• In August 2019, Mustafa Suleyman announced he was stepping away from DeepMind, adding that he needed "break to recharge."
• Less than half a year later, in December 2019, he announced that he was officially leaving the AI lab to join Google as VP of AI product management and AI policy.
• Suleyman's exit from Google, which acquired DeepMind in 2014, comes after he was accused of having aggressive management style by former colleagues at DeepMind.
• Explaining the rationale behind the move to VC, Suleyman told LinkedIn billionaire and Greylock partner Reid Hoffman on a podcast last week that he wants to be around founders who are visionary and fearless.

LONDON — DeepMind co-founder Mustafa Suleyman surprised many of his followers last week when he announced he's leaving his vice president role at Google to become a venture capitalist at Silicon Valley firm Greylock Partners, which has backed the likes of Facebook, Airbnb and LinkedIn since it was founded in 1965.

His exit from Google, which acquired the DeepMind artificial intelligence lab in 2014, comes after he was accused of having an aggressive management style by former colleagues at DeepMind.

Explaining the rationale behind the move, Suleyman told LinkedIn billionaire and Greylock partner Reid Hoffman on a podcast last week that he wants to be around founders who are visionary and fearless.

"I'm definitely somebody who likes to take risks," Suleyman said on the podcast, which was released last Thursday. "I find it super energizing when I'm around people who also have a courageous vision of the future, which sounds wacky or implausible, but are prepared to dedicate their lives to giving it a shot."

He added: "They're the kinds of people that I like to back and I think that's what we need. We need more people who are prepared to try and do bold things and tackle hard problems to try to improve our world."

Suleyman, widely known as "Moose," declined to talk to CNBC. However, in an exclusive interview with TechCrunch about his new role, he said that he thinks AI has a central role to play in gaming and the so-called metaverse.

One former DeepMind employee, who asked to remain anonymous because of the sensitive nature of the discussion, said they were surprised to read that Suleyman's main interests seem to be around the metaverse and gaming.

"In the past, Mustafa had a truly admirable focus on genuinely trying to make the world a better place," they said. "In particular, whilst at DeepMind, he played an essential role in health care and climate change projects. So I'm surprised that Suleyman's main interests now seem to be around the metaverse and gaming. I hope he will also find time to focus on fixing some of the deep problems facing the world."

Other tech investors said they think Suleyman, who has already made a number of personal investments, will make a good VC.

"I think Mustafa is likely to be a great investor given his track record in collaborating with exceptional founders and his early conviction as an investor in [start-up builder] Entrepreneur First," Ian Hogarth, an angel investor and the co-founder of concert discovery app Songkick, told CNBC.

Two of Suleyman's other public investments include music ticketing app Dice and health-care app Babylon Health.

Tom Hulme, a venture capital partner at GV (formerly Google Ventures), told CNBC that Suleyman has been excited about the VC industry for a while.

But another VC, who asked to remain anonymous because of the sensitive nature of the discussion, questioned how long Suleyman would remain a VC for. "My gut says that it's temporary while he looks for the next company to build or join as a founder," they told CNBC. "I think he has more left in the tank."

Selling DeepMind to Google

Suleyman co-founded DeepMind in London with childhood friend Demis Hassabis and New Zealander Shane Legg in 2010. In the lead-up to the Google acquisition, Suleyman helped DeepMind to raise millions of dollars from billionaires including Elon Musk and Peter Thiel.

Suleyman, who dropped out of his undergraduate philosophy degree at Oxford University to set up a Muslim helpline, led DeepMind's applied AI efforts for several years both pre- and post-acquisition.

This involved trying to find novel uses for the company's algorithms across Google's various products and services, as well as at other organizations including the U.K.'s National Health Service and National Grid.

While DeepMind found some clever uses for its technology in Google's data centers and within apps like YouTube, its external commercial endeavors have been less successful.

DeepMind is yet to make any serious revenue from selling its software to third-party organizations. Financial filings with the U.K. company registry show that it has operated at a loss every year since it was acquired except last year, when it posted a profit of £43.8 million ($59.6 million). Prior to that, it reported a loss of $649 million in 2019.

Beyond applying AI, Suleyman also oversaw DeepMind's work on AI ethics and that involved trying to set up an independent board to oversee the lab's research, which could one day have a huge impact on the world. DeepMind is ultimately trying to create super-intelligent machines that can outsmart humans on many levels and make even smarter versions of themselves.

"We made a lot of mistakes in the way that we attempted to set up the board, and I'm not sure that we can say it was definitively successful, but I do believe that radical experimentation is essential here," Suleyman said on the podcast. "We need new forms of governance and new forms of oversight that are fit for the modern age."

DeepMind experimented with different oversight boards, ethical charters and types research, Suleyman said.

Speaking about the wider technology industry, he said: "I definitely feel that we haven't really come close to cracking this nut of how we make technology platforms, software, and of course AI feel like it's happening with people, and where people have significant influence in shaping how it arrives in their world and doesn't just happen to people."

Controversial exit from DeepMind

In August 2019, Suleyman announced on Twitter that he was stepping away from DeepMind, adding that he needed a "break to recharge." Less than half a year later, in December 2019, he announced that he was officially leaving the AI lab he helped to build to join Google as VP of AI product management and AI policy.

The full circumstances of Suleyman's departure from DeepMind weren't disclosed at the time, but it later emerged that a number of his colleagues had taken issue with his management style, accusing him of harassment and bullying. In January 2021, DeepMind announced it had brought in a law firm to investigate his management style.

"I had a period in 2017-2018 where a couple of colleagues made a complaint about my management style" Suleyman said on the podcast. "You know, I really screwed up. I was very demanding and pretty relentless. I think that at times that created an environment where I basically had pretty unreasonable expectations of what people were to be delivering and when."

He added that he ended up being "pretty hard charging" and that this created a "rough environment" for some people. "I remain very sorry about the impact that that caused people and the hurt that people felt," Suleyman said.

Suleyman said the complaints gave him the opportunity to "take a step back and reflect" and to "grow and mature" as a manager and a leader. He admitted that he was "super focused on speed and pace over being caring and attentive to how people are feeling."

Suleyman says he has been seeing a coach for the last few years as part of an effort to address the issues that his former colleagues raised.