why aren't all 16 resolved? a statement only gets an assessment when the public
record can support or contradict it. opinions and what-ifs never can, and 0 checkable
ones are still open, waiting for their date. predictions held up or didn't;
assertions are supported or contradicted. on every card:
▮▮▮▮▮ certainty ·
▮▮▮▮▮ debate potential. speakers are clickable
Assertion Supported
DeepMind Employees Had Google-Wide Access While Google Staff Was Barred From DeepMind
“People that worked at DeepMind could get into any Google office globally, but people that worked at Google were barred from DeepMind's premises.”
What-if
DeepMind Resisted Pivoting to Large Language Models Out of Hubris
“At some point in this period, DeepMind should have pivoted to language models. Just as OpenAI did, but DeepMind was too excited by its own research. It was accustomed to being the world's top AI lab. It could scarcely imagine that a copycat outfit might overta…”
Insight
Demis Hassabis Warns That Powerful Tech Advisors Inevitably Become Rivals
“If you have powerful people who are able to understand the impact of the technology, they're not just going to sit on the sidelines. They won't be content to just be your advisors. So obviously what was going on was our supposed advisors were really our rivals…”
Assertion Not publicly verifiable
A Single DeepMind Team Used More Compute Than Google's Entire Gmail Network
“In some weeks, a single research team at DeepMind Might gobble up more computational resources than Google's worldwide Gmail network, which had nine hundred million users.”
Opinion
Academic Culture Is Cautious on Innovation but Ruthlessly Competitive in Publishing
“The culture of academia could be both boringly cautious and terrifyingly competitive. Boring because it pursued incremental advances. Terrifying because scientists cut each other's throats To be the first to publish.”
Assertion Supported
Founders Fund Invested $2.3M for Nearly Half of DeepMind in 2010
“Founders fund wired 2.3 million to deep mine and they assumed ownership of a bit less than half the company. There was no other capital available. This is December, 2010.”
Assertion Partly supported
DeepMind used 32 algorithms to predict nearly all protein structures
“In 2020, DeepMind solved a second grand challenge in biochemistry, stitching together 32 algorithms to define the shape of nearly all the proteins in nature.”
Assertion Supported
Bill Gates Claimed True AI Would Be Worth Ten Microsofts
“If you invent a breakthrough in artificial intelligence so machines can learn, that would be worth 10 Microsofts.”
Assertion Supported
Investor Required DeepMind Founders to Pitch His Teenage Sons
“In September, 2010, Demis and Mustafa appeared before a strange kind of investment committee. David Gammon declared himself ready to commit capital, but he'd only go forward if DeepMind did things his way. Entrepreneurs seeking his support were required to vis…”
Assertion Supported
Google Acquired DeepMind for $650M with Hassabis Netting $136M
“Google bought DeepMind for six hundred and fifty million dollars. Demis netted one hundred and thirty six million dollars.”
Assertion Supported
DeepMind created an AI system defeating top Go players in 2016
“In 2016, DeepMind solved a grand challenge in computer science, creating a system that surpassed the intuitive brilliance of the world's best players of the ancient board game Go.”
Assertion Supported
Demis Hassabis Worked 10 PM to 4 AM Night Shifts Alongside Normal Office Hours
“Like Ender, Demis has dedicated every fiber of his being to the accomplishment of a mission, which is why he worked night shifts from 10 in the evening until four in the morning, in addition to his normal office hours.”
Assertion Supported
DeepMind made AlphaFold research freely available to scientists worldwide
“And when DeepMind's project was completed, the company gifted its results to science, allowing researchers all over the world to make free use of its discovery.”
Assertion Supported
DeepMind's Annual Staff Costs Reached $260M Following Google Acquisition
“Not long after the Google acquisition, DeepMind was paying two hundred and sixty million dollars in staff costs annually, six times more. Then it's total spending during its first three years of existence.”
Insight
Why Atari Video Games Were the Ideal Initial Benchmark for DeepMind
“Given the primitive state of video graphics in that era, the computing power required to crack Atari would be affordable. Given that Atari had released dozens of games, an agent would have plenty of opportunities to prove it could be general. Given that most A…”
Assertion Not checkable as stated
Demis Hassabis Conceived DeepMind's Core Idea During His Teenage Gap Year
“I decided then that I was going to dedicate my career to working on AI, Demis recalled. I had already had the kernel of the idea for what eventually became DeepMind.”