• Venture capitalist Marc Andreessen said the US needs to open-source AI.
• Otherwise, the country risks ceding control to China, the longtime investor said.
• The stakes are high as AI is set to "intermediate" key institutions like education, law, and medicine, he said.

Venture capitalist Marc Andreessen has a clear warning: America needs to get serious about open-source AI or risk ceding control to China.

"Just close your eyes," the cofounder of VC firm Andreessen Horowitz said in an interview on tech show TBPN published on Saturday. "Imagine two states of the world: One in which the entire world runs on American open-source LLM, and the other is where the entire world, including the US, runs on all Chinese software."

Andreessen's comments come amid an intensifying US-China tech rivalry and a growing debate over open- and closed-source AI.

Open-source models are freely accessible, allowing anyone to study, modify, and build upon them. Closed-source models are tightly controlled by the companies that develop them. Chinese firms have largely favored the open-source route, while US tech giants have taken a more proprietary approach.

Last week, the US issued a warning against the use of US AI chips for Chinese models. It also issued new guidelines banning the use of Huawei's Ascend AI chips globally, citing national security concerns.

"These chips were likely developed or produced in violation of US export controls," the US Commerce Department's Bureau of Industry and Security said in a statement on its website.

As the hardware divide between the US and China deepens, attention is also on software and AI, where control over the underlying models is increasingly seen as a matter of technological sovereignty.

Andreessen said it's "plausible" and "entirely feasible" that open-source AI could become the global standard. Companies would need to "adjust to that if it happens," he said, adding that widespread access to "free" AI would be a "pretty magical result."

Still, for him, the debate isn't just about access. It's about values — and where control lies.

Andreessen said he believes it's important that there's an American open-source champion or a Western open-source large language model.

A country that builds its own models also shapes the values, assumptions, and messaging embedded in them.

"Open weights is great, but the open weights, they're baked, right?" he said. "The training is in the weights, and you can't really undo that."

For Andreessen, the stakes are high. AI is going to "intermediate" key institutions like the courts, schools, and medical systems, which is why it's "really critical," he said.

Andreessen's firm, Andreessen Horowitz, backs Sam Altman's OpenAI and Elon Musk's xAI, among other AI companies. The VC did not respond to a request for comment from Business Insider.

Open source vs closed source

China has been charging ahead in the open-source AI race.

While US firms focused on building powerful models locked behind paywalls and enterprise licenses, Chinese companies have been giving some of theirs away.

In January, Chinese AI startup DeepSeek released R1, a large language model that rivals ChatGPT's o1 but at a fraction of the cost, the company said.

The open-sourced model raised questions about the billions spent training closed models in the US. Andreessen earlier called it "AI's Sputnik moment."

Major players like OpenAI — long criticized for its closed approach — have started to shift course.

"I personally think we have been on the wrong side of history here and need to figure out a different open source strategy," Altman said in February.

In March, OpenAI announced that it was preparing to roll out its first open-weight language model with advanced reasoning capabilities since releasing GPT-2 in 2019.

In a letter to employees earlier this month announcing that the company's nonprofit would stay in control, Altman said: "We want to open source very capable models."

The AI race is also increasingly defined by questions of national sovereignty.

Nvidia's CEO, Jensen Huang, said last year at the World Government Summit in Dubai that every country should have its own AI systems.

Huang said countries should ensure they own the production of their intelligence and the data produced and work toward building "sovereign AI."

"It codifies your culture, your society's intelligence, your common sense, your history — you own your own data," he added.

• TikTokers are posting clips of interviews with glitching AI bots.
• These cases are rare and likely staged, professors told Business Insider.
• AI interviews are on the rise, and glitches can erode trust in the hiring process.

TikTok videos of glitchy AI interviews have gone viral in recent weeks, but don't worry if you're due for a job interview soon.

One user, who goes by Freddie, posted a video on May 3 of an AI assistant named "Catherine Appleton" glitching and spewing gibberish during his job interview. As of Thursday, his video had 8.8 million views.

"Should I email them? I was expecting a real human," he wrote in the caption.

Another TikTok user named Ken shared a clip of her interview, in which the AI assistant repeated the phrase "vertical bar pilates" on loop.

Neither responded to requests for comment from Business Insider.

Your next job interview probably won't involve a glitching AI bot

Yes, the viral TikToks are creepy. But they're probably not your future.

"The TikTok videos showcasing glitches or malfunctions are likely either doctored or represent rare, isolated incidents," said Sriram Iyer, an adjunct senior lecturer at the National University of Singapore Business School.

They "should not be considered a common phenomenon," he added.

Tan Hong Ming, the deputy head and senior lecturer in the department of analytics and operations at NUS Business School, said social media "tends to amplify things."

"It can make something appear far more common than it actually is through repetition and viral sharing," he said.

Tan, who also serves as lead advisor to a Singapore-based AI recruitment firm, said the looping audio is "likely dramatized or re-enacted to drive engagement and shares." He said he has not come across this specific glitch in AI interviews, but occasional breakdowns aren't surprising.

Many companies are using AI-powered recruitment tools which are often "wrappers around the same core models or APIs."

Some of them may not use the latest or most stable versions, which could explain why similar glitches show up across platforms, he said.

Unaizah Obaidellah, a senior lecturer specializing in AI at Malaysia's University of Malaya, said insufficient or irrelevant data could also be a culprit. If the bots are not trained with enough relevant examples, their quality suffers.

She added that the incidents portrayed on the videos could reflect the larger race to deploy AI faster than we're ready for, which is "quite worrying."

AI interviews on the rise

Emily DeJeu, an assistant professor at Carnegie Mellon University's Tepper School of Business who specializes in AI communication and etiquette, told BI earlier this week that AI-powered video interviews are likely to become more common as companies seek to streamline and automate early hiring stages.

Any time technology promises to save time and money and make everything faster, "we by default pursue it — there's a kind of inevitability to it," she said.

Despite what the TikToks might suggest, candidates aren't necessarily turned off by bots, said Iyer, who has worked in HR tech for 20 years.

What to do if your interview bot glitches

Glitches during AI interviews aren't just awkward.

"Glitches chip away at trust and can make the hiring process feel impersonal or even unfair," said Tan, especially if companies are not upfront about conducting an AI interview.

"They undermine the candidate's experience," he said, adding that employers need to "build in strong fallback options" and monitor these tools closely in real-world settings.

"Otherwise, what feels like a time-saving solution could quietly become a systemic problem," he added.

For candidates, the key is not to panic.

If an AI bot malfunctions mid-interview, Tan recommends emailing the hiring manager with a screenshot or recording of what happened.

"Most should offer a redo assuming the candidate isn't already put off by the idea of being interviewed by a bot in the first place," he said.

Unaizah, from the University of Malaya, said candidates can also request feedback from the HR team on their interview performance.

If there's clear evidence the interview wasn't properly assessed — or wasn't reviewed by a human — ask for an in-person interview, if possible, she said.

"If all fails or your gut feeling says otherwise, perhaps it's best to look for other companies," said Unaizah. "Target companies that prioritize human-centered hiring."

• Claude generated "an inaccurate title and incorrect authors" in a legal citation, per a court filing.
• The AI was used to help draft a citation in an expert report for Anthropic's copyright lawsuit.
• Anthropic's lawyer called it "an embarrassing and unintentional mistake."

A lawyer defending Anthropic had to clean up after the company's AI bot, calling it "an embarrassing and unintentional mistake."

In a copyright lawsuit over Anthropic's use of music lyrics, the company's legal team used its AI assistant, Claude, to help draft a citation in an expert report.

Claude provided the correct publication title, publication year, and link to the provided source, but "an inaccurate title and incorrect authors," Anthropic's lawyer said in a court filing on Thursday.

Attorney Ivana Dukanovic, of Latham & Watkins, said her team's "manual citation check" failed to catch the mistake and "additional wording errors introduced in the citations during the formatting process using Claude.ai."

"This was an honest citation mistake and not a fabrication of authority," Dukanovic wrote.

Music publishers Universal Music Group, Concord, and ABKCO sued Anthropic, saying the company used copyrighted lyrics to train Claude. The case is part of a wave of legal battles between copyright holders and AI companies.

The publishers' attorney told the court on Tuesday that Anthropic data scientist Olivia Chen may have used a fake source generated by AI to support the company's argument, Reuters reported.

On Thursday, Dukanovic responded that Chen cited a real article from the journal "The American Statistician," but Claude had made up the title and authors.

Anthropic and Dukanovic did not respond to a request for comment.

AI in the legal world

It's not the first time an AI tool has raised eyebrows in the legal world.

In March, a lawyerless man deployed an AI-generated avatar to argue his civil appeals case in a New York courtroom. A panel of stunned judges quickly shot him down.

AI hallucinations have also landed lawyers in hot water. An attorney was fired from Baker Law Group after he used ChatGPT to generate legal citations, which turned out to be fake.

Donald Trump's former fixer, Michael Cohen, also got into trouble when he used Google's AI chatbot, Bard, to find legal cases to support his arguments. The chatbot made up the cases, and his lawyer filed them in court without checking.

Daniel Shin, the assistant director of research at the Center for Legal and Court Technology at Virginia's William & Mary Law School, told Business Insider in a report last month that judges are concerned about the use of AI in the courts because of hallucinations.

Courts have shown they will not tolerate any improper use of AI tools, Shin said.

Still, lawyers are being told they need to start adopting AI.

At a legal-tech conference in March, lawyers were urged to embrace AI or risk falling behind, BI's Melia Russell reported.

"Lawyers need to wake up," Todd Itami, an attorney at the large legal defense firm Covington & Burling, said, adding that learning to use artificial intelligence was "imperative" for their success.

• Immad Akhund passed on Scale AI, now worth $14 billion, because the founders were "so young," he said.
• The miss taught him to rethink his bias against young founders.
• The CEO and founder of Mercury has backed more than 350 startups at their earliest stages.

Immad Akhund, an active angel investor since 2016, passed on an early bet in Scale AI — worth nearly $14 billion in last year's funding round — and it taught him an important lesson.

"I saw Scale AI. And I was like, 'good idea, but these people are so young,'" he said on an episode of the "Twenty Minute VC" podcast published Monday.

"I think they were like 19 and 20 at the time or something. I think I could run this company better if I was doing it and I didn't see how they were going to figure it out," he added.

Looking back, the serial entrepreneur said he was "just so wrong."

"There's some power to that youth that's hard to judge, to be honest. You kind of have to suspend belief and say, okay, this person's going to figure out how to run a huge company," he said.

The misstep was a formative one for Akhund, who has backed more than 350 startups at their earliest stages, including Rappi, Airtable, Rippling, Decagon, and Etched.

Akhund angel invests in "things that will seem inevitable 10 years from now and can be $10 billion companies," he told Business Insider in a May story about the most successful seed-stage investors.

Akhund is also the founder and CEO of Mercury, a banking startup that recently raised a $300 million Series C round at a $3.5 billion valuation led by Sequoia.

Leaving his ego at the door

Akhund said one of the biggest lessons in his investing journey was learning to check his ego.

As an entrepreneur, he was used to suggesting his ideas when speaking with startups, he said.

The founders, especially if they were young, would often concur. But "that's not their idea, and it's not even fair to push an idea on other people," he said.

He learned that it was important to back founders for their ideas, not for his.

"You really have to actually remove your ego and your ideas and really listen," he said. "You're much more along for their journey rather than a major part of it," he added.

Akhund said that he prefers serial founders to first-timers, especially those with "a chip on their shoulder."

"I have such a bias toward them," he said. "A serial entrepreneur knows how hard it's going to be, but they're willing to do it again."

That's "so unusual" and signals that "they must really want to do this," he added.

For investors, Akhund advises making diversified bets.

"Do at least 20 or 30 investments — that's when you start entering the game," he said.

"You learn a lot by doing subsequent ones," he said, adding that investors "need a diversified portfolio to have any return in this space."

A representative for Akhund declined further comment.

• Some startups are using the label "AI agents" to raise prices, said a partner at Andreessen Horowitz.
• The venture capital firm funds Sam Altman's OpenAI and Elon Musk's xAI.
• "There's a marketing angle to agents," said Guido Appenzeller, a partner at a16z.

"AI agent" may be the buzziest term in tech, but some startups are using the label to raise prices, said a partner at Andreessen Horowitz, one of Silicon Valley's most prominent venture capital firms.

"There's a marketing angle to agents," said Guido Appenzeller, a partner at a16z, on a company podcast released May 2. "A couple of startups are basically saying, 'Hey, we can price this software that we're building much, much higher because this is an agent.'"

Appenzeller said that a "continuum" of startups is cashing in on the hype by branding simple tools as AI agents.

"The simplest thing that I've heard being called an agent is basically just a clever prompt on top of some kind of knowledge base," he said. A user asks about a technical problem, and the "agent" looks at the knowledge base and comes back with a "canned response."

An agent for some of these startups "could just be a large language model with a chat interface," he added.

Appenzeller and partners Matt Bornstein and Yoko Li were discussing how to define AI agents — systems that are supposed to act autonomously to complete tasks.

Andreessen Horowitz is a backer of Sam Altman's OpenAI and Elon Musk's xAI. It is seeking to raise a $20 billion fund to invest more heavily in AI, sources told Reuters last month. The company did not respond to Reuters' requests for comment.

"I don't think anything we have are actually agents," said Bornstein. The term itself "may be poorly defined and kind of overloaded," he added.

Companies have been using AI agents to automate elaborate, multi-step tasks. For instance, Regie AI uses "auto-pilot sales agents" to automatically find leads, draft personalized emails, and follow up with buyers. Big Four professional services firm PwC unveiled "agent OS," a platform that makes it easier for agents to communicate with one another to execute tasks.

But the more steps an agent takes to complete a task, the more likely it is to make errors, Business Insider reported last month.

Despite the ambiguity, AI companies are betting big on agents to drive returns.

The Information reported in March that OpenAI plans to sell Ph.D.-level agents starting at $20,000 a month and eventually expects 20% to 25% of its revenue to come from agents.

"If 2024 was the year of LLMs, we believe 2025 will be the year of agentic AI," Praveen Akkiraju, a managing director at Insight Partners, previously told BI.

Bornstein, however, cautioned against letting hype lead strategy.

"Let's look at the actual technology underlying what we're calling agents," he said. "Where are they being deployed, and why?"

• Magnus Grimeland, the CEO and founder of the VC firm Antler, said demand for software engineers will only grow.
• AI will continue to make errors, and only software engineers will optimize this technology.
• AI will also lead to further specialization among software engineers, he said.

This as-told-to essay is based on a conversation with Magnus Grimeland, the CEO and founder of Antler, a global early-stage venture capital firm. He also cofounded Zalora, a fashion e-commerce platform in Asia. This interview has been edited for length and clarity.

There have been a lot of headlines about software engineering being replaced by AI, based on the assumption that anyone can just go in and code any program with natural language. It's actually much more likely that the need and demand for great software engineers will grow in the next couple of decades.

Even the best software engineers today make errors. AI models will also continue to make errors, at least for a very long time, and the only ones who will optimize this technology are software engineers.

At least over the next 20 to 30 years, what you will see is the best software engineers getting a tremendous amount of leverage to be more efficient and deliver better products faster. Software engineers will work in a different way than before.

In the not-too-distant future, we also need to adapt to an entirely new computer ecosystem, and the ones who are going to be able to do that are software engineers. We've already started investing in a few companies that are preparing for that.

Further specialization

AI will also lead to further specialization.

Today, software engineers are grouped a bit more generally. Some work on hardware, some on different types of software languages, and some are great mobile developers.

The complexity of the type of roles that you'll see for software engineers will increase significantly because the way this is being implemented in different industries will require specialized goals.

You'll also see fewer general engineers and more people who are really good at one specific thing.

Software engineers will work closer with businesses. AI will enable business leaders to work better with engineering departments because they can tinker with the early versions of the products themselves.

This should lead to more efficiency in terms of how the technical and less technical parts of the business work together, and that should actually give software engineering an even more important role in the business.

A new era of learning

When we were building Zalora and now at Antler, some of the best engineers we hired in Southeast Asia were self-taught.

They didn't have computer science degrees from universities. They read up on the internet, tinkered, and built their own programs.

AI has made it better than ever to teach people — as long as they have the right drive and basic intrinsics to learn how to become a great software engineer.

You'll see many more self-help people who are just as good as people who've done a full university degree.

• Today, Anastasia Soare is the CEO of the beauty giant Anastasia Beverly Hills.
• When she was starting out, she had to beg Wells Fargo for a $500 credit card.
• Soare, an immigrant from Romania, shared her journey of pursuing the American dream at the Milken conference on Wednesday.

Before she became the CEO of a global beauty brand with celebrity clients, Anastasia Soare needed a $500 credit card.

When she arrived in the US in 1989 as an immigrant, "nobody paid attention," the founder of Anastasia Beverly Hills said. At the time, she was an aesthetician with zero credit history and limited English — and a bold idea that needed backing.

"I went to Wells Fargo and had to beg the manager to give me a $500 credit card," she recalled. "Thirty-four years ago, they didn't throw at you with credit cards like what they are doing right now."

Soare made the comments on a panel at the 28th Milken Institute Global Conference on Wednesday. The conference, held in California this year, draws global leaders, tech executives, and experts to discuss finance, innovation, and global markets.

In 2000, Soare launched her first product line to bring her signature shaping techniques and tools to anyone who wanted beautiful brows. Since then, Anastasia Beverly Hills has become synonymous with eyebrows in the beauty world. The brand's celebrity clients include the likes of Jennifer Lopez, Kendall Jenner, and Kim Kardashian. In May 2024, Forbes estimated Soare's net worth at $900 million.

Soare shared the credit card anecdote as part of her pursuit of the American dream — one that started decades earlier in communist Romania.

At 15, her grandfather would take her into a small room, shut the windows and curtains, and tune into a banned radio station called "Voice of America."

"He used to tell me that I did a big mistake coming to Romania. The communist regime took everything away from us," she said.

"You should go to America," Soare recounted her grandfather saying. "That is the American dream."

Under the communist regime in Romania, entrepreneurship didn't exist. "You worked only for the government, and that's it. You couldn't be an entrepreneur," Soares said.

But Soare knew she wanted more. "I wanted first to show myself what I'm capable of doing," she said.

Chasing the American dream

When Soare arrived, she started working as an aesthetician — one of the few jobs that didn't require "perfect English," she said.

She noticed a gap in the beauty market — nobody paid attention to eyebrows. Believing in the potential of eyebrows as a business idea, she began doing clients' brows for free, as it wasn't "considered a service" then.

"My husband thought I am totally crazy," she said. "You don't know how to write a check, you don't have a credit card, you barely speak the language, and you want to open a business. Not even American people born here own a business," she recalled her husband saying.

But Soare went all in. "What do I have to lose?" she thought.

From that little room, she built a clientele. Then she opened a salon in Beverly Hills. "The rest is history," she said.

"I am proud to say that I invented a category in the beauty industry that didn't exist before," she said at the panel.

• Nvidia's CEO, Jensen Huang, said he uses AI "as a tutor every day."
• Huang was speaking at a panel at the 28th annual Milken Institute Global Conference on Tuesday.
• AI is the "greatest opportunity for us to close the technology divide," he said.

Nvidia's CEO, Jensen Huang, isn't just building the chips that power advanced AI models — he says he's also using those models as his personal tutor.

"I use it as a tutor every day," Huang said of tools like ChatGPT, Gemini Pro, and Perplexity. "In areas that are fairly new to me, I might say, 'Start by explaining it to me like I'm a 12-year-old,' and then work your way up into a doctorate-level over time."

Huang was speaking at a panel at the 28th annual Milken Institute Global Conference on Tuesday. The mega-investment conference held in California this year draws global leaders, tech executives, and experts to discuss the future of finance, innovation, and global markets.

That kind of access, Huang says, is what makes this generation of technology a potential equalizer.

"Artificial intelligence is the greatest opportunity for us to close the technology divide," he said.

"In this room, it's very unlikely that more than a handful of people know how to program with C++," he said. C++ is a specialized coding language used for building software.

"Yet 100% of you know how to program an AI, and the reason for that is because the AI will speak whatever language you wanted to speak," he added.

Huang said everybody should take advantage of AI's capabilities and "every student should use it as a tutor."

"Don't be that person who ignores this technology and this result loses you," he added.

The AI revolution will also touch every job, Huang said.

Some roles will be lost, others created — but all will be transformed. "You're not going to lose your job to AI," Huang said. "You're going to lose your job to someone who uses AI."

Huang has long been a proponent of using AI and co-pilots to automate tasks in the workforce. He previously said that he wanted Nvidia to be a company with "100 million AI assistants."

Huang said in October at an AI summit in India that AI will do parts of some jobs 1,000 times better — but it will never replace the actual humans doing those roles.

Nvidia produces some of the most sought-after chips for AI models. In June, on the back of the AI wave, Nvidia hit a $3 trillion market cap for the first time. Its value has since dipped back down to $2.8 trillion.

• Anysphere CEO Michael Truell said he took his time while hiring.
• "Many people you hear hired too fast. I think we actually hired too slow," he said.
• He also said he relied too heavily on conventional markers of talent like brand-name schools.

Michael Truell said he made several early hiring mistakes at Anysphere.

The CEO and cofounder of the startup behind Cursor, an AI-powered coding tool, said the team initially fussed over hiring in the hopes of assembling a high-performing team.

"We tried to be incredibly patient on the hiring front," Truell said in an episode of Lenny's Podcast that aired Thursday.

The goal was to build a world-class group of engineers and researchers — "a certain mix of intellectual curiosity and experimentation," he added.

But looking back, Truell said they may have been too patient. "Many people you hear hired too fast. I think we actually hired too slow," he added.

The San Francisco-based company is one of the leaders in vibe coding, or the practice of using AI to generate code quickly and build software. It has closed a new round of funding that more than triples its valuation to about $9 billion, the Financial Times reported on Sunday. OpenAI backer Thrive Capital led the round, with Andreessen Horowitz and Accel among the other investors, per the FT.

The startup has been doubling its valuation every eight weeks since August, per PitchBook.

Truell partnered with MIT classmates Sualeh Asif, Arvid Lunnemark, and Aman Sanger to start building Anysphere in 2022, according to TechCrunch. Cursor is Anysphere's first product.

Truell said another one of his early hiring missteps was relying too heavily on conventional markers of talent, such as brand-name schools.

The hiring team leaned toward people who fit the archetype of "well-known school, very young, had done the things that were high credential," Truell said.

But the hires who stood out — and stuck around — didn't always fit that mold. Some were later in their careers or looked "slightly different from being straight out of central casting," he added.

Truell said his early hiring lessons were "hard-won." It took time to refine what "greatness" looked like, how to spot it, and how to persuade someone to come on board.

"We kind of spent a bunch of time on the wrong profile," he said. "We were lucky early on to find fantastic people willing to do this with us, who were later-careered."

• Gavin Newsom said he's ready to team up with Trump to revitalize America's film industry.
• "Now it's time for a real federal partnership to Make America Film Again," Newsom wrote on X.
• Newsom's remarks come after Trump said he was considering imposing a 100% tariff on movies made outside the US.

Gov. Gavin Newsom of California said he is ready to team up with President Donald Trump to rejuvenate America's film industry.

"Now it's time for a real federal partnership to Make America Film Again. @POTUS, let's get it done," Newsom wrote in an X post on Monday night.

Newsom also said California is "ready to bring even more jobs home," adding that the state "built the film industry."

When approached for comment, a spokesperson for Newsom said the state is "eager to partner with the Trump administration to further strengthen domestic production and Make America Film Again."

"America continues to be a film powerhouse, and California is all in to bring more production here," Newsom's spokesperson told Business Insider.

Newsom's remarks come a day after Trump said he was considering imposing a 100% tariff on movies made outside the US.

In a Truth Social post published on Sunday, Trump said the US film industry is "DYING a very fast death" while other countries were using incentives to lure American filmmakers to shoot their movies outside the US.

"Hollywood, and many other areas within the USA, are being devastated. This is a concerted effort by other Nations and, therefore, a National Security threat," Trump wrote in his post.

On Monday, actor Jon Voight and his manager, Steve Paul, said they presented Trump with a "comprehensive plan" to "make Hollywood great again."

Voight and Paul's proposal includes "federal tax incentives, significant changes to several tax codes, the establishment of co-production treaties with foreign countries, and infrastructure subsidies for theater owners, film and television production companies, and post-production companies."

The pair also called for a "focus on job training" and imposing "tariffs in certain limited circumstances." They added that the White House is reviewing the proposal.

Newsom and Trump have had a strained relationship.

Last month, Newsom filed a lawsuit in San Francisco federal court challenging Trump's tariffs, calling them "unlawful and unprecedented."

When deadly wildfires ravaged Los Angeles in January, Trump attributed blame in part to Newsom.

"The fires in Los Angeles may go down, in dollar amount, as the worst in the History of our Country," Trump wrote in a Truth Social post in January.

"Let this serve, and be emblematic, of the gross incompetence and mismanagement of the Biden/Newscum Duo," Trump added.

Newsom's term as governor ends in about two years, and he is widely seen as a frontrunner for the 2028 Democratic presidential nomination.

Unlike most presidential hopefuls, Newsom has sought to engage his political opponents directly.

In February, Newsom started his podcast, "This is Gavin Newsom," which has featured guests like Charlie Kirk, a right-wing influencer, and Steve Bannon, Trump's former chief strategist.

The White House did not respond to a request for comment from BI.

• Jon Voight said he has presented Donald Trump with his plan to "make Hollywood great again."
• Trump appointed Voight, Mel Gibson, and Sylvester Stallone as his special ambassadors to Hollywood.
• Voight's proposal called for "tariffs in certain limited circumstances."

Actor Jon Voight said on Monday that he has presented President Donald Trump with a plan to revitalize the US film industry.

"The President loves the entertainment business and this country, and he will help us make Hollywood great again," Voight said in a statement.

Voight and his manager, Steve Paul, said in a press release that their proposal includes rolling out federal tax incentives, changing the tax code, inking co-productive treaties with foreign countries, and handing out infrastructure subsidies to cinema owners as well as production and post-production companies.

Voight's proposal also called for a "focus on job training" and the imposition of "tariffs in certain limited circumstances." The statement said the White House is reviewing the proposal. The White House did not respond to a request for comment from Business Insider.

"We look forward to working with the administration, the unions, studios, and streamers to help form a plan to keep our industry healthy and bring more productions back to America," Voight said.

Voight's remarks come just a day after Trump said he was considering imposing a 100% tariff on movies made outside the US.

Trump wrote in a Truth Social post on Sunday that the US film industry is "DYING a very fast death" while other countries were doling out incentives to lure American filmmakers to produce their work there.

"This is a concerted effort by other Nations and, therefore, a National Security threat. It is, in addition to everything else, messaging and propaganda!" Trump wrote.

In a post on X on Monday night, California's governor, Gavin Newsom, said the state is "ready to bring even more jobs home," adding that California "built the film industry."

"We've proven what strong state incentives can do. Now it's time for a real federal partnership to Make America Film Again. @POTUS, let's get it done," he wrote.

In October, Newsom announced a proposal to increase California's Film & Television Tax Credit Program from $330 million to $750 million yearly.

This proposal aims to incentivize production houses to keep their work in California instead of moving elsewhere.

This isn't the first time Trump has talked about boosting the US film industry.

In January, before his second inauguration, Trump announced that he was appointing Voight, as well as actors Mel Gibson and Sylvester Stallone, as his ambassadors to Hollywood.

"They will serve as Special Envoys to me for the purpose of bringing Hollywood, which has lost much business over the last four years to Foreign Countries, BACK—BIGGER, BETTER, AND STRONGER THAN EVER BEFORE!" Trump wrote.

"These three very talented people will be my eyes and ears, and I will get done what they suggest. It will again be, like The United States of America itself, The Golden Age of Hollywood!" he continued.

Trump did not specify how or when his tariff would be implemented. On Monday, he appeared to soften his stance when he told reporters that he would ask film studios if "they are happy" with his proposed tariff before rolling it out.

"So we are going to meet with the industry. I want to make sure they are happy with it because we are all about jobs," Trump said.

• The role of software developers might be broadened into "builders," said Windsurf's CEO.
• AI is about to "democratize" software creation, said Varun Mohan.
• Engineering might start to look more like a research-driven culture.

The days of the traditional software developer are likely numbered, says Windsurf's CEO, Varun Mohan.

"This notion of just a developer is probably going to broaden out to what's called a builder," Mohan said on a podcast episode of "Y Combinator" that aired on Friday.

Traditional developers may no longer be the only ones building software, he said: "I think everyone is going to be a builder."

Windsurf, founded in 2021 as Codeium, offers users an AI-powered coding development tool and has been riding the vibe coding wave. According to PitchBook, it has raised $243 million in VC funding. In April, Bloomberg reported that Windsurf is in talks to be acquired by OpenAI for around $3 billion.

Before cofounding Windsurf, Mohan was a tech lead manager at Nuro, an AI robotics company. He also had professional experience as a software engineer.

On the podcast, Mohan said AI is about to "democratize" software creation. Instead of downloading an app, people might simply ask their AI assistant to build a custom tool tailored to their needs — one they can keep tweaking over time.

"I can imagine a future like that where effectively everyone is building but people don't know what they're building is software," he said.

Mohan also said vibe coding is going to get more capable. AI is supercharging every stage of the software development process — writing, reviewing, testing, debugging, and designing code, he said.

"AI is going to be adding 10 times the amount of leverage very shortly. It's going to happen much more quickly than people imagine," he said.

Vibe coding, a term coined in February by the OpenAI cofounder Andrej Karpathy, refers to the process of feeding prompts to AI in order to write code. As Karpathy puts it, developers can "fully give in to the vibes" and "forget the code even exists."

The rise of vibe coding has shaken up the way people think about software development. It has left some engineers wondering if AI could put them out of a job and sparked debate among investors over whether technical skills are still a must-have for startup founders.

Don't hire for boilerplate coding

If AI can take over repetitive tasks like boilerplate coding, developers will be freed up to focus on what really matters — testing bold ideas, said Mohan.

Engineering starts to look more like a research-driven culture, one where they are testing hypotheses, evaluating them, and getting user feedback. Those are things that make a product significantly better, he said.

This also changes what startups should look for when hiring engineers, Mohan added.

"For engineers that we hire, we want to look for people with really high agency that are willing to be wrong and bold," he added.

Startups should never be hiring engineers to "quickly write boilerplate code," he said. "A startup can succeed even if they have an extremely kind of ugly code," he added.

"The reason why a startup fails is that they didn't build a product that was differentially good for their users," he said.

• Longtime technologist John Fitzpatrick said he has seen companies exaggerating their AI capabilities.
• Some businesses rebranded as AI, despite few real changes.
• Regulated industries need to be especially watchful for hallucination, privacy, and other issues.

This as-told-to essay is based on a conversation with John Fitzpatrick, the chief technology officer at document management startup Nitro. It has been edited for length and clarity.

I have worked in AI for 15 years and am one of the original engineers behind Apple's Siri. I'm currently the chief technology officer at Nitro — a software company that helps businesses manage and secure documents more efficiently.

Over the last year, I've seen a lot of AI washing, especially after ChatGPT took off.

AI washing is when companies exaggerate or misrepresent what their AI can actually do, just so they can say they're using AI.

Suddenly, tons of apps that were just a new skin slapped on top of ChatGPT popped up. Businesses started rebranding their existing automation features as AI without making any real product enhancements.

I see this as similar to the "cloud" hype many years ago. Suddenly, every business became a cloud business. We're seeing that with AI today. If you listen to earnings calls, every company's talking about AI.

Recent AlphaSense data shows a 779% increase in mentions of terms like "agentic AI," "AI workforce," "digital labor," and "AI agents" during earnings calls in the past year.

Almost every single startup now has to have an AI angle to secure funding.

Telltale signs of AI washing

There are a few different examples of AI washing.

One example is thin user interface layers on top of ChatGPT and maybe a small amount of prompt engineering. In some cases, that can be really valuable, but in many cases, it doesn't add any particular value.

The other challenge with AI washing is companies rushing AI features to the market with these simple integrations without considering customer privacy or security.

In the worst cases, major players launch assistant features and update their terms and conditions to allow them to use customer data for training.

Then there's the problem of relying on third-party public APIs and services vendors don't control. This would mean sensitive documents would be sent to third parties, which is a major security risk.

In regulated industries, where they often deal with extremely important documents, you want to be very careful of hallucination and ensure you're getting things like confidence scores from the models.

Many of our customers have invoices and financial data in PDF documents. It's really important to be accurate when that data is extracted, or where the model has low confidence, to be very obvious.

Companies are also trying to do full automation flows without having that human check, and that's where mistakes can happen.

In these types of industries, those mistakes could be really costly,

We're moving beyond the hype and into the adoption part — where AI becomes an implementation detail of building really powerful product features.

Companies are learning what AI can and cannot do, building good features, and leveraging AI.

Because of that, investors and the market are starting to understand where there's superficial AI versus actually adding real utility.

• McDonald's and Coca-Cola have pushed back on talk of returning to the Russian market.
• Their responses follow speculation in Russian media that they were preparing a return to the country.
• Russian officials said the country is prioritizing domestic companies over returning Western firms.

McDonald's and Coca-Cola, two of the biggest companies that left Russia after it invaded Ukraine in 2022, have pushed back on talk of returning to the Russian market.

In March, McDonald's and Coca-Cola separately told the Ukrainian advocacy group B4UKraine Coalition they do not intend to reenter Russia. B4UKraine Coalition shared both companies' responses on its website and wrote on Monday about McDonald's correspondence.

In McDonald's response, as shared on the Ukrainian site, the company said it is sticking to the principles that guided its decision to exit Russia in 2022. At the time, the company evaluated five questions: whether it's legal to operate in the country, whether it has the freedom to run its business, whether its presence is helpful to its brand, whether staying aligns with its values, and whether it makes good business sense.

In 2022, "the answer to each of those questions was no," wrote Jon Banner, the fast food giant's global chief impact officer.

"The principles behind these questions, which guided our decision to 'de-Arch' a major market and sell our portfolio of McDonald's restaurants, still apply today," wrote Banner.

In an email reply to the Ukrainian group in late March, Coca-Cola wrote that "sanction regimes and other legal hurdles" stand in the way of any return to the Russian market.

Both companies' responses came after B4UKraine Coalition asked various companies about their Russian market plans. The requests came as Washington and Moscow's warming relationship sparked speculation in Russian media that Western businesses might reenter the country.

McDonald's and Coca-Cola did not respond to requests for comment from Business Insider.

'A price to pay for past decisions'

Three years into the war in Ukraine, nearly 475 foreign companies have left the Russian market completely, per the Leave Russia database from the Kyiv School of Economics. Those that have made a complete exit include McDonald's, Starbucks, Ikea, the British energy giant Shell, and the Japanese tire maker Bridgestone.

In February, Russia said it is not in a hurry to receive Western companies that had left the country.

"We are not waiting for anyone with open arms. There will be a price to pay for past decisions," Anton Alikhanov, the Russian industry and trade minister, told reporters, according to the state news agency TASS.

Alikhanov also said Russia was prioritizing domestic brands instead of waiting for foreign brands to return.

Analysts have said that businesses are likely to be cautious about returning to the country even if sanctions are lifted.

Russia's wartime economy is facing multiple issues, including high inflation, currency volatility, and sky-high interest rates.

• OpenAI said on Monday it rolled out new shopping feature for ChatGPT.
• The chatbot now curates product recommendations based on your preferences.
• I gave it a try — it was efficient, but not quite personalized.

OpenAI is jumping into the online shopping game.

The company said on Monday it rolled out a new shopping feature for ChatGPT. Users can ask for product recommendations and receive results with images, reviews, and direct links.

ChatGPT will provide a curated list of products based on what it remembers about users' preferences and what it pulls from online reviews, said Adam Fry, who leads ChatGPT's search product team, in a prelaunch demo for Wired. Users are then sent to the retailer's site to check out.

The feature works across categories like fashion, electronics, and home goods.

Unlike Google Shopping's sponsored links, the product listings shown in ChatGPT are not paid placements, Fry said.

I tried out the new shopping feature to see if it could make shopping smarter and easier. It was helpful, but I missed a few parts of non-AI online shopping.

Getting ChatGPT's techy recommendation

As an audiophile, I love a good pair of headphones.

I asked ChatGPT, "What are the best headphones under $500 USD?"

The response was impressive — and a bit overwhelming. It broke down options into categories like best wireless noise-canceling headphones, best audiophile and studio headphones, and budget-friendly picks.

It even gave me a handy "Top Picks Summary," which helped cut through the clutter. The Sony WH-1000XM5 clinched the "Best Overall" — a model I'd heard people rave about.

But ChatGPT didn't list where I could buy them, so I followed up with, "Where can I buy them at the best price?"

This time, ChatGPT served up live prices and retailer links in Singapore, where I live.

ChatGPT also suggested a rival to the Sony XM5 — the Bose QuietComfort Ultra.

I wanted to compare both in detail, so I asked: "What's the difference between the Sony WH-1000XM5 and the Bose QuietComfort Ultra?"

ChatGPT came back with a side-by-side comparison that felt lifted from a high-end tech blog, covering everything from noise cancellation to fast-charging stats.

One gripe, though: I had no idea where the information came from. It said Bose had "plush cushioning and lighter clamp force," while Sony offered "synthetic leather" and a lighter build. Those were objective details, but who decided Bose was more comfortable? What if I actually like the snug fit of Sony's headband?

Buying a really personal item: running shoes

Next, I put ChatGPT to the test with running shoes. For me, this was make-or-break.

I'm obsessed with running shoes. I've gone too deep into the rabbit hole of reading shoe reviews, binge-watching YouTube breakdowns, and analyzing specs like a full-time job.

I told ChatGPT I was looking for shoes for easy, short-distance runs under $200.

My search yielded a solid list: Nike Air Zoom Pegasus 41, Brooks Ghost 16, Hoka Clifton 9, ASICS Gel-Nimbus 26, and New Balance Fresh Foam X 1080v14.

All great picks, but I was a little disappointed that my favourites, like the Adidas Ultraboost, didn't show up.

When I refined my search and mentioned that I was a neutral pronator — my foot rolls naturally to absorb shock evenly — ChatGPT stuck to the same lineup with short, summarized notes about each shoe's strengths.

I asked ChatGPT to break down the differences between the Brooks Ghost 16 and Hoka Clifton 9 since they were "Best Overall" and "Best Cushioning."

The side-by-side comparison covered everything from cushioning and stability to ride feel.

But one big thing was missing: visuals.

As someone who cares a lot about how my shoes look — the sole design, the lacing, the vibe — I missed scrutinizing images beyond the four at the top.

I don't blame ChatGPT for giving me a wordy review. But for something as personal as running shoes, I'd still rather scroll through a store's website and rummage through every photo and video.

Google vs ChatGPT

Google's shopping experience still feels a lot more fun. Although there are paid ads, there are also endless visuals — different colorways, angles, and even photos from real buyers showing off their pairs in the wild.

Plus, I love seeing user reviews. They're raw, relatable, and sometimes way more helpful than expert reviews.

ChatGPT wins in the heavy-lifting department. It provides detailed comparisons and spec breakdowns without making you dig through 10 different websites.

It feels like an extension of what ChatGPT already does best — gathering info and summarizing it neatly. An AI built solely for shopping — or OpenAI's next version of this feature — might provide a better experience.

I'm still an old-school shopper. I gather all possible sneaker info online, then head to the store to do a few laps.

No review can tell me the best — for me — pair until it's on my feet.

• DeepSeek is hiring for a job in product management and design.
• It's a major shift from the startup's focus on AI model research.
• The rush to hire product talent mirrors a broader trend in the US.

DeepSeek, the Chinese startup that rattled the AI industry earlier this year, is hiring for a product role that illustrates the company's shift from research to commercialization.

In a job notice posted Tuesday on its official WeChat account, DeepSeek said it is looking to fill a "product and design" position on its teams in Beijing and Hangzhou. It is unclear from the notice if the job refers to a single role or multiple positions.

The Hangzhou-based firm labeled the job notice "urgent." The company wrote that it wants people to help create the "next generation of intelligent product experience" centered on large language models.

Candidates are expected to have product management experience and be proficient in product and visual design, the notice said.

DeepSeek did not respond to Business Insider's request for comment.

DeepSeek is also hiring a chief financial officer and chief operating officer — jobs not labeled urgent. The company is expanding its research and engineering teams, according to other listings on its WeChat account.

The move marks a major shift for the company, which has been focused on fundamental AI model research. Last month, DeepSeek released an upgraded version of its open-source V3 large language model, boosting its reasoning and coding capabilities.

Founded in 2023 by Chinese entrepreneur Liang Wenfeng, DeepSeek made headlines and disrupted markets in January after unveiling its low-cost reasoning model, R1. The startup claims R1 can rival top competitors like OpenAI's GPT-4 — but at a fraction of the cost.

An analyst told Business Insider earlier this month that DeepSeek's latest models — especially the reasoning-focused R1 and R2 set to launch later this month or in May — mark a "significant inflection point" in China's AI ambitions.

"These models not only match the best-in-class performance globally, but are also open-sourced under the most permissive MIT License," said Wei Sun, the principal analyst for AI at Counterpoint Research.

"That changes the game," she added.

Unlike flagship models in the US, which are typically closed-sourced and monetized through APIs or enterprise licensing, DeepSeek's models like R1 and V3 are free for anyone to download, modify, and integrate.

DeepSeek has been quiet about the progress of its next-generation R2 model.

Amid high costs and chip shortages, Chinese firms have been prioritizing AI integration and consolidation to stay competitive, an analyst told BI earlier this month.

Tencent has deployed its Hunyuan model and DeepSeek R1 across its massive ecosystem, including WeChat, said Ray Wang, a Washington-based analyst who specializes in AI and US-China tech statecraft. WeChat, China's biggest social media app, is used by nearly 1.4 billion people.

Baidu has also integrated DeepSeek R1 into its search engine, he said.

While details about DeepSeek's hiring process are scant, Liang, the founder, has made it clear that he values creativity over experience when it comes to hiring.

In a 2023 interview with Chinese tech publication 36KR, he said that "experience is not that important," even in a similar role. "Basic skills, creativity, and passion are much more important," he added.

"Our core technical positions are mainly filled by fresh graduates or those who have graduated one or two years ago," he said.

Rise of product managers

The rush to hire product talent mirrors a broader trend in the AI world.

In the US, product managers are seen as increasingly critical for some companies in the AI era, helping bridge the gap between rapidly advancing AI technology and real-world user needs.

"The future really does belong to product managers," Frank Fusco, a product manager turned CEO of a software company called Silicon Society, told BI in November.

As AI becomes more capable of handling coding and other engineering tasks, Fusco said it's an opportunity for product managers to take on an even greater role.

OpenAI is hiring seven product manager roles in its New York and San Francisco offices, and Anthropic is hiring 11 product-related roles, according to the companies' websites.

However, some tech companies are revisiting their reliance on product managers. Microsoft wants to increase the number of engineers relative to product or program managers, BI's Ashley Stewart reported last month. Other companies like Airbnb and Snap have been rethinking the need for product managers.

• AI could become the way people interact with software, said Meta's chief technology officer.
• That would turn the app model on its head — a nightmare for some companies.
• Andrew Bosworth said this shift is "net positive."

AI could replace apps as the main way people use technology, said Meta's chief technology officer.

Right now, people pick out software from "a garden of applications," Andrew Bosworth said on a podcast episode published Thursday by venture capital firm Andreessen Horowitz. The firm was an early Facebook investor.

Instead of opening a specific app, like Spotify, to listen to music, Bosworth said he'd rather just tell an AI what he wants and have it handle the rest. Spotify did not respond to a request for comment from Business Insider.

"I don't want to be responsible for orchestrating what app I'm opening to do a thing," he said. "We've had to do that because that's how things were done in the entire history of digital computing."

This shift, Bosworth said, could turn the app model on its head — and that may be great for users, but a nightmare for some companies.

"It abstracts away a lot of companies' brand names, which I think is going to be very hard for an entire generation of brands," he said.

Brands "want me to have an attachment. I don't want to have an attachment," he added.

That could upend the ways apps like Spotify and Netflix monetize their services, whether through ad revenue, subscriptions, or "freemium" models that offer a free core product with paid tiers for premium features.

Software services 'squeezed' by AI

The rise of reasoning models and AI agents is beginning to erode the core assumptions that have defined the software-as-a-service business model for decades, Business Insider reported on Monday.

A study released by consulting firm AlixPartners said this was affecting more than 100 midmarket software companies.

AlixPartners said these companies are caught in a "big squeeze," pressured on one side by nimble, AI-native entrants that can replicate applications at a fraction of the cost and on the other by tech behemoths, such as Microsoft and Salesforce, that are pouring billions of dollars into the AI arms race.

"We believe many midsize enterprise software companies will face threats to their survival over the next 24 months," the firm added. It declined to identify specific companies.

Ultimately, Bosworth thinks this shift from a smattering of apps to a powerful AI interface is worth it.

"That's net positive because what matters now is performance on the job and price per performance," Bosworth said. "A lot of companies won't love that," he added.

• Cluely is an AI tool built for cheating in live conversations, and it has raised $5.3 million.
• The startup is founded by Chungin "Roy" Lee, who was suspended from Columbia.
• I put the app through a mock interview to see if it could help me land a job.

An Ivy Leaguer just released an AI app to feed live answers to users. I put it to the test to see if AI could interview as well as I did.

Chungin "Roy" Lee — the Columbia student who went viral for creating an AI tool to "cheat" on job interviews — was suspended in March for posting content from a disciplinary hearing, the university said.

His new app, Cluely, helps users by analyzing what's on their screens, hearing audio, and suggesting answers to questions, all without detection from the other side. In a LinkedIn post, Lee said Cluely is a "cheating tool for literally everything."

Lee said on Monday that San Francisco-based Cluely has raised $5.3 million, backed by Abstract Ventures and Susa Ventures.

Speaking to Business Insider on Wednesday night, Lee said that Cluely has gained about 70,000 users since its launch on Sunday.

"It is pretty crazy. And this is just a lot more than we expected," he said.

Lee said the concept of "cheating" needs to be rethought in the AI era.

"Using AI is just inevitable and something that we should just all embrace," he said.

I put Cluely to the test, and I saw promise — but also huge privacy concerns.

First impressions

I tested Cluely on Tuesday morning and paid $20 for the Pro version — a monthly subscription — since the free tier doesn't offer much.

The interface is slick and minimal: a thin bar on the top of my screen. With one keyboard shortcut, Cluely scans my screen, then generates answers. I could also speak to it through my mic.

It only works with what's already on my screen — it doesn't crawl the internet like ChatGPT.

It also reads everything. I kept my Signal group chats closed.

That's what makes it powerful — and also kind of terrifying.

Lee told BI that Cluely doesn't save any data, and if it does, it's "heavily censored."

"If there's something that you don't want to be seen, then you shouldn't use the tool, and it shouldn't be up while you're on sensitive information," Lee said.

I deleted it right after writing this story.

Flopping my mock interview

Cluely's killer pitch is real-time assistance.

I let it scan my LinkedIn profile and kept it open while my editor put me through a mock interview with the same questions I answered to land this job.

First up: What is my media consumption like? Could I share what I like to read or watch? What keeps me interested and busy?

The app starts "thinking." Five seconds later, it gave me a summary of the question, not an answer.

My editor tried again, more succinctly: "What media outlets do you read?"

This time, Cluely came through:

"I enjoy staying up to date with a mix of international and local news — typically from outlets like the New York Times, BBC, and CNA. I love longform journalism from The Atlantic and Wired, especially on tech and social issues. I also listen to 'The Daily' and 'Planet Money' during my commute."

Not bad, though I don't listen to podcasts on my way to work.

My editor asked me to talk about a difficult story I've worked on. 10 seconds of awkward silence later, Cluely offered: "One difficult story I worked on was about the impact of layoffs in the tech industry."

I never wrote that story.

The second hallucination came when the app said I had a "working knowledge of Malay." My editor congratulated me on my surprising — including to me — third language skill. It completely missed the elementary Korean listed on my LinkedIn.

When my editor asked if I had questions for her, Cluely suggested a few basics: What do you enjoy most about working here? What's the team culture like? What does success look like in this role?

Not worth the $20 — yet.

Cluely's biggest flaw is speed. A five to 10-second delay feels like forever in a live interview.

The answers were also too generic, occasionally wrong, and not tailored enough to me.

It did generate decent answers to common questions. When I read them aloud, my editor said the biggest clue I had help was the delay, not the substance. She also said my real answers were better than Cluely's.

Lee told BI that Cluely is in "a really raw state."

"Our servers are super overloaded, so there's a lot of latency," he said.

But there have been "significant performance updates" since the app went out on Sunday, he added.

"We've upgraded all our servers, we've optimized the algorithms, and right now it should be about three times faster, which makes it much more usable in conversations."

Lee said hallucinations will "exist insofar as the base models that we use allow for them."

"The day that the models get better is the day that our product will get better," he added.

There's definitely potential. If Cluely got faster, smarter, and could pull info from beyond just my screen, it could become a game-changing AI assistant. If I were hiring, I might think twice about conducting remote interviews because of these sorts of apps.

But between the privacy risks, laggy performance, and random hallucinations, I'm keeping it off my computer.

• An OpenAI exec said the company would consider acquiring Chrome if Google is forced to sell it.
• The DOJ had asked a judge to force Google to divest Chrome as part of its antitrust case.
• Integrating ChatGPT with Chrome "could offer a really incredible experience," said ChatGPT's product head.

OpenAI would be interested in buying Google's Chrome browser if antitrust regulators force the tech giant to sell it, a company executive said in testimony during Google's antitrust trial in Washington on Tuesday.

Nick Turley, OpenAI's head of product for ChatGPT, was asked whether the company would consider acquiring Chrome.

"Yes, we would, as would many other parties," Turley said, Bloomberg reported.

In the landmark antitrust case against Google, the Department of Justice is seeking remedies to address what it says is the company's illegal monopoly in online search and advertising markets.

Last year, the DOJ asked the judge in the case to force the company to sell its Chrome browser. Judge Amit Mehta ruled in August that Google had violated antitrust laws. Google has not offered Chrome for sale and it plans to appeal the ruling.

Chrome is a hugely popular Google product that the company relies on to grow and maintain its search-advertising empire. Mehta's August ruling cited data from StatCounter, which analyses web traffic, that said Chrome held 61% of the US browser market share, while 20% of general search queries came through user-downloaded Chrome browsers.

Rebecca Haw Allensworth, a law professor at Tennessee's Vanderbilt University, told Business Insider it's a "toss up" on whether Mehta would force Google to divest Chrome. The DOJ is leaning on competitors like OpenAI to strengthen its case.

According to Bloomberg, Turley said an integration between ChatGPT and Chrome "could offer a really incredible experience." OpenAI would "have the ability to introduce users into what an AI first experience looks like," he added.

Turley also said that dominant tech players like Google could lock OpenAI out of key distribution channels, Bloomberg reported.

OpenAI has powerful competitors "who control the access points for how people discover products, including our product. People discover via a browser or an app store," Turley said. "Real choice drives competition. Users should be able to pick."

Industry experts told Business Insider last year that a forced sale of Chrome would be a major blow to Google, but a potential win for competitors. Advertisers and search rivals would likely cheer the news.

Ben Thompson, a tech blogger, wrote last year that it was "unlikely" regulators would allow Meta to acquire Chrome, leaving companies like OpenAI as potential buyers. The "distribution Chrome brings would certainly be welcome, and perhaps Chrome could help bootstrap OpenAI's inevitable advertising business," he wrote.

• Windsurf's CEO wants to keep his company lean and light.
• The goal is to be the smallest company possible to meet its ambitions, Varun Mohan said.
• Windsurf builds AI tools that help developers write code.

Windsurf's CEO and cofounder, Varun Mohan, wants to keep his startup lean — and dry.

"I want the company to almost be like this dehydrated entity," said Mohan on an episode of "Lenny's Podcast" published Sunday. "Every hire is like a little bit of water, and we only go back and hire someone when we're back to being dehydrated," he added.

The goal isn't to idolize small teams for the sake of it, he said. It's to "be the smallest company we can be to satisfy our ambitions."

Windsurf, formerly known as Codeium, builds AI tools that let developers write code using natural language prompts.

It's part of a new wave of startups leaning into "vibe coding," a term coined by OpenAI cofounder Andrej Karpathy to describe giving AI prompts to write code. As Karpathy puts it, developers can "fully give in to the vibes" and "forget the code even exists."

Founded in 2021, Silicon Valley-based Windsurf has raised more than $200 million in venture capital funding, according to PitchBook data. The company was valued at $1.25 billion in a deal led by General Catalyst last year, with backing from Greenoaks and Kleiner Perkins.

Revenue per employee has become an important metric for investors, especially after many tech companies grew rapidly during the pandemic. Maintaining smaller teams is often preferred if it leads to the same output level.

"If we can crack actually being a fairly sizable company but still operate as if we're a startup," Mohan said, "that's the dream."

Windsurf has 170 employees, per data provider PitchBook, and its website lists more than 30 open roles on its website, including software engineers and recruiters.

Lean teams, less drama

Mohan also said keeping head count low isn't just a financial decision — it's a way to avoid unnecessary problems.

Hiring for teams where there are already enough people often leads to "weird politics," he said.

When there's no real need for their role, people may manufacture something to work on.

"Realistically, it's not that important, but they'll go out and try to convince the rest of the organization that it is," he said.

That kind of distraction, he said, can slow a company down more than it helps. "As a startup, we don't have the bandwidth to go out and deal with that," he added.

Hiring should only happen when "everyone's just almost raising their hands and being like, 'I'm dying, we need one more person,'" he said.

Windsurf did not respond to a request for comment from Business Insider.