Tuesday, May 28, 2024
HomeAIGoogle’s Demis Hassabis is the man tasked with turning bleeding-edge AI research...

Google’s Demis Hassabis is the man tasked with turning bleeding-edge AI research into profits

Demis Hassabis is well-known for his contributions to artificial intelligence research. He is a chess master and a neurologist. On Wall Street, he is lesser known.

That may not hold true for long. Hassabis, the face of Google’s AI initiative, will speak for the first time at the annual developer conference, Google I/O, on Tuesday.

Hassabis, a scholar recognized with some of the most significant advances in AI over the last decade, is quite clear about his goal ahead: deliver the latest AI technology to every corner of the Google universe to serve its billions of users.

“We’re like the engine room of the company,” Hassabis told CNBC, referring to his newly merged AI section within Google.

Last month, Alphabet CEO Sundar Pichai merged Hassabis’ DeepMind with Google Brain, a different AI team, and appointed Hassabis to run it. It’s now up to Hassabis to rebuild Google as the leader in generative AI, after the corporation was taken off guard by the rapid emergence of OpenAI, which is supported by Microsoft.

There may not be a more vital responsibility at Google, especially when new generative AI services provide users with new and innovative ways to search for content online. The business question is whether a long-time researcher like Hassabis can ship popular products.

Geoffrey Hinton, renowned as the “Godfather of AI,” says Hassabis’ determination to win is unquestionable.

“I don’t think I’ve ever met anybody more competitive,” said Hinton, who encouraged Google to buy DeepMind approximately ten years ago. “Demis is competitive at the level of people who get gold medals in the Olympics.”

He developed such tendency at an early age. Hassabis was a juvenile chess prodigy who was once regarded as the world’s number two player. He has also participated in the World Series of Poker.

Hassabis claims to have some consumer experience as well. In the 1990s, at the age of 17, he released the popular video game “Theme Park.” Hassabis stated that games at the time have to be enjoyable and simple to use in order to be successful. Following Theme Park, he founded his own video game firm, Elixir Studios.

Hassabis would eventually co-found DeepMind, which became well-known as the world’s leading AI research center, drawing some of the most notable deep learning researchers. When Google bought the lab for a rumored $500 million in 2014, DeepMind was granted a lot of freedom to operate freely.

Under Hassabis, DeepMind focused on building technology through games such as Breakout and AlphaGo, which defeated the world’s top Go player.

There was a practical rationale for focusing on gaming. According to Hassabis, “simulations are totally safe, have no consequences, but they can still learn from it.”

Hassabis led the field of artificial intelligence during his tenures at DeepMind and Google.

In 2018, Tesla CEO Elon Musk told OpenAI co-founders that competing with Hassabis and Google will require billions of dollars.

Google, on the other hand, would lose that advantage as competitors such as OpenAI and Microsoft introduced headline-grabbing technologies like ChatGPT and Copilot to the market.

According to someone who has worked with Hassabis but wishes to remain anonymous, Hassabis may have been more interested in earning academic awards than producing things that consumers could utilize at one point. Nature magazine, one of the world’s most renowned scientific periodicals, has published Hassabis’ work several times in the last decade.

“Accolades were never the end goal,” DeepMind told CNBC. “They simply reflect the importance and impact of the research they recognized.”

One of DeepMind’s most important products, AlphaFold, was a game-changing piece of technology that used AI to assist scientists in predicting the structure of proteins, which had been a major difficulty in biology for decades.

DeepMind open-sourced AlphaFold, effectively giving it out for free.

In 2017, a team of Google researchers working independently from DeepMind produced a groundbreaking study on Transformers, a method enabling AI models to better process training texts.

That sparked a wave of AI innovation that followed, particularly at OpenAI. Transformers are the “T” in ChatGPT.

Critics have claimed that Google was giving away valuable products and research while assisting its main competitors.

“They had the lead and they’ve been very cautious,” Hinton stated. “They’ve been very cautious with both generative AI images and large language models.” And when OpenAI partnered with Microsoft and ChatGPT was adopted by Microsoft, Google could no longer afford to remain careful.”

Hassabis and Google went on the offense.

The Washington Post reported in May 2023 that Google was internally announcing a significant break from previous policy. Employees were required to stop discussing their findings with the public, only publishing papers after the research had been turned into products.

AlphaFold became a significant financial prospect, attracting commercial partners like as Eli Lilly and Novartis for medication discovery.

During a TED lecture, Hassabis stated that the ChatGPT moment demonstrated that the public saw value in LLMs and was eager to adopt them.

“When we’re working on these systems, mostly you’re focusing on the flaws and the things they don’t do and hallucinations,” he stated. “We wanted to improve those things first before releasing them.” Surprisingly, despite these faults, tens of millions of people still find them extremely valuable.”

Hassabis stated that it was time to move these items “beyond the rarefied world of science.”

Investors are now waiting to see if Google can achieve its most essential goal: turning cutting-edge science into money.



Please enter your comment!
Please enter your name here

Most Popular

Recent Comments