DeepMind Co-Founder Suleyman Departs For Investment Firm

DeepMind co-founder Mustafa Suleyman has left parent company Google after seven years to become a partner at Silicon Valley venture capital firm Greylock.

One of three co-founders of AI company DeepMind in 2010, Suleyman joined Google when it acquired the firm in 2014 for $500m.

His tenure was marked by struggles to ensure the firm’s parent company would use its AI ethically and by complaints by staff of an aggressive management style.

DeepMind’s efforts to create an independent AI ethics board to oversee its work were slow to build, and it made an unsuccessful attempt to create a legal structure to give it more independence within Google.

Data controversies

At the same time DeepMind was hit by controversies over the way it gathered data from partners such as the NHS.

Last October law firm Mishcon de Reya launched a class action claim with the UK High Court against Google and DeepMind over the way DeepMind accessed medical records under an NHS partnership.

In a recorded interview with Greylock partner Reid Hoffman released to coincide with Suleyman’s joining the company, Suleyman said “mistakes” had been made with its attempts to set up the oversight board.

“I’m not sure that we can say it was definitely successful, but I do believe that radical experimentation is essential here,” he said. “We’ve experimented with different oversight boards, with different ethical charters, with types of research.”

He said the broader industry has not “come close” to a situation where “people have significant influence” in shaping AI and its role in their lives.

AI oversight

Google’s own plans for AI oversight have struggled, with two heads of AI ethics leaving the company amidst criticism of the way their work was handled.

Suleyman was amongst those who resisted Google’s AI work with the US Department of Defense.

He was relieved of some of his responsibilities in 2019 after complaints from DeepMind workers about his management style, and moved to take a position at Google’s headquarters in California at the end of that year.

Addressing the complaints, he said, “I really screwed up. I was very demanding and pretty relentless.”

He added that he had set “unreasonable expectations” leading to “a very rough environment for some people. I remain very sorry about the impact that caused people and the hurt that people felt there.”

DeepMind, currently led by co-founder Demis Hassabis, created a system that beat the world champion at the board game Go in 2016, while its 2020 research into using AI to predict how protiens fold into different shapes was considered a potential breakthrough for drug discovery.

Matthew Broersma

Matt Broersma is a long standing tech freelance, who has worked for Ziff-Davis, ZDnet and other leading publications

Recent Posts

Creating Deepfake Porn Without Consent To Become A Crime

People who create sexually explicit ‘deepfakes’ of adults will face prosecution under a new law…

4 hours ago

Google Fires 28 Staff Over Israel Protest, Undertakes More Layoffs

Protest at cloud contract with Israel results in staff firings, in addition to layoffs of…

5 hours ago

Russia Already Meddling In US Election, Microsoft Warns

Microsoft warns of Russian influence campaigns have begun targetting upcoming US election, albeit at a…

6 hours ago

EU To Drop Microsoft’s OpenAI Investment Probe – Report

Microsoft to avoid an EU investigation into its $13 billion investment in OpenAI, after EC…

9 hours ago

US Provides Assurances For Julian Assange Extradition

As President Biden 'considers' request to drop Julian Assange extradition, US provides assurances to prevent…

11 hours ago