data available on the internet, robotics faces a shortage of real-world training data to teach humanoids to navigate the physical world.

Other startups, such as Texas-based Apptronik, have also set up "data factories" where robots load conveyor belts and sort objects to generate training data to combat the data shortage.

• Hugging Face released a new AI model that allows users to "chat with DNA."
• Carbon-X can analyze and generate DNA and allows researchers to ask questions in plain language.
• Hugging Face said the model can help researchers who work outside the big AI labs.

Hugging Face is releasing an AI model that could make it easier for scientists to study DNA and develop new genetic treatments, the company announced Thursday.

Carbon-X is a "first of its kind" model that combines DNA analysis and generation with plain language, allowing researchers to ask questions and essentially "chat with DNA," according to Lewis Tunstall, a senior research scientist at Hugging Face.

Hugging Face, which Nvidia agreed to acquire for nearly $13 billion in September, is a platform for sharing and developing AI models and also releases its own. The company worked with research labs in Italy and China on Carbon-X, which is an open-source model. Hugging Face built Carbon-X by training an existing language model, Qwen3-14B, on genomic sequences and natural language. It was trained on DNA sequences from humans, other mammals, plants, and fungi.

As an example of how Carbon-X can be used, Tunstall said a researcher could provide a DNA sequence and ask whether a particular mutation could increase a person's risk of disease. The model would then offer a prediction and explanation. It can also generate new DNA sequences, which could help researchers develop new gene therapies.

Tunstall said the company was partly motivated to build Carbon-X to prevent biology researchers from becoming dependent on the major AI labs, noting how AI has disrupted mathematics.

"If you want to really push the frontiers of mathematics, you likely have to be at one of these AI labs like OpenAI or Anthropic, because they're the ones who have the best mathematics models," he said, suggesting a similar situation is likely to play out in other fields of science.

"We wanted to provide the open research community with tools that don't make them fully dependent on some external single provider," he added.

Researchers can run Carbon-X on their own computers, avoiding the need to share sensitive patient data with external AI companies. He also said that some of the misalignment behavior observed in larger models, such as the OpenAI agents that famously hacked Hugging Face in July, has not yet been observed in smaller, more specialized models.

Tunstall said generalist chatbots also sometimes refuse to answer legitimate biology questions from researchers because they mistake them for attempts to "design a bioweapon." He said viruses were explicitly excluded from Carbon-X's training data in order to avoid misuse of the model.

While Carbon-X is likely to be used by researchers, Tunstall said the model represents a small step toward a future where AI can help provide more personalized medical care.

"I think the holy grail is, in general, this concept of personalized medicine where we're able to, as individuals, take our own patient records, our own genomes, and use this to ultimately find or even maybe design our own treatments," he said.

• Google CEO Sundar Pichai said no Trump asked him to rename the Atlantic Ocean, the president said.
• Trump has floated the idea of renaming either the Atlantic or the Pacific to the "Trump Ocean."
• Earlier this year, he issued orders to change the names of the Gulf of Mexico and Lake Ontario.

President Donald Trump's plan to rename the Atlantic Ocean hit choppy waters when he took it to Google.

At a Wednesday rally in South Carolina, Trump said he considered putting his name or "America" on one of the two oceans that flank the US: the Atlantic and the Pacific.

"I’m not going to put either Trump or America on any other ocean but these two," he told the crowd.

Then, he said, he called Google to ask about making the change on Google Maps.

The tech giant had previously changed its US labels for the Gulf of Mexico and Lake Ontario to the Gulf of America and Lake America after executive orders from the president.

"I spoke to a brilliant guy, Sundar. He heads all of Google," Trump said, referring to CEO Sundar Pichai. He also praised Google cofounder Sergey Brin, calling both men "totally brilliant guys," before joking that they were "still rattled from the Gulf."

Trump recounted putting his idea directly to Pichai: "Sundar, I’m thinking of the Atlantic Ocean. Can you handle it?"

"No, no, no. This time I’m not doing it," Trump quoted Pichai as replying. "Oceans are different, sir. They cover really a lot of countries. Please don't ask us to do that."

"So I'll probably pass," Trump told the crowd. "Should I pass it?" He then answered his own question: "I think let's keep it the Atlantic Ocean."

Google did not immediately respond to a request for comment.

It was not Trump's first ocean pitch. At a Tennessee rally in early October, he floated "Trump Ocean" or "Ocean of the Americas" for the Atlantic or Pacific.

He has also mused about renaming a river, calling New Mexico "New America," and rechristening the Strait of Hormuz the "Trump Strait."

While those ideas remain proposals, Google Maps did adopt "Gulf of America" and "Lake America" for US users after the names changed in federal records. Users in Mexico and Canada still see the original names for the gulf and lake, respectively. Users in the rest of the world see the original names, followed by Trump's new names in brackets.

The changes have created a dilemma for people who make maps for an international audience. "The issue of international acceptance is also a factor, as some nations may not accept it," Joseph Wright, CEO of Replogle Globes, told Business Insider's Joshua Nelken-Zitser after the lake order.

"We sell globes worldwide."

For now, Trump's Atlantic idea appears to have sunk before it reached Google Maps.

• Elon Musk said he's going to give every child in the US a share of his space company, SpaceX.
• That would amount to $10.7 billion, based on SpaceX's stock price and the 72 million US children.
• Governments and billionaires have previously explored ways to spread corporate ownership.

Christmas came early: Elon Musk wants American children to own stock in his space company, SpaceX, next year.

Musk wrote on X around 1 a.m. Monday that he planned to make the donation next year — "the first of many," he said. SpaceX went public this summer.

Giving one share to each of the estimated 72 million US children would cost roughly $10.7 billion at SpaceX's closing price of $147.92 on Friday. Musk's 4.77 billion SpaceX shares were worth about $706 billion at that price, so the gift would reduce his holdings by about 1.5%.

President Donald Trump said on Truth Social on Monday that the shares would be transferred through Trump Accounts, tax-advantaged investment accounts for eligible American kids launched this summer. Under temporary Treasury rules, donated shares generally cannot be sold for five years.

"This is exactly what was envisioned when we created TRUMP ACCOUNTS," Trump said. "Bringing together the GREAT PRIVATE SECTOR, Elon and so many more, to invest in our children."

The White House said nearly 70 million Trump Accounts have been created. Numerous entities have agreed to contribute, including a $6.25 billion donation from billionaire couple Michael and Susan Dell for children in lower-income zip codes. In July, SpaceX President Gwynne Shotwell and her husband pledged personal SpaceX shares worth at least $321 million, prioritizing lower-income communities.

The White House, Elon Musk, and SpaceX did not immediately respond to a request for comment from Business Insider.

Musk's plan to give away SpaceX stock raises questions about a new kind of billionaire philanthropy: Gifting people shares in a company instead of cash.

Giving millions of children a stake in SpaceX could give them money to use later in life — but the value is tied to a single company's stock. And history shows that efforts to spread corporate ownership can yield uneven results, from big investment returns to stakes being sold for quick cash.

Here are other historical efforts to share wealth or spread corporate ownership among everyday people.

US Sen. Mike Gravel's 1978 share proposal to Alaskans

In 1978, US Sen. Mike Gravel proposed a statewide ownership corporation that would have given Alaskans, including children, stakes in major energy and other business projects. The Trans-Alaska Pipeline System was one potential investment.

The idea was that Alaskans would benefit from an economic boom and any rising asset values. However, the plan never materialized.

Alaska instead developed a dividend system that draws on its preexisting Permanent Fund, a state fund established in 1976 with oil royalties that invests in stocks, bonds, and other assets. It began paying annual cash dividends to eligible Alaskans in 1982.

The Permanent Fund still continues today, with a $1,200 dividend distributed to every eligible Alaskan adult and child who applied in 2026. The initial dividend this year was set at $1,000, but recipients also got an extra $200 energy-relief payment.

Eastern Europe's 1990s mass privatization experiment

Following the collapse of the Soviet Union in 1991, several countries in Central and Eastern Europe used voucher programs to transfer former state-owned companies into the hands of everyday people.

Russia began distributing privatization vouchers to roughly 144 million people in 1992. Russians could use them to buy shares in former state companies, exchange them for shares in investment funds, or sell them for cash.

Czechoslovakia used a similar system, allowing adults to buy low-cost voucher books and use the points to bid on company shares or invest through funds. After Czechoslovakia split in early 1993, the Czech Republic continued its voucher program, while Slovakia canceled its planned second wave of voucher privatization.

Other countries, including Poland, Romania, Bulgaria, and the Baltic states, adopted their own versions of mass privatization.

The programs were designed to rapidly create broad public ownership. But many recipients sold their vouchers for cash amid economic uncertainty, rather than retaining shares.

Evžen Kočenda, a professor at Charles University's Institute of Economic Studies, said that the program was a good idea at the time, given the region's poor economy, but that ultimately, the vouchers were not 'real money,' but rather a financial asset.

Kuwait's 2019 national stock exchange privatization

Kuwait privatized Boursa Kuwait, the operator of its national stock exchange, in 2019, and gave Kuwaitis the opportunity to purchase a stake in the company. The only requirement for registration was that the individual was a Kuwaiti citizen.

Those who participated — about 83,000 of the country's 4.4 million — were initially allocated up to 70 shares each. The initial public offering share price was set at 100 Kuwait fils, roughly $0.33 per share.

Following the privatization, Kuwaiti citizens and private-sector investors together owned roughly 94% of Boursa Kuwait. The Public Institution for Social Security held the remaining 6%.

The Kuwaitis who participated saw a huge return on investment: When Boursa Kuwait debuted on the public market in September 2020, its shares rose more than 10-fold from their 100-fils IPO price, closing at 1,055 fils, or about $3.41.

Today, the privatization is widely regarded as a major success in the investment world, and Boursa Kuwait has described it as the Middle East's only privately owned stock exchange.

• This post appeared in the Business Insider Today newsletter.
• You can sign up for Business Insider's daily newsletter here.

Some pairings feel inevitable. "Friends" had Ross and Rachel. "The Office" had Jim and Pam. Elon Musk has SpaceX and Tesla.

But a deal between Musk's two giant companies could end up looking more like Frankenstein's monster than a Hollywood love story. And investors might struggle to understand and value it.

The SpaceX-Tesla tie-up isn't a shoo-in, but both sides are officially considering it. That raises questions about what's gained from a merger.

Combining the companies turns Musk's two jobs into one. Their somewhat overlapping AI plans could create significant efficiencies in areas like compute and talent. Joe Ciolli, author of the First Trade newsletter, broke down the pros and cons for SpaceX and Tesla investors.

These companies have A LOT under their hoods, though. Here are just some of the businesses or long-term bets that SpaceX and Tesla are pursuing:

• Space-mission services (Falcon 9, Dragon, Starship)
• Internet service (Starlink broadband)
• Mobile carrier (Starlink Mobile)
• Electric vehicles (Model 3, Model Y, Cybertruck, Roadster)
• Government satellite systems (Starshield)
• Autonomous vehicles (Robotaxi)
• Humanoid robots (Optimus)
• Energy storage (Powerwall, Megapack)
• Frontier AI models (Grok)
• Social media (X)
• Payments app (X Money)
• AI infrastructure (Colossus)
• AI coding tools (Cursor)

And don't forget Musk's grandest ambition: colonizing Mars.

A conglomerate with such a diverse set of businesses can become a headache to value.

Let's start simple: How would you describe a SpaceX-Tesla entity?

A space company? An AI company? A robotics company? Saying it's "a little bit of everything" might not sit well with investors.

When a company plays in a bunch of industries, the market doesn't always give it the benefit of the doubt. There's a term for this phenomenon: conglomerate discount.

It means a company's stock is valued lower than the sum of its individual parts. Some investors fear the company will be too complex to manage, or too expensive to run. Others worry its combined financials will make it hard to understand its true health, or stronger parts of the business will subsidize weaker ones.

A widely cited 1995 paper on the topic found the discount to be between 13% and 15% among a sample of diversified companies from 1986 to 1991.

Some academics and economists have argued the conglomerate discount is a relic of the past. A 2024 McKinsey report said the discount wasn't necessarily automatic. Instead, heavily diversified companies often had weak performance or lacked financial transparency, which made it harder to assess their true value.

Either way, these types of companies require investors with a broad appetite. (Or a tough stomach.)

Imagine you're excited about Starlink's potential, but not as keen on humanoid robots. Expensive robotics projects could suck up the budget you were hoping would go to Starlink. Or the money generated by Starlink could subsidize the losses accrued in an underperforming division.

It's why giant, sprawling companies are often less appetizing than a pure-play stock.

Ultimately, a deal is often about selling a vision to investors. And there's arguably no one better at doing that than Elon Musk.

Now he just needs to prove SpaceX and Tesla were always meant to be together.

• After expanding Trump Accounts to all children, the administration may look to adults next.
• We spoke to five experts about what should be in a Trump Account program for adults.
• Accounts should support the millions of Americans who don't have retirement savings, they said.

Trump Accounts, created for children, may be coming to adults, too.

Brad Gerstner, an architect of Trump Accounts, said on Wednesday that the tax-advantaged accounts should be rolled out to every US citizen under 40, though he flagged that Congress would need to pass new legislation to make that happen.

President Trump similarly said on Tuesday that the White House is exploring creating an "alternative" Trump Account for adults.

"We are looking at something different because it really is somewhat of a different thing having to do with adults," Trump said.

So what might that look like? We spoke to five experts about how Trump Accounts could work for adults, who they should be for, and how they could figure into financial planning.

Ray Boshara, a senior policy advisor at the Aspen Institute and Washington University in St. Louis' Center for Social Development, said the accounts could be a boon to the 46% of Americans without money in a retirement account, according to Federal Reserve data.

"I really like the idea of all Americans having a lifelong investment account through which they can build wealth and achieve financial security," said Boshara, who advocated enrolling all American children in Trump Accounts, a change that was rolled out earlier this month.

He sees expanding Trump accounts as an extension of the same push, as long as the accounts are automatically created for those adults, too.

Focus on those without retirement savings

There's a "pretty robust retirement savings system" in place for employees, said Boshara, but nearly half of Americans don't have retirement savings, many because their employer doesn't offer a plan.

Trump Accounts could "fill that huge gap," Boshara said, but unlike the plans for children, it may be worthwhile to focus on those without 401(k) or similar plans.

"One of the benefits of the Trump accounts for kids is that it's universal, but I'm not so sure you'd want it to be universal for an adult," Boshara said. Traditional IRAs already limit tax deductions for some higher earners, and that's a "good thing," he added.

"That way, you're not just giving wealthier Americans yet another tool," Boshara said.

If the goal is to help Americans without retirement, then Boshara's push away from universalism makes sense.

But Cem Karsan, founder of the hedge fund Kai Volatility Advisors, sees this as part of a broader effort to stem the tide of populism by giving regular Americans a greater stake in the stock market.

From that perspective, a universal plan is essential.

"It's essentially the infrastructure that's being laid, the pipes, for delivering ownership broadly to the masses," Karsan said.

For those with retirement accounts, a Trump Account, technically another type of IRA, could allow them to contribute more than the current cap, said Stephanie Link, chief investment strategist at Hightower Advisors.

"There would be an additional layer of tax-deferred wealth accumulation," Link said.

Don't mess with benefits

One of the main concerns with rolling out Trump Accounts to low-income Americans is that they could be kicked off public benefits like food stamps and Supplemental Security Income. This wasn't an issue for children, but many benefits available to adults can be means-tested, and contributing to an account may make you ineligible.

Right now, for traditional IRAs, the rules are a "patchwork," Boshara said. A single person with more than $2,000 in a traditional IRA can lose access to Supplemental Security Income, while Medicaid rules vary from state to state.

"What I'd call for is a categorical disregard," said Boshara, meaning money in the accounts wouldn't count against eligibility for public benefits.

"If you expect low- and moderate-income people to save and build wealth through an IRA, you have to protect their public benefits."

Maybe most importantly, Boshara said these should be a "supplement" for Social Security.

"We never see these as a substitute for Social Security," Boshara said.

Fund them

The most transformative part of the Trump accounts so far has been the $1,000 available to every US citizen born between 2025 and 2028. But Boshara's biggest critique of the program is that families have to open up to receive government payments and employer contributions.

"Opting in is suppressing participation, especially among lower-income families because they don't have the awareness, or they don't interact with the tax system," he said.

He's hoping for Treasury rules that will activate accounts, so contributions can be added automatically, and that the $1,000 funding for newborns be continued indefinitely. And similarly, he's hoping that anyone can put money into adults' funds, like with children.

"The best thing about Trump Accounts for kids is that there are multiple investors in somebody's future," Boshara said. This could even include employer contributions.

"I'm not sure if all employers would contribute, but if they did, it's another savings benefit plan, HSA or 401K type of plan, which is again, more money in their pockets," Link said.

For those who already contribute to their 401(k) and other vehicles, like Roth IRAs, which can be a better fit for early career savers, the outside funding is the big beneficiary, said Christina Taylor, VP of tax development and delivery at tax fintech April.

"For adults saving their own money, a Roth IRA or a 401(k) with an employer match is often still the stronger tool," Taylor said. "Savers should think of a Trump Account as a collector for other people's contributions, not a replacement for retirement accounts they already have."

Treat adults like adults

Current Trump Accounts come with a range of restrictions that keep investors focused on the longer term, including a five-year holding period for individual stocks. That restriction on selling ends when they turn 18, and Boshara said that adults should be able to decide what they want their portfolio to look like.

"You want investor choice," he said. "You want to empower families to make good choices."

Trump Accounts hold only equity investments, so adult versions would require greater diversification.

"A 10-year-old can ride out a bad market, but someone five years from retirement needs some bonds, some cash, and a plan for withdrawals," said Brian Boswell, a wealth manager at Savvy Wealth.

It's a "great idea to make sure all Americans have a lifelong investing account," but the key is to ensure it meets adults' needs, Boshara said.

"Why not extend it and keep the great features of Trump accounts and discard the ones that don't make sense for adults?" Boshara said

• Walmart chief people officer Donna Morris oversees more than 2 million workers globally.
• She shared three green flags that make her lean in during a job interview.
• Morris said she likes to see a track record of grit in a candidate.

Donna Morris oversees over 2 million employees, so she knows a strong candidate when she sees one.

Walmart's chief people officer identified three green flags that make her lean in during a job interview — regardless of the role they're applying for. First, she told Business Insider that she looks for a genuine alignment with the company's purpose and values.

Second, she wants to see an ability to take on multiple responsibilities and make things work.

"They demonstrate a track record of grit," Morris said, defining those individuals as people who have overcome adversity.

In an ideal world, she said that means they've worked on the frontline, since so many of Walmart employees have customer-facing roles. She said the company looks for candidates with a focus on service.

Morris, who started working at 13 at a dry cleaner's and then worked at a drugstore for eight years, said that if a candidate has never experienced what it's like to be a frontline hourly worker, it may be challenging for them to lead. Morris said those jobs gave her purpose, and she found it hard to resign.

Now, when she interviews candidates, Morris said her go-to question is "If I was to ask people who've worked for you before, what will they tell me?" She said the open-ended question generates a genuine response and helps her gauge whether the person is focused on servant leadership, which is a "big green flag."

Asking this question allows her to hear more about a candidate's first job or a work experience. She said that while graduating from an Ivy League school is valuable, she's more focused on hearing about what they've learned from work experience.

"What was the grit they had to persevere through?" Morris asked.

The chief people officer, who worked at Adobe for more than 17 years before joining Walmart, said that her third green flag is candidates who are passionate about making a difference.

"That, to us, would be really, really key," Morris said.

This as-told-to essay is based on a conversation with Bridget Griswold, the 25-year-old CEO of Lean In, who lives in San Francisco. It's been edited for length and clarity.

I fell in love with technology at Brown University, where I studied computer science and business. I worked with two female professors to build a robot. I interned at a semiconductor manufacturing company and also loved design and art.

Product management felt like a really wonderful combination of all those interests, and that was my role at Meta a year after graduating from college. My team and I redesigned the Instagram search grid, so now when you search for large Instagram accounts, you're routed to the search results page rather than directly to those accounts. I'm very proud of that.

While at Meta, an incredible female mentor connected me to Sheryl Sandberg. Meeting her felt so surreal and is what led me to start my role as CEO at Lean In in March. It feels like a once-in-a-lifetime opportunity to do mission-driven work.

I was a swimmer as a kid, and as my swim coach used to say: "You get out of swimming what you put into it." I think that applies to many things in life: The more you put in, the more you get out.

Here's a day in my life.

I wake up to my needy, angelic cat around 7:30 a.m.

The second I wake up, I pet my crying and whining cat. She's an angel but very needy.

I'm not a breakfast person, but I need coffee immediately. I have many cups of coffee throughout the day. One of the first things I do is make some at home.

I always tidy up my apartment while brushing my teeth

The most productive two minutes of my day are when I brush my teeth. I clean and tidy my apartment at the same time.

I'm probably not super focused on brushing my teeth, which is maybe a problem, but I'm also making my bed, emptying the dishwasher, feeding my cat, or starting the coffee maker.

I always move really quickly; I don't like to sit still. Brushing my teeth and standing still would make me antsy, so I like to get something else done at the same time.

I'm still adjusting to West Coast hours

I moved from New York City to San Francisco in August.

In New York, I went to Solidcore classes before work every day. Working on the West Coast, the workday starts earlier than on the East Coast, so I'm still learning whether I like to work out in the morning or after work.

When I take the bus to the office three days a week, I often use that time to catch up on personal housekeeping tasks like scheduling appointments or texting back my friends who live on the East Coast.

I come to the office around 8 a.m. or 8:30 a.m.

A majority of my in-person time is spent in meetings. As CEO, I enjoy being involved in all aspects of the team.

I'll meet with our content and research team — we spend a lot of time digging into data, such as some really exciting new research on trad wives in the manosphere. I also often meet with our product and tech team, as well as many of our partners, to help grow our programs.

We have lunch in the office — my favorite is a salad-plus-smoothie combo — and it's a really wonderful time to connect with the team.

I also love connecting with our Lean In community. One thing we're focused on is closing early gender gaps in AI usage. Our flagship program, Circles, consists of groups of 8 to 12 women who meet regularly to support each other in their careers — we have over 170,000 members. It's a perfect forum to hold each other accountable for upskilling in AI.

Every week, I send AI tips to our network leaders. A recent one encouraged the use of AI notetakers, as women are more likely to take notes during meetings and miss out on participating.

Lean In is in over 180 countries, and it's really heartwarming to see that our commitment to our mission has a positive impact on women's careers. Getting to speak with them is really energizing and rewarding.

I rehearse for meetings by talking to myself on walks

Another habit of mine I've been told is weird is how I prepare for big meetings.

If I have a presentation or interview coming up, I'll set up a Zoom and practice speaking to myself. Talking for a long time is sort of an unnatural act, and the first time you do it is probably not going to be your best.

Sometimes, I'll even go on a walk and wear headphones, pretending like I'm in a meeting or on a phone call, but I'm actually talking to myself and practicing.

When there's a bug, I usually fix the code myself

Because of my background in computer science, I spend a lot of my time finding bugs or things I believe we should change on our platform. I'll code and fix those on my own. Thankfully, with AI, it's so much faster to code now than it was a few years ago.

Our team is busy, and my goal is to enable everyone to have as much impact and move as quickly as possible. If I can help unblock the tech team or resolve an issue rather than add it to their plate, I'm always more than happy to do so.

I leave work around 6 p.m.

I continue to work until I go to bed pretty much every night.

It's nice to have that time at home, when I'm not running around between meetings, to focus, give feedback on documents the team has sent, write, respond to emails, and brainstorm my own ideas.

I take breaks to tidy up my apartment and make dinner. Living in New York got me out of the habit of cooking, so I'm taking the initiative to cook more meals, like rice bowls. I would say about half the time I order takeout, like pho, and the other half I cook, which is progress for me.

If my work doesn't require intense focus, I'll do it from the couch and put a happy TV show on, which right now is "Schitt's Creek." I think it's a nice balance between relaxation and getting things done.

I usually get in bed around 10:30 p.m.

I've been trying to make a habit of reading every night instead of going on my phone, which has been a really positive change. I like anything I read to be a distraction. I'm currently reading "Big Little Lies" and enjoying it.

It's nice to spend 30 minutes reading something like fantasy that's completely different from what I think about at work. That way, I'm not spinning on all the things I need to do or coming up with new ideas that can't get done in my sleep anyway.

Weekends are for the beach and farmers market

My boyfriend works in Los Angeles and flies up to San Francisco every weekend. We've made a tradition of going to the Fort Mason Farmers Market on Sundays for fresh produce to cook an at-home meal that night.

I also enjoy seeing friends over the weekend. My preference is to go on a long walk or go to the beach. I also enjoy fitting in some work — emails, deep-focus work, or reading — throughout the weekend.

• Anthropic platform engineering head Katelyn Lesse said that Claude now writes all her team's code.
• Lesse shared ways that Claude Code has helped the company become more efficient.
• Anthropic nears an initial public offering at a potential $2 trillion valuation.

A top Anthropic engineer says that Claude Code has already wiped out all human coding at one of the AI lab's central teams.

Katelyn Lesse leads the AI startup's platform engineering group, which handles the company's core services, developer tools, and product infrastructure — it's essentially the engine room of Anthropic. Lesse said at Grafana Labs' ObservabilityCON event on Tuesday that among her staff, which tallied 200 as of this summer, the era of human-written code is fully over.

"When it comes to our team and how we're actually building and maintaining our systems, all of our code is written by Claude," Lesse said. "Like, literally all of it."

Anthropic's heavy reliance on Claude Code has helped it remain relatively small, as it approaches an initial public offering that could value it at nearly $2 trillion. The company's revenue growth slowed in the third quarter of this year. Maintaining its lofty valuation may depend on selling more companies on the AI-first approach espoused by Lesse and other Anthropic leaders.

Lesse said that people on her teams are still designing systems and verifications, and then offloading the actual coding to Claude. She said that when the company changed the programming language of a major codebase from Python to Rust, workers developed a way to verify the final product.

"But the grunt work of building the pieces and actually monitoring the pieces and actually making sure that we felt good about it, that was agents all the way down," Lesse said.

Lesse said Anthropic already lets agents fix tests autonomously, with no human interaction, and it's thinking about how staff can get comfortable with agents autonomously making changes to the actual codebase, given the improvements in AI models over the last year.

"As engineering leaders, we're like terrified, and we're like, let's be really careful about this, but I think the line is starting to push," Lesse said.

Anthropic has also embraced Claude Code companywide. It released a report last week saying that as AI agents have worked more inside the company, it has hired for a smaller share of junior technical positions. And in May, Anthropic said that Claude then wrote more than 80% of the entire company's production code.

Have a tip? Contact this reporter via email at scouncil@businessinsider.com, or over text, Signal, Telegram, or WhatsApp at 415-757-8198. Use a personal email address, a non-work WiFi network, and a non-work device; here's our guide to sharing information securely.

• A sign instructing San Francisco gym-goers not to vibe code between sets went viral.
• The sign was an "AI-generated fake," according to the gym.
• Amid tech's vibe-coding obsession and fitness boom, it's no surprise the image has gotten heat.

Don't hog the leg press while prompting Claude.

A screen with the logo of Fitness SF, a Bay Area chain of gyms, went viral among techies on X and Instagram on Monday. The image instructed gym-goers to refrain from writing code or prompting AI during their workouts. "Be present. Lift harder. Ship later," the image reads.

Troy Macfarland, Fitness SF's director of marketing, said, however, that the image was an "AI-generated fake." The gym does have etiquette signs that mention phone usage and being courteous, he wrote, but none mention vibe coding.

"This person altered our actual images displayed on the screen," Macfarland wrote. "I will give an 'A' for creativity to him."

It's not clear whether the image was generated with AI, Photoshop, or some other tool. Ian Brash, whose post with the image got almost 7,000 likes, did not respond to a request for comment.

In San Francisco's AI mania, it's no surprise a sign like that blew up on X. Techies weighed in, cracking their own jokes about how tied they were to their coding agents.

One X user told Fitness SF to give their copywriter a raise. A Cognition engineer said that vibe coding between sets was encouraged at his company's gym.

"Vibe coding at the gym?" one X user wrote. "You're worried about the wrong weights."

One reaction image often appears in the comment sections of posts like this one. It shows a muscled man, holding a giant nuclear bomb in his hands, jumping into the city of San Francisco. One VC posted it with the message: "Even I have limits."

Claude Code and cardio

The sign tapped into two of the buzziest subcultures in tech.

First, there's AI coding. Ever since the launch of Claude Code and OpenAI's Codex, engineers have been tied to their phones (or walking around with half-open laptops) to continue building. Some techies told Business Insider that their love of vibe coding created distance between them and their girlfriends.

Then there's the gym. Tech bros are getting fit, pumping themselves full of peptides and getting shredded at early-morning Equinox sessions. A developer at the gym used to be an anomaly; now, it's the norm.

Some techies seemed skeptical from the start. One X user quoted the post, writing that there was "no way this is real."

Another user responded that they didn't know about that specific sign, but they did know about the underlying issue. "This is a real problem at the Equinox in San Francisco," they wrote.

• One of Meta's AI researchers left the company this month to start a legal tech company.
• Ethem Coban worked on teaching Meta's language models to reason through legal questions.
• Since then, he's had talks to raise $6 million in funding for a startup, according to a person familiar.

The legal industry's artificial intelligence gold rush is luring talent from Silicon Valley's biggest labs.

Ethem Coban, a researcher at Meta, left the company earlier this month to launch a legal tech startup, according to a person familiar with the matter. Coban, who worked on teaching Meta's language models to reason through legal questions, has been in talks to raise $6 million at a $25 million valuation.

Talks are still early, and the terms of the deal could change.

The startup is the latest entrant in an increasingly crowded legal tech market. Long seen as laggards, lawyers have spent the past two years experimenting with tools built on large language models. Early movers like Harvey and Legora saw a crush of demand, setting off a wave of investment in startups promising to transform how lawyers work.

Now, that opportunity is attracting researchers from the companies building the underlying models.

Coban spent almost two years at Meta, most recently at its Superintelligence Labs, where he helped train models to give better answers to legal questions and reduce hallucinations, the person said. He previously studied law at American University and Goethe University in Frankfurt.

A spokesperson for Meta didn't respond to a request for comment.

Coban has told investors that his new company aims to help law firms and corporate legal teams build and customize models around their expertise. The bet is that companies will increasingly want to own their AI rather than license it from outside companies, because it gives them more control over their data.

Law firms also see a chance to turn mountains of case files into a competitive advantage that other firms can't simply buy off the shelf. Kirkland & Ellis, the industry's revenue leader, has committed $500 million to building proprietary software. Latham & Watkins has been buying its own Nvidia servers to customize models in-house.

For Coban, the challenge won't be convincing law firms that they need artificial intelligence. It will be persuading them that someone who knows how the models work from the inside is best suited to build them.

• Newcomers like Kevin O'Leary face skepticism due to a lack of experience in data center development.
• Bernstein analysts dubbed O'Leary and others "bragawatts."
• Most announced projects will never happen. "Press releases are cheap…concrete is not."

The AI industry has a problem with bragging.

Companies are announcing enormous new data centers to power the AI boom. Some promise facilities bigger than anything the industry has built before.

There's just one problem: Most of these projects may never happen.

Bernstein Research analyzed thousands of planned US data centers and concluded that only about 35% of the announced capacity is likely to get built. That's roughly 170 gigawatts of data center capacity out of a total of almost 500 gigawatts.

The rest? A significant chunk of it belongs to a category the industry has a wonderfully dismissive name for: "Bragawatts." These are ambitious announcements, often from newcomers with little experience building data centers, that have a slim chance of becoming reality.

"Press releases are cheap…concrete is not," Bernstein analysts wrote in a research note.

Bernstein began its analysis with more than 7,000 data center projects across North America, drawing on information from Aterio, a company that tracks construction through satellite imagery, permits, and other records. The analysts narrowed that to more than 5,000 projects (partly by excluding those already canceled).

Then they assessed the chances of these remaining projects being completed. The analysts considered factors such as whether construction had started, whether electricity was available, and whether the companies involved had successfully built data centers before.

The results revealed a striking divide.

The 'bragawatt brigade'

Experienced operators including Amazon, Google, Microsoft, and Meta landed at the top. Bernstein gave these companies perfect scores of 5 out of 5 for their development credentials. Other data center specialists, such as Digital Realty, Equinix, and CoreSite, got top marks, too.

O'Leary Digital, associated with celebrity investor Kevin O'Leary, ended up at the bottom, with a credibility score of 0.35 out of 5, according to Bernstein. His company has announced 15 gigawatts of data center projects, and Bernstein estimates that only 489 megawatts, or about 3%, will actually materialize.

"Sorry, Kevin," the analysts wrote.

Although the term "bragawatts" has been doing the rounds in the data center industry lately, it's unclear who coined it, Bernstein analyst Madison Rezaei told me.

"I wish I could take credit for that," she said. "I believe I originally heard it from an Equinix executive, but I don't think he owned it either."

She said that while some ambitious newcomers could succeed, they haven't yet demonstrated they can deliver.

"It's not that all of these guys are sort of terrible and not going to deliver, but we just have lower confidence because we don't know them," Rezaei said.

O'Leary Digital responds

O'Leary Digital disputed Bernstein's assessment, saying the research doesn't account for substantial work that occurs before construction begins.

While acknowledging that O'Leary Digital is a young company, CEO Paul Palandjian said the team has decades of experience building major infrastructure projects and that its proposed 15 gigawatts of capacity would be built in stages over several years.

The company said it has secured binding agreements covering about 11,000 acres in Utah and 13,000 acres in Alberta, obtained a 30-year development agreement in Utah, and made progress on planning, engineering, and permits at both sites. It also said the projects would generate their own electricity.

The company said its early development work is funded, additional capital is committed to reaching key milestones, and experienced engineering and construction partners are involved.

"We would rather be judged on milestones than on a model," Palandjian said.

'An announcement in search of a project'

Other low scorers in the "bragawatt" brigade included Vermaland and Joule Capital Partners, according to Bernstein analysis based on Aterio data.

Joule has announced 50 data center projects with a combined capacity of 11.7 gigawatts. Bernstein estimates that only about 11% of that capacity will actually get built. Almost all of Joule's planned capacity remains at the announcement stage, with little evidence of construction progress or of major customers committing to use the facilities, Bernstein wrote in a July note to investors.

The analysts gave Joule a credibility score of 2.05 out of 5, placing it among the lowest-ranked developers in its database.

"An announcement in search of a project," the analysts wrote.

Vermaland has announced plans for a massive data center project in Arizona. Bernstein estimates that only 4% of its proposed 2.2 gigawatts of capacity will get built. The company is primarily a landowner, not a data center developer. Its proposed $33 billion La Osa project near Eloy had already been sharply scaled back amid local opposition and without construction approval. In late August, regulators formally rejected the project. Even before that, the analysts gave Vermaland a credibility score of 0.75 out of 5.

Joule and Vermaland didn't respond to emails seeking comment this week.

A yawning gap for data centers

Sergio Toro, CEO of Aterio, sees similar patterns in his firm's voluminous, detailed data.

Aterio's tracking shows that roughly 30 to 40 gigawatts of new data center capacity get announced each month. Yet only about 7 gigawatts begin construction each month, while 2 to 3 gigawatts actually start operating.

This leaves a yawning gap.

Many announced projects get withdrawn before construction begins — and before real money has been spent.

"They just make the announcement, but they didn't get their approvals. Or they make multiple filings in multiple locations to see which one gets selected," Toro told me. "That's why we track with satellite if there's real movement on the ground or not."

Indeed, Bernstein analysts wrote recently that the amount of "stranded" data center capacity has surged. These are projects that are either canceled, delayed, or not approved.

The AI boom's three tiers

Toro sees three broad groups competing to build AI infrastructure.

At the top are the established technology giants, including Amazon, Microsoft, and Google. They've built data centers for decades and generally deliver on their commitments. Most of their projects will get done, he said.

In the middle are "neoclouds," newer companies such as CoreWeave that rent out computing power specifically for AI. Although they're more vulnerable to financing problems and delays, many are making solid progress, especially because some of these players already have access to electricity.

"We see them advancing quite well," Toro said. "Sometimes we see some delays, but it's not something significant or any kind of big capacity that is at risk."

Then come the newcomers making extraordinary promises. That's where most delays and cancellations happen, Toro said.

Even Amazon needs a backup plan

Not every project that disappears was a fantasy, though. Even a top player like Amazon struggles sometimes to secure electricity and other supplies on time for these huge projects.

Business Insider's Eugene Kim reported exclusively this week on the challenges Amazon faces with its record AI buildout.

The biggest, most experienced tech companies deliberately pursue more potential data center capacity than they need because they know some developments will fail or get delayed.

Major AI companies were committing to three or four times the space and electricity they needed over the following two years, Jeffrey Moerdler, a data center lawyer, estimated during an interview with Bernstein analysts.

His reasoning: Of data center projects that haven't started construction, roughly a third will arrive on schedule, another third will be delayed, and the remainder won't happen.

Rezaei and Toro estimate that the top players don't over-commit that much — more like double the capacity they need at most.

The strategy provides a "Plan B," as Toro put it, in case of delays involving electricity, equipment, permits, local opposition, or other factors.

Texas has a ghost problem

This helps explain another extraordinary number.

Data center companies have requested connections for roughly 450 gigawatts of electricity demand on Texas' main power grid, more than five times the grid's roughly 85 gigawatts of existing capacity, according to Toro and others. 

Many of those requests are placeholders rather than firm construction plans. Toro calls them "ghost projects."

These speculative projects clog the waiting lists for electricity connections, making the electricity shortage look a lot worse than it actually is.

"Developers routinely spam projects across several sites, and many are speculative builds with no customer yet," Ryan McEntush, a partner at venture capital firm Andreessen Horowitz, wrote in a recent blog. "Often these new developers have never even plugged in a GPU, let alone a power plant."

Texas has paused applications to review and weed out projects that aren't serious. Toro, from Aterio, expects this to favor the established players at the expense of the bragawatts.

The real AI boom

None of this means the AI infrastructure boom is imaginary.

Aterio tracks thousands of buildings, including major projects already under construction. Bernstein estimates that nearly 80% of the capacity the largest technology companies are directly developing is credible.

And when speculative projects disappear, Rezaei doesn't expect a wave of unpaid debt. Many will fail before building begins, the stage when developers borrow serious money through project financing.

Some of those missing data centers were never really coming in the first place.

Sign up for the AI Insider newsletter here. Reach out to me via email at abarr@businessinsider.com

• Gen Zers are much less likely to have at least one kid than older generations at the same age.
• Not wanting kids, affordability problems, and medical challenges are plausible reasons.
• Having at least three kids is becoming less common across income percentiles, except in the top 1%.

Birth rates in the US have been falling for decades, but there's been an especially large drop in the likelihood of Gen Zers in their 20s having a kid compared with their older peers at the same age.

For 26-year-old Gen Zers, the likelihood of having at least one kid was 26%, down from 47% for millennials, 50% for Gen X, and 46% for boomers at that age. There were also differences between Gen Z and older generations in the probability of wanting another kid.

That's based on an analysis by a group of researchers published by the Federal Reserve Bank of Atlanta. Julie Hotchkiss, a research economist at the Atlanta Fed, told Business Insider the analysis controls for education, partnership status, and some health conditions. The analysis uses National Survey of Family Growth data up to 2023.

From affordability challenges to concerns about the future, there are several likely reasons behind the lower probabilities. Hotchkiss said that a lower desire to have kids is one explanation.

Gen Zer Brooklen Farley said in an as-told-to Business Insider essay in 2023 that she decided during college she didn't want to have kids. Farley is worried about issues like the economy and climate crisis.

"Many people call us Zoomers, but honestly, a lot of us are doomers: We feel pessimistic about the future and don't want to subject our children to it," she said.

The research also notes that there's been a rise in medical challenges related to having kids. The researchers looked at impaired fecundity, or trouble becoming pregnant or carrying to term. The average probability of impaired fecundity at 26 was 18% for Gen Z compared to 10% for boomers at that age. The probabilities are higher for older ages — the likelihood for 40-year-old millennials was about 30%, higher than the 26% for Gen Xers at that age.

"If we want to talk about birth rates, we can't just talk about preferences," Hotchkiss said. "It's not just that darned younger generation not wanting, just being selfish and not wanting to have children. There's also really a very real medical issue that needs to be addressed."

Another problem is all the financial costs that come with being a parent. LendingTree found the cost of raising a kid for over 18 years adds up to $303,418 after tax exemptions and credits, or about $17,000 a year.

Affordability issues aren't just a Gen Z problem. Surrounded by a large family growing up, millennial Lizzie Lawson said in a Business Insider essay that she assumed she would also have a large family one day.

"When I was a kid in the early 2000s, my dad was an electrician and sole provider for our growing family of eight," Lawson said in a Business Insider essay, adding that her family got by with the help of coupon cutting and hand-me-downs. "Money was tight, but we also owned a house in a safe suburb, and my parents sent six kids to Catholic school from K-8."

Lawson, a college instructor whose husband is a physical therapist with $100,000 in student loans, said the economy is different from the one she grew up in.

"We can't afford to have one of us stay at home, and there aren't enough coupons in the world to cover the rising cost of day care," she said.

Having large families is rarer than in the past

Michael Konczal, vice president of policy and research at the Economic Security Project, said in his newsletter that across income percentiles, families with at least three kids have become less common since the 90s, except at the top 1%.

"A third kid isn't new for them; what's new is that it has become rarer for everyone else," he said. "Compared with the 1990s, families with three or more kids became less common overall and among the rest of the top 10%. The top 1%'s estimated share barely changed, though the uncertainty is substantial. Maybe all the stresses that made childbearing more difficult over the past 30 years just bounced right off the 1%?"

The US fertility rate has cooled to 1.6 births per woman, below the 2.1 replacement rate, which Hotchkiss said would translate to slower population growth. The Congressional Budget Office projects that just a few years from now, the number of births in the US could fall below the number of deaths, which could have implications in future decades.

"Over time, the negative net contribution of births and deaths increasingly offsets the positive contribution of net immigration, until population growth slows to zero in 2056. The population is projected to shrink thereafter," the CBO said.

• I tested out the leading AI agents for a while, and gave them each the same tasks.
• OpenAI's Dot saved me from failing to file my taxes, but was the worst at ordering lunch.
• Ultimately, dealing with multiple AI agents at once is a hassle!

I've been playing around with various AI agents — Muse, Instinct, and OpenAI's Dot, and I've found them genuinely really useful for certain things.

But which one is best?

I put them to some, uh, semi-rigorous testing. Ok, not totally rigorous or controlled, but I gave them all the same 3 tasks — triage my email, order me lunch, and build me a personal website — and it was interesting to see which worked best. It was also interesting to see how differently they functioned at these tasks.

If you haven't started using an AI agent yet but are curious, well, I'm not sure that this is going to answer all your questions, but maybe is a place to get started.

A big spoiler here for my ultimate recommendation: right now, Meta's Muse is the only free one of the bunch, so in my mind, that wins for most normal people who just want to try an agent out. OpenAI's Dot is only for the $100/month plan (ouch!) and Instinct is currently invite-only.

To me, that's a no brainer. Do the free one (unless you're already paying for the ChatGPT plan, in which case you're probably already a pretty pro user and probably don't need my advice).

Challenge #1: Manage my email

I get a lot of email. Most of it I don't need to bother opening. Help me figure out which are the important ones.

Muse: Muse has been a dutiful soldier here, and this was pretty much the first task I gave it when I started using it a few weeks ago. So far, "triage my email" has been the one truly useful thing an agent has done for me.

Instinct: Overall, I don't like the fact that Instinct works mainly through your text messages vs. a dedicated app. While that's very convenient in certain ways, I don't like the idea of seeing email (bad, no fun) in my texts (good, fun) — mixing the two feels wrong.

However, I have to hand it to Instinct — it found a Paperless Post invite for a party that I would otherwise have completely missed in my personal email. It was the only agent to pull that one out for me.

Dot: Dot did something truly useful that required a lot of context, and it did all on its own. It noticed an older email exchange from last month with my tax accountant (I had filed for an extension, due in October). Dot noticed my accountant never confirmed he actually filed my taxes — yikes! This actually really saved me! (Don't worry, the taxes had indeed been filed.)

Finding an older — but important — email exchange that I hadn't closed the loop on is exactly what I want from an AI agent/assistant. I feel like 90% of things I'd rather do myself, but my human brain gets easily distracted and forgetful, and there are plenty of things I forget to deal with or get distracted.

Winner: Dot. Also me, for not going to jail for tax evasion.

Challenge #2: Order me lunch.

I asked each agent to order me a salad at a specific local restaurant. Because I didn't want to annoy real humans, I did the orders for pickup, to avoid any confusion or multiple separate deliveries.

This turned out to be a real fiasco, and I learned my lesson: it would have been infinitely easier for me to just do it myself.

Muse: Initially, Muse was very helpful. I accidentally gave it the wrong restaurant name, but Muse helped me find the right place and pulled up the correct menu options for me to see.

I had already connected my credit cards with Stripe, so I figured this would be easy. It offered up my debit card or another Mastercard, which I assumed was my regular credit card (I had just gotten a new card in the mail and didn't recognize the last four digits).

When I finally went to check my wallet, I discovered it was a mistake — it was actually the card connected to my health insurance-related HSA account, which is only approved for medical expenses.

By the time I straightened this mess out, I had already moved on to placing orders with Instinct and Dot. When I went to pick up my orders at the restaurant, there were only two bags with my name on them. Sure enough, I had gotten so distracted with the other orders that I never confirmed with Muse about placing the order, and it never sent through. (Can you see now what I was saying before about how I get distracted and don't finish tasks?)

Instinct: The most annoying part about Instinct, again, in my opinion is that it's over text messages. That means there were some quirks when I tried to text on my laptop — sometimes it failed to send as SMS. But after a few fails, Instinct was able to place the order relatively quickly. Well, sort of.

All told, this took 9 separate text messages on my end — not including Instinct's responses. It is hard for me to imagine that this was at all easier than just going onto the restaurant's website or UberEats and doing it myself.

Although the process was fairly painless, there was one thing I didn't like: it put the other through with no tip (!!!) and didn't ask me if I wanted to add one.

Dot: Somehow, this was the most difficult to get to work. It kept getting tripped up by the 6-digit code sext by text needed to login, but it kept failing (I had to get the code texted 3 separate times), and the browser it created for me to enter wouldn't let me type into it. A mess, basically, but finally it worked.

Winner: None. I feel very confident in saying that it would have been infinitely easier to just place the order myself through regular means. It's possible that it would be easier to order in the future, since I've already done the work of setting it up once. But I've learned my lesson.

Challenge #3: Build me a website

I have an old personal/professional website, but it's due for a refresh. I have terrible aesthetic sense and don't want to spend much time designing a site. Seems like a good job for AI to do for me!

Muse: I appreciated that, unlike Dot and Instinct, Muse didn't use the photo of my face from my byline photo (it seems very weird to have a giant photo of your own face on a professional website).

But all three basically did the same thing: pulled the info from my author bio on Business Insider, selected some recent articles as clips, pretty basic stuff.

Instinct: The design here was the most boring, and in a way I wouldn't even know how to make it look better (terrible aesthetic sense!). Weirdly, the selected article it pulled for me were random ones from over a year ago.

Dot: The look and design were probably the best here, and it pulled a decent mix of articles I'd want to showcase. But GOOD GRIEF the tagline here is so corny that I'm sort of embarrassed to even show you, even though I didn't write it myself:

Winner: Based on the mockups it made with design alone, I'd go with Dot. I think I could work with this. But all three of these made such cringey, cheesy copy about me that it honestly made me feel genuinely depressed. Is this who I am? Yuck!

Obviously, I'd have to rewrite all the copy myself, and pick the articles for clips — and at this point, is this even saving me any time beyond being a very generic template? Ah well.

Conclusion: The biggest challenge with all of them was figuring out what kind of task an agent is well-suited for.

Most of all, I found that trying to manage three different agents at the same time was a huge hassle. I want AI agents to be like a wife — making my life easier by anticipating my needs, not a husband who simply does chores when told. But this has backfired on me: now I have three wives, and they're all nagging me and I'm forgetting which is which. Ultimately, I don't recommend AI polygamy. May my experiment with it help you choose your bride.

• Barclays said it's cutting 5,000 jobs, or 5% of its workforce, in an AI push.
• Most of the cuts will impact support functions, including developers.
• Wall Street at large is grappling with how AI will transform both back- and front-office roles.

Barclays said it's cutting 5,000 jobs, or around 5% of its full-time workforce, as it increases its reliance on AI.

The British bank said in its third-quarter earnings statement that it is "structurally reducing" its costs through "technology-enabled" head count reductions.

Barclays laid out the numbers in its accompanying earnings presentation, linking the cuts specifically to "organisational simplification and technology change efficiency driven by AI."

Around 60% of the cuts will impact support functions, including developers, according to the presentation. Coding has become one of the main business use cases for AI across industries, causing some companies to cut back on developers.

Anna Cross, Barclays' group finance director, said the layoffs come as "AI makes our people more productive."

"Nowhere in the bank is this more relevant than in Barclays UK, where we have taken more structural cost actions than planned in '26," she said on the earnings call.

Wall Street is grappling with how AI will impact head count for both back- and front-office roles, as much of junior bankers' technical, repetitive workload can be automated. Some executives have said they're looking for young people with sharp judgment and communication skills rather than just analytical brainpower, as AI places an even greater premium on client interactions.

Goldman Sachs said in a memo last year that it plans to "restrain head count growth," and JPMorgan CEO Jamie Dimon has said he will likely hire fewer bankers and more "AI people."

Barclays joins a growing list of companies that have announced AI-related layoffs this year, including Snap and Coinbase. Business Insider analyzed 15 layoff memos earlier this year and found that "AI" was the most commonly used term.

AI hasn't always led to companies cutting back. Some companies have increased hiring, while others have cut some roles and hired more for others.

When Rishub Jain worked on safety at Google DeepMind, he held one of the most coveted jobs in tech. He was among a small group of researchers working to mitigate the most consequential risks of AI.

But he left to cofound a nonprofit, Sampura Research. Jain traded the prestige and access that come with a Google job because he feels he can have a sharper, more profound impact on AI safety from the outside. Jain says he wants the autonomy to pursue research on a specific issue in AI alignment, which he felt had been deprioritized.

"All the top AI companies do care a lot about safety, and they invest in it, but they could be investing a lot more, and that's kind of what was inhibiting me," Jain says. "If they invested more in safety, I would still be at the lab and working on what I'm interested in with a large team." Google did not respond to a request for comment about whether it maintains a focus on this specific area of research, but DeepMind's principles state that it maintains "a holistic approach to responsibility, safety and security."

Jain's founder story echoes concerns raised by safety researchers who have quit OpenAI and Anthropic, including dire warnings that these companies are "gambling with our lives," and could develop technology that ends humanity. Once dismissed as fringe and doomsday-obsessed, the AI safety profession is growing as hacks by rogue agents and a string of departures have vaulted the issue into mainstream concern.

Open roles in AI safety jumped by 91% on LinkedIn over the past year. The industry expects a flood of cash from much-anticipated IPOs at Anthropic and, potentially, OpenAI, as those made absurdly wealthy by the equity returns may funnel money into potential clean-up crews for the messes created by the technology's rapid development.

The AI safety problem is existentially big, but the people who have spent time honing AI expertise are few. These are people who test and train AI systems to operate ethically. Some focus on technical research, others on policy and governance. And they're increasingly sounding alarms.

"The people who are most worried about AI are often the people who have been closest to the technology," Laura González Salmerón, advising manager at 80,000 Hours, a nonprofit focused on transitioning people into careers with high social impact. The scarcity issue is less about funds and more about people. "Talent is the main bottleneck."

Startups and nonprofits in AI safety must fight for talent in a shallow pool.

Jain's Sampura lists a salary range of $273,000 to $395,000 for an experienced team leader, which would put that worker in the top 3% of earners in the US, but likely pales in comparison to offers from larger companies. Meta reportedly offered $100 million in signing bonuses last year to poach the world's top experts. OpenAI gives employees an average of $1.5 million in stock-based compensation, and technical staff at Anthropic can earn more than $1 million annually.

Money aside, researchers are divided over whether they're more effective when working inside tech companies to mitigate risks as the latest models develop, or stepping back and attempting to steer the ship toward safer outcomes from the outside, like Jain did.

"I'm a little bit concerned about the pace of AI, and if we're able to actually come up with something useful before the rug gets pulled beneath us," Jain says. He's worried that AI's advancement will outpace the work to keep it in line with human values. "That's kind of what's been happening already."

David Robinson, who led transparency work for OpenAI's safety team, wrote an essay for The Atlantic explaining his departure. He said that AI companies should have more layers of checks and safety measures, akin to those in airports and nuclear energy, rather than the steamrolling culture of Big Tech.

"Perhaps I should have stayed and fought for fundamental shifts in our staffing and culture, but in practice, my colleagues and I were so busy sprinting that we seldom had the chance to consider big changes, much less to actually make them," he wrote. Robinson has said he plans to work on these issues from outside of the company.

As the issue grows in prominence, so has a niche of organizations helping train and recruit workers — and the opportunities aren't just for AI experts. Neav Topaz, co-director of Kairos, a nonprofit that upskills people to work in AI safety and recently raised $50 million from Coefficient Giving, says he speaks with about 10 to 20 people each month who want to move to AI safety, and interest is growing. He tells me that people are asking him more frequently about what the work entails.

"People are watching AI get good at their own jobs and starting to ask themselves what comes next," he says. The emerging jobs in AI safety "pay well, the field is growing fast, and you get to work on what might be the defining problem of our time."

Kairos this summer launched its Generator Residency, a cohort of 27 people who spent three months training for career transitions into AI safety. It's less about the research and technical work and more about merging other expertise, like in consulting, with AI safety principles.

He says that of the 27 residents, a dozen have already landed roles in the AI safety space, and many of the rest have offers to do so.

"Most of the people who will do this work are doing something else today," Topaz says. "The biggest talent gap in the field isn't researchers. It's people who grasp how high the stakes are and can translate their varied expertise into things that could make a difference in this space."

The major AI companies maintain that safety is a top priority. Last month, Anthropic CEO Dario Amodei wrote an essay calling for a slowdown in AI development, and Sam Altman and Elon Musk have agreed with him. Amodei and Altman have both said recently that they will bring third-party evaluators into their companies to bolster safety protocols.

These companies have their own safety teams, but safety work competes with business priorities to build the most advanced models and win in the market. The tension, along with safety issues like OpenAI's rogue agents hacking Hugging Face, has led to more scrutiny of internal safety work. Within days of Robinson's departure, OpenAI fired three researchers, claiming they shared sensitive internal data improperly with third-party AI safety researchers. The ousted researchers wrote a letter, reported by The Wall Street Journal, to OpenAI's board, arguing that their engagement with safety researchers fell within the scope of their roles, and that their terminations were "chilling those who remain at OpenAI." OpenAI said that the decisions "were not about raising safety concerns or speaking out," and that the company would not fire people for doing so.

While some have obsessed over AI risks for years, 2026's breakthrough moment has accelerated funding for and interest in the industry. 80,000 Hours has had calls with some 8,000 people since its founding in 2011, the majority of which have focused on AI in some way, González Salmerón says. Philanthropic groups like Coefficient Giving offer massive grants to startups in the AI safety space. METR, a nonprofit AI safety group that investigated the OpenAI agentic hack on Hugging Face, says it raised $71 million in six months this year.

Much of the funding flowing into AI safety filters through effective altruism, the Silicon Valley-tied philanthropic ideal focused on using evidence and logic to donate money where it will achieve the most good. Critics say the philosophy can place too much emphasis on avoiding future problems rather than addressing imminent needs. With more alarms sounding around the existential threat of AI, more people view those theoretical, distant fears as urgent causes.

AI safety groups and individual donors have spent nearly $50 million in this year's election on candidates focused on regulating AI, Axios reported. Jaan Tallinn, a billionaire who invested early in Anthropic and could soon have even more billions as a result, has donated $170 million to nonprofits focused on AI safety. Tallinn's view, as he recently told The Wall Street Journal, is that he can have the biggest impact on the future of AI and its risks by investing in AI companies, and then donating returns to nonprofits working in the AI safety space.

Coefficient Giving, the philanthropic venture from Facebook cofounder Dustin Moskovitz rooted in the effective altruism movement, announced its AI-safety-focused arm, Project Tailwind, in September. It says it's on track to give more than $1 billion this year to help companies of all sizes scale up in preparation for what it believes will be a deluge of philanthropic capital. Sampura Research has received an $11 million grant from the group. If startups don't begin hiring now, the logic goes, they won't move quickly enough when massive wealth hits the industry.

The dizzying development of AI has created safety risks, and there are "gaps in the ecosystem that aren't being filled and that aren't being owned by people," says Rossa O'Keeffe-O'Donovan, the program director in technical AI safety at Coefficient Giving. There are few people, he says, who sit at the intersection of technical expertise, risk analysis, a solutions-oriented vision, and the capacity to hustle as a startup founder.

New roles will emerge not just for technical talent, but also for consultants, policy wonks, communications experts, and founders who can run a business alongside technical experts. Working from the outside, they'll face hurdles in buying compute power, a lack of direct access to the latest AI models, reduced leverage, and smaller paydays.

The world feels unmoored, and the future perhaps more malleable than ever. The appeal of doing work that does good could stave off feelings of uncertainty, dread, and anxiety that people increasingly feel toward AI. In one doomer camp, AI will obliterate white-collar jobs. The optimistic view is that AI drastically reshapes work and spurs the creation of jobs we haven't yet thought of. AI safety is one of the first.

"There are lots of people beyond AI researchers who can help, like cybersecurity experts, like biosecurity experts, like economists," says Mantas Mazeika, a researcher with the Center for AI Safety, a nonprofit focused on AI research and advocacy. "They're making very significant contributions that the pure AI researchers, who were most of the people focusing on the problem beforehand, would be completely unable to do on their own."

• The director behind Yung Lean's "Storm" video said AI copies made him consider quitting filmmaking.
• Romain Gavras said AI imitations of "Storm" on social media made his work look "lame."
• The original video, released in April, has amassed more than 40 million views on YouTube.

The director behind Swedish rapper Yung Lean's viral "Storm" video says AI almost drove him from filmmaking.

In an interview with Vulture published on Tuesday, Romain Gavras said he was disheartened by the wave of AI-generated copies of the music video.

Gavras directed the video for "Gener8ion," an audiovisual project he runs with French producer Surkin.

The video, which has amassed over 40 million views since its release in April, follows Lean as a bully at a boarding school before the boys he has been tormenting break into an elaborate dance sequence around him during a class photo.

In September, social media users began using AI to remake the dance scene from "Storm," swapping Lean and the schoolboys for a wide range of new characters, including animals and celebrities.

"My daughter sent me the first one I saw," Gavras said. "She was like, 'Look, this dog is doing it.'"

It was funny for a while, he said, but the novelty soon wore off as the copies seemed to spread everywhere.

"It made me sad for all the people who put in so much work on our video," Gavras said. "A trend like this can make even good work look lame. Part of me wanted to quit filmmaking, move back to Greece, and become an olive farmer."

Gavras said he wasn't trying to make any commentary on AI, but when people started calling "Storm" the kind of video AI couldn't pull off, he found himself drawn into the debate.

"The tech people were like, 'Oh yeah? AI could never?' Then they started replicating it. So we were forced into this conversation. It's a nightmare," Gavras said.

He added that he has turned down a few AI companies that have approached him about licensing the video.

"There's an economic model coming that we didn't want to be a part of," he said.

However, even as AI-made versions of his work proliferate online, Gavras said the technology can imitate elements of the video but has yet to create anything with the same originality or human spark.

"They're missing the swagger of Yung Lean. He looks like a dangerous baby. There's something in the imperfection of humanity, and the interior life of someone who dances or even someone who just smokes a cigarette, that's not quite there yet," he said.

Last month, Lean and the dancers performed the choreography at the MTV Video Music Awards. A video of the performance has since racked up nearly 14 million views.

Gavras is not the only filmmaker who has weighed in on the impact of AI on creative work. Other major Hollywood directors have raised similar concerns: Guillermo del Toro has said AI could rob future filmmakers of the chance to learn their craft, while Tim Burton has said it strips away the humanity behind art.

• Japan's aging population and rural exodus are fueling a small-business succession crisis.
• With no heirs, many aging business owners are listing their companies on succession platforms, often at low prices.
• Meet the young entrepreneurs betting on themselves to keep these businesses alive.

Six days a week, Kengo Ohama is up before the sun. He takes 5 minutes to get ready, and is soon out the door for an hourlong motorbike ride to the neighboring town of Mihara in Hiroshima, Japan.

By 6 a.m., he pulls up outside a log-cabin-style building, home to Country Road, the café and roastery that he's been managing with his wife since April.

He cleans the space, makes pastries, roasts coffee beans, and serves guests throughout the day. After the last table has cleared, Ohama stays behind to prepare for tomorrow. By the time he gets home, it is already 11 p.m.

Ohama had always dreamed of opening a café. When he found an established one listed on Relay, a Japanese business-succession platform, it felt like fate.

The café's owners, an older couple looking to retire, no longer had the physical stamina to run the business."Their biggest wish was simply for someone to take good care of the shop and keep it a lively place," Ohama, 31, told Business Insider.

The terms of the takeover were simple: He had to keep the business operating, but was free to redesign the interiors as he saw fit. "That flexibility made the opportunity very attractive," Ohama said.

Ohama now covers 50,000 Japanese yen in monthly rent for the building. The business itself — including the brand, furniture, kitchen equipment, recipes, and supplier information — was transferred to him for zero Japanese yen.

A new generation steps in

Ohama is part of a younger generation taking over Japan's small businesses as their longtime owners age out with no successor in sight.

The country's low birth rate and rapidly aging population have left many business owners without adult children to inherit their companies. Even when they do have kids, those heirs may live elsewhere, have established careers, or simply not want to take on the demands of running a local business.

In 2023, the Japan Finance Corporation found that 57.4% of small and midsize businesses it surveyed expected to close when their current owners stepped away, up from 52.6% in 2019.

That gap has created an opening for platforms such as Relay and BATONZ, which connect owners looking to hand over their businesses with prospective successors.

Relay has listed about 900 businesses to date, including roughly 350 active listings, a representative told Business Insider. Owners looking to pass on their businesses are often in their 60s and 70s, while successors are often in their 30s and 40s.

BATONZ, which also handles more conventional acquisitions by companies looking to expand, has listed more than 42,000 businesses and matched 3,421 buyer-seller pairs as of March 2026, a spokesperson told Business Insider.

Keeping a beloved place open

For some business owners, these platforms can offer a lifeline when an unexpected crisis forces them to confront succession planning sooner than they expected.

Akitaka Mitsui spent years working as an outdoor tour guide before opening Mori to Mizu Backpackers in Nagano, a city in central Japan known for its alpine landscapes, in 2011.

When the pandemic brought travel to a standstill, he temporarily closed the guesthouse. Even after it was safe to reopen, he had to scale back the operation and cut staff.

By 2022, visitor numbers were picking up again, and Mitsui was considering expanding once more. Then he suffered a stroke.

"It was not severe, but I had to be hospitalized," Mitsui, 67, told Business Insider. "The cause of the stroke was never identified, which meant it could happen again. I began to think it would be better to stop running the business."

He turned to BATONZ to find a successor outside his family.

"I never intended to make my children take over the business," he said. "I wanted them to do what they wanted to do with their own lives."

Enter Kota Fujisawa, a Nagano native who moved to Tokyo for university and later worked in consulting.

Fujisawa, 29, had been looking for a way to help revitalize his home region. Along with two childhood friends, he took over the guesthouse in May 2024, though one of them has since left the business. They renamed it Local Knot Backpackers.

He said they paid less than 10 million Japanese yen for the business, which came with its website, booking systems, and staff. The building was not part of the transfer, and he declined to disclose the monthly rent.

It is a drastic change from his stable office job, where he worked on just one part of larger projects. Now, as a small-business owner, Fujisawa has more freedom to shape the entire operation and decide where to take it next.

"My previous work was mainly desk-based, but I now spend much more time on-site and have more opportunities to communicate directly with a wide range of tourists," Fujisawa told Business Insider.

Still, there is a balancing act: Taking over an established business means deciding what to preserve and what to rework.

"If we changed everything, it would become a completely different place," he said.

Getting a foothold in a new industry

Beyond handing over entire businesses, succession platforms can also help aspiring entrepreneurs gain access to the infrastructure they need to get started.

For Takayuki Otawara, a 29-year-old former civil servant in Miyazaki, it opened a path into farming despite his lack of agricultural experience. His interest in cacao farming began during a stressful stretch at work, when he discovered craft chocolate and fell in love with it.

"I had never farmed before," Otawara told Business Insider. "The only thing I had ever grown was a small tomato plant for a primary-school project."

His parents initially opposed his decision to leave a stable government job for farming. His friends, though, were more encouraging. "They mostly thought it sounded interesting," he said.

To get started, Otawara needed a greenhouse and farmland, neither of which his family owned. After struggling to find one through local farmers, he turned to Relay and found a listing from a strawberry-farming family looking to scale back their operations as they got older.

Under a business succession agreement, Otawara bought a greenhouse and irrigation equipment from the original owners for 3 million Japanese yen.

He used the facilities to start his own farm, where he now grows cacao alongside strawberries, bananas, rice, and other crops.

He is at the farm every day, where his work begins at 6 a.m. and often ends around 7 p.m. He is usually in bed by 10 p.m.

The work is more demanding than his former office job, which came with weekends and public holidays off. However, farming has given him a kind of satisfaction he never found as a civil servant.

"The work I put in comes back to me directly as results," Otawara said. "In the government, you have a clearer idea of what your salary and career will look like until retirement. Now, how much I do and how much I earn can change. I find that exciting and rewarding."

• Rand Wacker, a Box product executive, wants his potential hires to go beyond being just AI-literate.
• Wacker likes when candidates show him their side projects, especially if they've monetized them.
• He's dipped his toes into building, too, with an Apple TV app that evokes the feel of old-school TV.

Knowing how to use AI is only enough to get a foot in the door for interviews, according to a Box executive.

Rand Wacker, Box's vice president of product strategy, said that for candidates applying for a job at Box, AI literacy is a bare-minimum requirement. Having a fully developed side hustle or project — something he practices himself — is what gives potential hires a leg up in interviews.

Case studies, which he said used to be a staple in Box interviews, are now a relic of the past. Now, he expects candidates to be prototyping out their ideas.

What he really likes to see is candidates showing him a product they've built themselves using AI, even if it has nothing to do with Box's business. He said he's had potential hires show him mobile apps that they've built, gotten onto the App Store, and are monetizing to some degree.

"They're running this little side hustle overall," he said. "And it shows the passion."

He said that people who explore and experiment with AI are the ones who win in the AI world, as they figure out how to use it most effectively in their personal and professional lives.

He practices what he preaches

Wacker, who's spent almost 30 years in the software engineering and product management fields, said he started out as a "computer nerd" and loved playing around with software tools and building things.

As he progressed in his career, he said he had less time to experiment with building, but AI has recently allowed him to test small ideas he didn't have time to explore before vibe coding existed.

He recently made an app to bring back old-school TV because he missed the days when he'd turn it on, pick one of three cable channels, and it would play a random movie from a random point.

"These days, whenever I'm trying to find something to watch on the weekends, it's decision fatigue. Which of the six streaming services do I go to? What are the 20 million titles I scroll through?" he said.

Using Cursor, he built an Apple TV app that would access his entire movie library, pick a random movie, and play it from a random point.

Working on these kinds of projects is a great way to learn the tools available on the market, get ideas for how to use them in other areas, and scratch the itch to create something of his own.

Wacker is one of many executives who have dipped their toes back into building and writing code in recent years. Leaders who haven't written lines of code in decades have started experimenting with AI coding to stay closer to their products and feel like individual contributors.

• Tesla's Full Self-Driving feature still requires drivers to pay attention to the road.
• During its Q3 earnings call, the company said unsupervised FSD for personal cars is coming in 2027.
• The catch? Your Tesla must remain within the company's Robotaxi geofences.

When I reviewed Tesla's Full Self-Driving software in January, I hailed it as an "impressive" driver-assistance feature that left me wanting more.

I dreamed of being able to flip the autonomous driving system on whenever I make my routine 400-mile trip from San Francisco to Los Angeles — kick my feet up, listen to music, and ideally fall asleep without having to supervise the road.

That's been part of Tesla's ultimate pitch for the future. "With autonomy, you get your time back," Elon Musk has said.

And the pitch may soon come true. Just don't expect to make long self-driven road trips quite yet.

On Wednesday, during Tesla's Q3 earnings call, Ashok Elluswamy, the company's VP of AI, painted a more modest picture of what FSD unsupervised will look like for a personal car — at least at first.

Elluswamy said unsupervised FSD is coming "somewhere around Q2 next year" for privately owned Teslas, with the rollout of FSD V15.

The caveat is that it will only be operable in areas where Tesla already operates its Robotaxi service, which represents a small fraction of square miles for the total number of Teslas on US roads.

The Robotaxi platform is available in parts of seven metropolitan regions: Austin, Dallas, Houston, Miami, Orlando, Tampa, and the San Francisco Bay Area. In the Bay Area, all Tesla robotaxis are supervised by a human operator.

"The only additional requirement is to make sure that, for example, the customer Model Ys can be part of the geofences and the fleet management that Tesla has for it to be fully autonomous," Elluswamy said.

In other words, your Tesla may be able to drive itself, but only if you're within the zones where Tesla Robotaxi is operating. If you're in Austin, that could cover about 260 square miles of the city, excluding high-speed roadways.

That's similar to what other aspiring self-driving car companies have pitched.

Tensor, a San Jose-based automaker, aims to sell a luxury sedan that it says can drive itself. However, the self-driving feature will only operate within restricted geofences, the company told Business Insider.

Tesla didn't say why it needs to limit the release of unsupervised FSD to geofences at first, although there are likely questions about liability that the company will have to address once it declares, Yes, our technology now lets you take your eyes off the road.

Autonomy is a difficult problem to solve after all. During the call, Elluswamy shut down the idea that competitors could catch up to Tesla's FSD.

"When they get in the long tail of this problem — it's so long that people don't realize — anyone with a sane mind should not attempt to solve this problem," he said. "It's so difficult."