A man who gave up a slice of Anthropic so he could work for Rishi Sunak has described the decision as 'very expensive' – after it cost him $10million.
uncertain
it → give → 10million
Ian Hogarth, chairman of the UK's AI Security Institute (AISI), was an early investor in the leading artificial intelligence firm when very few people had heard of it.
However, he was asked by then-Prime Minister Mr Sunak to set up AISI in 2023 – and subsequently sold his shares in Anthropic to avoid any conflict of interest.
asserted
he → lead → interest
Mr Hogarth has now lost at least $10million after the company, which created the Claude AI platform, raised an unprecedented $81billion as it opened for trading – selling 980 million shares at $83 a piece on Wednesday.
asserted
it → lose → Wednesday
I've got a mortgage, right?
asserted
I → get → mortgage
' he told the Amol Rajan Connection podcast.
asserted
he → tell → podcast
He explained one of his motivations to sell the shares was witnessing 'grotesque' examples of 'graft' during the pandemic.
asserted
one → explain → pandemic
Despite tech outages and software blackouts sowing chaos across the US, Anthropic's shares opened at $111.50, 34 per cent above the $83 IPO pricing level, stunning Wall Street and giving the company a total market cap of more than $2trillion.
asserted
shares → sow → 2trillion
The AI revolution, Anthropic says, could lead to Nobel Prize-winning breakthroughs in medical research, potentially curing Alzheimer's and cancer – though there are also warnings of a 10 per cent chance it could kill humans within the next decade.
uncertain
it → say → decade
Ian Hogarth (pictured in 2023), chairman of the UK's AI Security Institute (AISI), was an early investor in the leading artificial intelligence firm when very few people had heard of it
Anthropic opened for trading this morning on the Nasdaq stock exchange, completing the largest initial public offering in history.
CEO Dario Amodei rang the Nasdaq opening bell at 9:30am to mark the occasion, as his company began trading under the ticker ANT.
AISI describes itself as the first state-backed organisation dedicated to ensuring advanced AI is 'safe, secure and beneficial'.
asserted
AI → picture → organisation
The organisation currently has over 100 technical staff, including 'senior alumni from OpenAI, Google DeepMind and the University of Oxford'.
asserted
organisation → have → Oxford
Some entrepreneurs who helped Mr Hogarth set up AISI are now working for Anthropic.
asserted
Hogarth → help → Anthropic
Mr Hogarth grew up in south London and attended Dulwich College before studying engineering at Cambridge.
asserted
Hogarth → grow → Cambridge
He worked at Bain before co-founding the music app Songkick, sold to Warner Music in 2017.
asserted
He → work → 2017
He is now a partner at Plural Platform, a venture capital firm which has raised €1billion to invest in European start-ups.
asserted
which → raise → ups
The unlucky decision comes as former British state school boy Jack Clark was seen beaming from ear to ear on Thursday during the market debut of Anthropic – which saw his fortune swell to $26billion.
asserted
fortune → come → 26billion
Former British state school boy Jack Clark beamed from ear to ear today during the stunning market debut of world-leading AI company Anthropic
Co-Founder of AI research company Anthropic, Jack Clark pictured with wife Olivia, the mother of his two young daughters
From a reporter earning less than £20,000 a year to a billionaire tech tycoon in little more than a decade, Jack Clark's rise has seen him branded the world's most unlikely tech bro.
asserted
him → beam → decade
Brighton-born Jack, 37, became one of the richest men on the planet as his artificial intelligence giant Anthropic undertook its hotly anticipated stock market debut in New York.
asserted
giant → bear → York
He credits part of his success to a hobby of printing out academic papers on AI, particularly research into how computers competed against humans in simple online games.
asserted
computers → credit → games
He believes it helped him spot the technology's potential earlier, between 2013 and 2016 when he was working as a writer for Bloomberg.
asserted
he → believe → Bloomberg
'I wasted a lot of Mr Bloomberg's printer ink by printing out archive papers about AI research,' he said recently.
asserted
he → waste → research
Despite his now vast fortune, he lives a notably frugal life outside work, spending time with his wife and daughters, hiking, watching Doctor Who and catching garage-band gigs.
asserted
he → live → gigs
Approximately 5,300 Anthropic employees own equity grants that will be worth tens of billions of dollars once their lockup periods expire.
asserted
periods → own → dollars
Around 1,000 early hires own stock grants worth $100million or more a piece, while thousands of newer hires hold smaller packages averaging around $20million each.
asserted
thousands → own → 20million
All in all, the company’s employees hold around $216billion in stock at the IPO price, with another $160billion reserved for future hires.
asserted
160billion → hold → hires
Ashton Kutcher’s venture capital firm Sound Ventures invested $435million in Anthropic back in 2023, when the company was valued around $5billion – and after the IPO the firm’s position is worth around $7billion, firmly establishing Kutcher as a billionaire.
asserted
position → invest → billionaire
In addition, as a public benefit corporation Anthropic allocated 5 per cent of its shares to a dedicated fund aimed at supporting the responsible development of AI.
asserted
Anthropic → allocate → AI
In its investment prospectus, Anthropic wrote that 'We believe that AI is one of the most consequential technologies in human history.'
asserted
AI → write → history
The company believes it could rapidly accelerate scientific innovation, greatly increase economic growth rates, bring cures for major diseases within reach and solve problems that have resisted generations of human effort.
uncertain
that → believe → effort
But the same document offers a stark warning: 'Sufficiently advanced AI systems could... pose catastrophic or existential risks to humanity.'
uncertain
systems → offer → humanity
And in a video posted earlier this week, one Anthropic researcher warned that AI would likely leave 'pretty much all humans on Earth' permanently unemployable and dependent on government welfare.
asserted
AI → post → welfare
Tao Lin, who works on safeguards for Anthropic’s Claude models, said in an interview with Palisade Research that he expects 'broadly superhuman AI' in almost every field by around the end of 2027.
asserted
he → work → 2027