From a tense corporate split to a viral photo: A timeline of Anthropic and OpenAI's budding rivalry

Sam Altman and Dario Amodei's hands did not make contact, and the internet noticed.

Ludovic MARIN / AFP via Getty Images

• Sam Altman and Dario Amodei went viral for refusing to join together in a sign of unity.
• Once colleagues, the two AI CEOs have helped nurse their companies' rivalry.
• Here's how OpenAI and Anthropic finally reached this point.

If you want to know one of the biggest rivalries in AI, just ask for a show of hands.

On Thursday, OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei provided what is sure to become an iconic image of their feud. The pair went viral for refusing to join hands as the rest of the world's tech leaders gathered for a moment of unity, sparked by Indian Prime Minister Narendra Modi.

Here's a timeline of how Altman and Amodei went from colleagues to becoming the face of AI's Cold War.

July 2015: Decisions happened over dinner

Elon Musk, Sam Altman, and the New York Times financial columnist Andrew Ross Sorkin speak onstage during "What Will They Think of Next? Talking About Innovation" at the Vanity Fair New Establishment Summit at Yerba Buena Center for the Arts on October 6, 2015, in San Francisco, California.

Michael Kovac/Getty Images

In July 2015, Tesla CEO Elon Musk, Altman, and a group of elite AI researchers all gathered at the swanky Rosewood Hotel in Menlo Park, California.

According to The New York Times, Musk had a falling out with then-Google CEO Larry Page. Weeks later, Musk, Altman, Amodei, Greg Brockman, Ilya Sutskever, and others discussed the formation of a new AI lab to ensure Google had a worthy competitor in the AI space. Musk invited Amodei, per tech journalist Alex Kantrowitz.

Their vision became OpenAI, though Amodei initially elected not to join the startup research lab. Roughly a year later, he changed his mind and joined OpenAI as "Team Lead for AI Safety."

September 2018: Amodei rises up

Anthropic CEO Dario Amodei

Markus Schreiber/File/AP

Amodei quickly moved up the ranks at OpenAI. In September 2018, the startup named him its research director.

Altman, who cofounded OpenAI while still serving as president of Y Combinator, began to devote more time to the startup.

In March 2019, Altman stepped down as YC's leader. He then became CEO of OpenAI and led the startup's pivot to a capped for-profit structure.

In November 2019, OpenAI released GPT-2, which Amodei played a major role in developing. A month later, OpenAI named Amodei as its Vice President of Research.

June 2020: OpenAI releases GPT-3

In June 2020, OpenAI began to show just how far the technology had come with the release of GPT-3, considered to be the first highly capable Large Language Model (LLM).

Amodei told The New York Times that the model had "this emergent quality." Independent researchers told the publication that GPT-3's capabilities surprised them, even as the model still showed signs of struggle.

To address safety concerns, OpenAI initially controlled access through a private beta.

December 2020: Amodei goes his own way

Anthropic CEO Dario Amodei at The World Economic Forum in Davos, Switzerland.

Fortune via Reuters Connect

The release of GPT-3 solidified OpenAI's standing, but behind the scenes, tensions were rising.

The rifts began when Amodei successfully lobbied to keep Greg Brockman, an OpenAI cofounder, off the team that developed GPT-3, according to Keach Hagey's biography of Altman, "The Optimist: Sam Altman, OpenAI, and the Race to Invent the Future."

Hagey wrote that Amodei's stunning power within OpenAI had started to ruffle feathers. Differences continued to escalate over Amodei's long-held views on safety, Hagey wrote, especially regarding slowing the pace of updates to prevent malicious uses of the AI models.

Amodei told friends that he "felt psychologically abused by Altman," Hagey wrote. Altman, in turn, was telling colleagues that the tension "was making him hate his job."

On December 29, 2020, OpenAI made it official. Amodei was leaving, and a "handful" of other colleagues were leaving.

Amodei has since suggested that his vision became incompatible with OpenAI's direction.

"It is incredibly unproductive to try and argue with someone else's vision," Amodei told podcaster Lex Friedman in 2024, when asked why he left OpenAI.

Early 2021: Anthropic is created

Jakub Porzycki/NurPhoto via Getty Images

With seven other former OpenAI employees, Amodei founded Anthropic in early 2021. The group was extremely close and included Daniela Amodei, Dario's sister. Daniela Amodei later said the name was chosen to emphasize their company's focus on humans.

Only one of Anthropic's initial employees hadn't worked at OpenAI, according to AI Business.

Despite starting from scratch, Amodei said that by the Summer of 2022, the company's chatbot, Claude, had finished training. Amodei said he was worried about what the release of a powerful AI could mean. Anthropic held off on a release.

"I suspect it was the right thing to do," Amodei told Time Magazine in 2024. "But it's not totally clear-cut."

Months later, OpenAI released ChatGPT, kicking off the AI race and making Amodei's former employer a household name.

May 2024: Amodei takes a shot … or did he?

As Anthropic began to establish itself in its own right, Amodei began to use his public appearances to take what were widely viewed as implicit shots at OpenAI.

During an appearance at a Bloomberg event, Amodei noted how Anthropic had kept its leadership intact.

"We have 7 cofounders," he said, Gizmodo reported. "Three and a half years later, we're all still at the company."

While never calling out by name, OpenAI was experiencing upheaval at the time. Months earlier, Andrej Karpathy, an OpenAI cofounder, had left the company. And in November 2023, Altman was briefly pushed out of OpenAI, an effort fellow cofounder Ilya Sutskever assisted. (Sutskever later expressed regret over his role. He formally left OpenAI just days after Amodei's jab, though there had been months of speculation surrounding Sutskever's standing.)

December 2025: 'We don't have to do any code reds'

OpenAI CEO Sam Altman

Justin Sullivan/Getty Images

By the end of 2025, OpenAI's lead in the AI race was slipping. When presented with the opportunity, Amodei seized the moment to troll his rival.

Journalist Andrew Ross Sorkin asked Amodei about OpenAI's decision to declare a "code red" to marshal resources for ChatGPT amid Google's rising strength.

We have a little bit of a privileged position where we can just keep growing and just keep developing our models, and we don't have to do any code reds," Amodei told Sorkin during an appearance at The New York Times' DealBook summit.

Earlier in their conversation, Amodei appeared to take another swipe at Altman when he talked about some players "who are YOLOing" by making too risky bets on future demand based on their current revenue.

"Who is YOLOing?" Sorkin asked.

"I'm not going to answer that," Amodei replied.

February 2025: A snarky Super Bowl ad gets a response

A still from Anthropic's ad is set to air during the Super Bowl. The ad features a scrawny man who wants to get a six-pack quickly, but a helpful trainer gives him more than just the advice he needs.

Anthropic

Anthropic used the biggest stage available to take its most direct shot yet at OpenAI.

Ahead of the Super Bowl, Anthropic revealed it was spending millions on an advertising campaign to denounce AI chatbot ads. While OpenAI was not named directly, it was clear who the intended target was, given that just months earlier, OpenAI had said it would begin testing ads on ChatGPT.

The ads featured real human actors mimicking the voice of product-pushing AI chatbots when asked questions like how to get a six-pack quickly or how to better connect with your mom.

"First, the good part of the Anthropic ads: they are funny, and I laughed," Altman wrote in a lengthy post on X." But I wonder why Anthropic would go for something so clearly dishonest."

Altman wasn't done.

"Anthropic serves an expensive product to rich people," he continued. "We are glad they do that and we are doing that too, but we also feel strongly that we need to bring AI to billions of people who can't pay for subscriptions."

Days later, the companies went head-to-head again. This time, with the release of major updates to their coding-focused model within minutes of each other.

February 2025: I (don't) want to hold your hand

In a moment before the viral photo, Altman holds hands with Indian Prime Minister Narendra Modi with his back facing toward Amodei

Ludovic Marin/AFP/Getty Images

A who's who of AI and tech elite gathered in India for a major summit on artificial intelligence.

Indian Prime Minister Narendra Modi took the opportunity to orchestrate a classic image of unity: competing CEOs with their hands raised together. (It's something Modi has done before with other world leaders, and politicians have been doing forever.)

Modi almost got his moment. While Altman held the prime minister's hand, the OpenAI CEO didn't grasp Amodei's hand, who was positioned to his other side. Amodei grasped the hand of the other person next to him, but not Altman's.

The internet, predictably, had a field day. And the world got a perfect encapsulation of one of AI's bitter rivalries.

Read the original article on Business Insider

from Business Insider https://ift.tt/qcoaFT9

Is taste a 'new core skill'? Techies debate — and quickly get memed

Discourse about "taste" is growing among techies on X.

AnnaStills/Getty Images

• Tech leaders from Y Combinator cofounder Paul Graham to OpenAI president Greg Brockman are posting about "taste."
• As life becomes automated with AI, they say, having good taste will become more important.
• Others on X have memed the discourse, pointing out some of Silicon Valley's bland taste in style.

Good news, connoisseurs — your judgment may be highly prized in the AI age.

A debate about the importance of "taste" has seemed inescapable on X in recent days, prompting bold declarations, a fair share of eye-rolling, and, yes, plenty of memes.

But there might be some truth to the idea.

The most recent wave of taste discourse began when Y Combinator cofounder Paul Graham — who coined "founder mode" — posted a prediction on Saturday.

"When anyone can make anything, the big differentiator is what you choose to make," Graham wrote on X.

This wasn't Graham's first time writing about the importance of taste; in his post over the weekend, he linked to an old essay of his from 2002, titled "Taste for Makers," and he also wrote about the topic again in 2021.

Then, OpenAI's president weighed in.

Will good taste get you a job?

OpenAI president Greg Brockman went even stronger, declaring, "Taste is a new core skill."

A variety of leaders across Silicon Valley also came out in support of the benefits of cultivating taste. As engineers grow into managing more agents and making more decisions, having strong judgment skills (or "taste," as they'd call it) could be crucial.

Cloudflare CTO Dane Knecht agreed with Graham and linked back to an earlier post he'd written in January that stated, "In 2026, taste is the engineering differentiator."

Engineers across the industry chimed in. "The AI will make anything but it takes human taste to decide if it's worth keeping," one Google software engineer wrote.

taste is a new core skill

— Greg Brockman (@gdb) February 16, 2026

What even is taste? The term is slippery.

"The taste thing works because it's nebulous, unassailable, and it feeds the ego," Poggio cofounder Matt Slotnick wrote on X, weighing in on the discourse.

Graham gave one glimpse, writing that it was about "being honest with yourself" and moving beyond the mindset of "I like what I like."

Not everyone agreed that taste would become a future-proof skill of the future. Linear head of product Nan Yu wrote that "you probably don't have better taste than AI."

"There are plenty of other distinctly human things that we can contribute, but 'having better taste' isn't one of them," he wrote.

Matt Schumer, the author of the viral "Something Big is Happening" essay on AI, agreed.

"There's a good chance AI will have better ideas than us within a few years," he wrote. "I don't see why "taste" and direction are uniquely human, like many people say. If an AI can train on it, it can learn it."

Taste memes are on the rise

There's another high-profile figure who championed taste (and basically built his career around it): Def Jam cofounder Rick Rubin.

There's an interview making the rounds of Rubin, in which he says he has "no technical ability" when it comes to music. Anderson Cooper asks what he's being paid for. Rubin responds: "The confidence that I have in my taste."

Is that what software engineers will soon look like? Several meme accounts on X seem to think so, commenting photos of Rubin's bearded face on Brockman's post.

Software engineers in 2030 be like pic.twitter.com/Vmfx8Qydzg

— prayingforexits 🏴‍☠️ (@mrexits) February 18, 2026

It also begs the question: Do engineers even have good taste to begin with? The answer depends on who you ask — and how you define taste.

General Catalyst creative director Reggie James wrote that techies would be "ruffled" when they learn they're not the top of the taste pyramid.

Stripe alum Sam Gerstenzan wrote that taste was "so rare" in Silicon Valley that, if asked to name the five people with the best taste, everyone would name the same people.

its almost like these tech guys understand that they dont have good taste so they are trying to turn it into this quantifiable set of data points instead of this mysterious, ineffable quality derived from going out and engaging with culture!

— Max Berlinger (@Berlinger) February 19, 2026

The irony, of course, is that tech bros aren't known to have the best taste in one specific arena: style.

The same people prizing taste are wearing backpacks to the bar, Very AI growth head Abril Zucchi wrote. Another user shared a "Person in tech that has 'taste' starter pack."

Person in tech that has “taste” starter pack https://t.co/jQw0PYESAM pic.twitter.com/4gPjDWAZOB

— VCs Congratulating Themselves 👏👏👏 (@VCBrags) December 22, 2024

The sneaker brand Allbirds was a popular punching bag.

"'Taste is the new core skill' says men who kept Allbirds afloat," Vercel CMO Keith Messick wrote.

“in the ai era, taste is the new core skill” pic.twitter.com/1vsvJZmnK8

— sophie (@netcapgirl) February 18, 2026

Graham responded to one of the memes, which showed "taste" as a polo and cargo shorts.

"Taste in clothing isn't important," he wrote. "If your goal is to think well, clothing should just be as comfortable as possible."

Rick Rubin — famous for his disheveled, DGAF appearance — would likely agree.

Read the original article on Business Insider

from Business Insider https://ift.tt/Q0AUR6P

A Japanese toilet maker and seasoning giant are unlikely winners of the AI boom

AI demand is boosting unexpected Japanese companies — including a toilet maker and a seasoning giant.

Smith Collection/Gado/Getty Images

• A toilet maker and seasoning giant are Japan's unlikely winners in the AI boom.
• Toto, famous for its bidets, has drawn investor attention because it makes key components for memory chips.
• Food giant Ajinimoto produces an insulating material used in advanced semiconductor packaging.

The AI boom isn't just lifting chipmakers and Big Tech. In Japan, it's flushing gains into a toilet manufacturer and a seasoning giant.

As demand for AI chips surges, investors are piling into companies that sit inside the semiconductor supply chain — even if they're better known for bathrooms and soup stock.

Toilet maker Toto, famous for its high-tech bidets and heated seats, has drawn investor attention. The company makes electrostatic chucks, which are critical components used in the production of NAND memory chips.

Memory prices have climbed sharply in recent months, driven by AI-related demand.

Last week, UK-based activist fund Palliser Capital called Toto "the most undervalued and overlooked AI memory beneficiary," according to reports by Bloomberg and the Financial Times.

After news broke on Tuesday that Palliser Capital had taken a stake and was pushing Toto to promote its chip-parts business, the toilet maker's stock jumped more than 5%. Its shares are up more than 54% over the past year.

It's not just Toto. Japanese food giant Ajinomoto, better known for its umami seasonings and soup bases, has become an unlikely AI infrastructure play. The company produces an insulating material used in advanced semiconductor packaging.

Ajinomoto's latest financials point to strength beyond its core food business. For the nine months ended December, the company reported an 8.9% rise in net profit, while operating profit increased 5.6% year-on-year. The gains were partly driven by its "Healthcare and Others" segment which includes electronic materials used in semiconductors, the company said in a February earnings statement.

After Ajinomoto posted its earnings on February 5, the company's stock rose 13%. Its shares are up more than 56% over the past year.

Not all non-tech companies are benefiting equally from the AI boom. Daikin, best known globally for its air conditioners, supplies high-purity chemical materials used in semiconductor manufacturing. It recently trimmed its outlook, citing uncertainty over US tariffs as a drag on demand.

The Japanese air conditioning maker reduced its operating profit forecast by about 5% to 413 billion Japanese yen, or $2.6 billion, for the fiscal year ending in March.

"Operating profit was significantly affected by the decline in semiconductor demand, decreasing by 44.6% year over year to ¥18,102 million," the company said in its financial report in February.

"Net sales of fluoropolymers fell year over year, despite focused Group efforts to capture strong new demand in the data center field, and was due to the stagnation in the construction markets of the United States and China and the significant overall impact of delays in the recovery of semiconductor demand," it added.

The company said it plans to cushion the blow through price increases and cost reductions.

Daikin's stock dropped as much as 8.4% in Tokyo following its financial results.

Read the original article on Business Insider

from Business Insider https://ift.tt/Rbl7JPa

Here's what smart people are saying about a software apocalypse

OpenAI CEO Sam Altman, Nvidia CEO Jensen Huang, and AWS CEO Matt Garman

AP and Getty Images

• Wall Street has cooled a bit after a massive sell-off wiped more than $1 trillion in Big Tech valuations.
• Recent AI advancements have undermined some investors' faith in established software names.
• Jensen Huang, OpenAI CEO Sam Altman, and Figma CEO Dylan Field have all weighed in.

Big Tech is continuing to bounce back after a brutal sell-off.

Wall Street's fears of AI-related disruption drove a sell-off of software stocks after the release of Anthropic's new industry-specific plug-in.

Not everyone in finance and tech is sold on the idea that AI is going to kill the software business.

From Nvidia's CEO dismissing the concerns, to Zoho's founder acknowledging the industry is "ripe for consolidation," here's what leaders in tech and finance are saying:

Jensen Huang

Nvidia CEO Jensen Huang

Steve Marcus/REUTERS

Nvidia CEO Jensen Huang said software is a tool for AI to use, rather than replace.

"There's this notion that the tool industry is in decline and will be replaced by AI," Huang said during a recent Cisco AI event. "You could tell because there's a whole bunch of software companies whose stock prices are under a lot of pressure because somehow AI is going to replace them. It is the most illogical thing in the world and time will prove itself."

Huang named ServiceNow, SAP, Cadence, and Synopsis, as bright spots in the industry.

Sam Altman

OpenAI CEO Sam Altman

Mandel Ngan/AFP/Getty Images

OpenAI CEO Sam Altman said volatility will remain in the software market.

"It's different, it's definitely not dead," Altman said during an interview on TBPN. "How you create it, how you're going to use it, how much you're going to have written for you each time you need it, versus how much you'll want sort of a consistent UX —yeah, that's all going to change."

In the meantime, sell-offs like the one Wall Street has continued now are likely to continue.

"I think it's just going to be volatile for a while as people figure out what this looks like."

Dylan Field

Dylan Field

Dylan Field, Kimberly White/Getty Images for TechCrunch

Figma CEO Dylan Field said it's not just about building something quickly, it's about "building the right thing."

"Good enough, it works, it's not enough," he told CNBC in February. "You really have to focus."

Overall, Field said volatility is good for companies. Shares of Figma have dropped by over 79% over the past year, illustrating the growing pains for the design company since its highly touted IPO in July 2025.

"I think if you look back on this time, we'll just be a more resilient company overall," Field said.

Figma also announced a partnership with Anthropic on a tool that converts AI-generated code into editable designs.

Sridhar Vembu

Sridhar Vembu, founder of Zoho, a cloud-based software company, said SaaS was "ripe for consolidation" long before the rise of AI.

"An industry that spends vastly more on sales and marketing than on engineering and product development was always vulnerable," he wrote on X. "The venture capital bubble and then the stock market bubble funded a fundamentally flawed, unsustainable model for too long. AI is the pin that is popping this inflated balloon."

Vembu said he asks his employees to consider the possibility of the company's death.

"When we accept that possibility, we become more fearless and that is when we can calmly chart our course."

Steven Sinofsky

Steven Sinofsky

Reuters

Steven Sinofsky, who helped lead the development of Windows 7 and 8, said AI may change "what we built and who builds it," but tales of software's demise are just "nonsense."

"Wall Street is filled with investors of all types. There's also a community, and they tend to run in herds. The past couple of weeks have definitely seen the herd collectively conclude that somehow software is dead. That the idea of a software pure play will just vanish into some language model," Sinofsky wrote in a lengthy post on X. "Nonsense."

Sinofsky said it is true some companies will fail. He also noted that such cycles have happened in retail and media.

"Strap in," he wrote. "This is the most exciting time for business and technology, ever."

Rene Haas

Arm CEO Rene Haas

Reuters

Arm CEO Rene Haas isn't panicking.

"As I look at enterprise AI deployment, we aren't anywhere close to where it can be," Haas told the Financial Times.

Haas, who leads the SoftBank-owned semiconductor company, said the current market reaction is "micro-hysteria."

Stephen Parker

JPMorgan analyst Stephen Parker said investors shouldn't be too worried by the sell-off.

"We're seeing a rotation," Parker told CNBC. "It's about a broadening of the recovery story. Cyclicals are picking up the slack, and it's not just the AI infrastructure plays and the hyperscalers that are driving markets higher."

Parker, the co-head of global investment strategy at JPMorgan Private Bank, said AI developments are likely to continue to cause disruption in the software industry.

Anish Acharya

Anish Acharya

Harry Murphy/Sportsfile for Collision via Getty Images

Anish Acharya, a general partner at A16z, said the sell-off was an overreaction based on a misunderstanding of how AI will be deployed.

"You have this innovation bazooka with these models," Acharya told podcaster Harry Stebbings during an episode of "20VC." "Why would you point it at rebuilding payroll or ERP or CRM, right? You're going to take it and use it to extend your core advantage as a business, or you're going to take it to optimize the other 90% that you're not spending on software today."

Acharya said there "will be secular losers," but overall, the sell-off was misguided.

"I think the general story that we're going to vibe code everything is flat wrong and the whole market is oversold software," he said.

Spenser Skates

Amplitude

Amplitude CEO Spenser Skates said the sell-off correctly identified that many SaaS companies are moving too slowly.

"The median SaaS company their innovation has actually slowed to a standstill," Skates told TBPN in February. "I don't know if you guys have ever been inside of these, but it's crazy how little they ship in terms of net new products."

Skates said AI has placed a major emphasis on the speed of innovation.

"It's like sushi," he said. "Buyers are always going to want the best thing. So, if you're keeping up with innovating the best thing, you will be able to charge a premium. It's fine that the 7-Eleven at the gas station now sells sushi. It's not going to put Jiro in Japan out of business."

Matt Garman

AWS CEO Matt Garman

Noah Berger/Getty Images for Amazon Web Services

AWS CEO Matt Garman said the current fears are "overblown."

"AI is absolutely a disruptive force that's going to change how software is consumed and how it's built," he told CNBC in February.

The top Amazon exec said current SaaS companies can still survive this moment.

"They have to innovate, just like the rest of the world," he said. "They can't stand still. If they stand still, they're absolutely going to be disrupted."

Read the original article on Business Insider

from Business Insider https://ift.tt/t8W71dL

AI's top leaders got corralled into holding hands. It made for a photo op for the ages.

Sam Altman and Dario Amodei's hands did not make contact, and the internet noticed.

Ludovic MARIN / AFP via Getty Images

• Sam Altman and Dario Amodei's awkward moment at the India AI Summit went viral.
• The two AI leaders — and former colleagues — raised their arms but did not hold hands.
• Tech leaders, including Sundar Pichai, gathered onstage with Narendra Modi in New Delhi.

The world's biggest AI leaders gathered in New Delhi this week, prepared to talk about the latest models and their impact on societies. They seemed less prepared for a 14-person hand-hold that tech circles will remember for a long time.

On Thursday, top executives, including Demis Hassabis, Sundar Pichai, Brad Smith, Sam Altman, and Dario Amodei, lined up on stage with Indian Prime Minister Narendra Modi at the India AI Impact Summit.

In his signature style, Modi held hands with Pichai on his right and Altman on his left and began raising their linked arms for a celebratory photo. Modi has previously taken photos this way with world leaders, including former US President Joe Biden and EU Commission President Ursula von der Leyen.

The other tech execs were quick to catch on to Modi's directive, looking right and left before grabbing their neighbour's hand.

The photo op's most meme-worthy scene was the OpenAI and Anthropic CEOs not managing — or refusing— to hold each other's hands. After a pause, they raised their arms without making contact.

The moment was widely screenshotted and shared on social media.

LOL at Sam and Dario not holding hands pic.twitter.com/dmmUYHGWkj

— Sheel Mohnot (@pitdesi) February 19, 2026

The awkward moment followed a Super Bowl advertising jab between the two AI giants earlier this month. Anthropic's 30-second commercial roasted OpenAI over its decision to bring ads to ChatGPT.

After Anthropic released a series of Super Bowl ad teasers, Altman responded with a lengthy post on X, calling the Anthropic ad "dishonest."

Amodei cofounded Anthropic in 2021 after leaving OpenAI, citing disagreements over AI safety priorities and the lab's leadership style.

Read the original article on Business Insider

from Business Insider https://ift.tt/VTMbPds

OpenAI, Meta, and Apple's latest battle: Breaking your phone addiction

DeepMind's CEO said there are still 3 areas where AGI systems can't match real intelligence

DeepMind's CEO said AGI still lags behind real intelligence in three areas.

LUDOVIC MARIN/POOL/AFP via Getty Images

• DeepMind's Demis Hassabis said artificial general intelligence efforts still leave a lot to be desired.
• Current systems cannot learn continuously, cannot plan long-term, and lack consistency, he said.
• He said last year that it would take five to 10 years for the world to see real AGI in play.

True artificial general intelligence is on the way, but it still has some ways to go, said Google DeepMind's CEO.

Speaking at an AI summit in New Delhi, Demis Hassabis was asked whether current AGI systems can match human intelligence. AGI is a hypothetical form of machine intelligence that can reason like people and solve problems using methods it was not trained in.

Hassabis' short answer: "I don't think we are there yet."

He listed three areas where current AGI systems are falling short. The first was what he called "continual learning," saying that the systems are frozen based on the training they received before implementation.

"What you'd like is for those systems to continually learn online from experience, to learn from the context they're in, maybe personalize to the situation and the tasks that you have for them," he said during the discussion.

Secondly, Hassabis said current systems struggle with long-term thinking.

"They can plan over the short term, but over the longer term, the way that we can plan over years, they don't really have that capability at the moment," he said.

And lastly, he said that the systems lack consistency. They're adept in some areas and unskilled in others.

"So, for example, today's systems can get gold medals in the international Math Olympiad, really hard problems, but sometimes can still make mistakes on elementary maths if you pose the question in a certain way," he said. "A true general intelligence system shouldn't have that kind of jaggedness."

Humans, in comparison, would not make mistakes on an easy math problem if they were math experts, he added.

Hassabis said in a "60 Minutes" interview last year that true AGI would arrive in five to 10 years.

The executive cofounded DeepMind, an AI research lab, in 2010. The lab was acquired by Google in 2014 and is the brains behind Google's Gemini. In 2024, Hassabis won a joint Nobel Prize in chemistry for his work on protein structure prediction.

AGI is a disputed topic in Silicon Valley. Databricks CEO Ali Ghodsi said at a September conference that current AI chatbots already meet the definition of AGI, but Silicon Valley leaders keep "moving the goalposts" and pushing toward superintelligence, or AI that can outthink humans.

The AI Summit in India, from Monday to Friday this week, has attracted big names from the tech and AI spheres. Notable speakers on the summit's agenda include OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Google CEO Sundar Pichai, and Meta's chief AI officer, Alexandr Wang.

Read the original article on Business Insider

from Business Insider https://ift.tt/cdbnrAP