Facts
Stated by the speakers. Each quote matches the transcript, which shows what was said, not that it is true.
In 2020 DeepMind combined 32 algorithms to predict the shapes of nearly all proteins found in nature.
“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.”
0:04:13 · David Senra · AlphaFold · 2020
Demis Hassabis and John Jumper were awarded the 2024 Nobel Prize in Chemistry for protein structure prediction.
“So in 2024, Demis and John Jumper were awarded the Nobel Prize in chemistry for protein structure prediction.”
0:04:37 · David Senra · Nobel Prize · 2024
Demis Hassabis discovered chess at age four.
“Luckily for Demis, he discovers chess when he's four.”
0:12:25 · David Senra · Childhood
At age twelve Demis Hassabis built a computer program that played Othello, applying ideas from David Levy's chess computer book.
“Twelve-year-old Demis sets out applying Levy's principles. He built a computer program to play a simpler game, the game Othello.”
0:15:53 · David Senra · Early programming
Bullfrog's founder offered Demis Hassabis a check for 500,000 pounds to stay and work on its next game, and Hassabis refused in order to attend Cambridge.
“He writes out a check for 500 ,000 pounds to get him to work on Bullfrog's next game.”
0:18:40 · David Senra · Bullfrog
Venture capitalists offered Demis Hassabis money for his first company, Elixir, only if he gave up more than half its equity.
“they said they'll give him some money, but you have to give up more than half the equity of the company.”
0:21:36 · David Senra · Elixir funding
The overly ambitious game Republic led to the failure of Demis Hassabis's first company, Elixir.
“Republic was this overly ambitious, basically technically impossible idea of a game. It's going to lead to the failure of his first company.”
0:22:50 · David Senra · Elixir failure
Demis Hassabis approached Peter Thiel at a party by talking about chess, which Thiel was obsessed with.
“So he starts talking about chess with Peter because he knew that Peter was obsessed with chess.”
0:25:14 · David Senra · DeepMind funding · 2010
DeepMind's early pitch deck quoted Bill Gates saying a breakthrough in machine learning would be worth 10 Microsofts.
“If you invent a breakthrough in artificial intelligence so machines can learn, that would be worth 10 Microsofts.”
0:26:24 · David Senra · Pitch deck · 2010
In September 2010 investor David Gammon said he would fund DeepMind only if it pitched to his whole family, each with an equal vote.
“David Gammon declared himself ready to commit capital, but he'd only go forward if DeepMind did things his way.”
0:27:51 · David Senra · Fundraising · 2010
Founders Fund wired $2.3 million to DeepMind in December 2010 and took ownership of slightly less than half the company.
“founders fund wired 2.3 million to deep mine and they assumed ownership of a bit less than half the company”
0:29:11 · David Senra · Founders Fund · 2010
DeepMind chose Atari's 1970s and 80s video games as its first testing ground for a general learning agent.
“And so they start out on what they think is the perfect environment for testing an agent. All the video games that were designed in the 1970s and 80s by Atari.”
0:33:08 · David Senra · Atari games · 1970s-1980s
Elon Musk told Demis Hassabis that humans needed to colonize Mars in case of disaster, while Hassabis argued AI mattered as much.
“Elon had declared that humans needed to colonize Mars in case disaster struck Earth.”
0:35:10 · David Senra · Elon Musk
When Elon Musk offered to have Tesla acquire DeepMind, Hassabis said Tesla did not generate enough cash to support its research.
“Demis pointed out that Tesla was not generating enough cash to support DeepMind's research.”
0:39:02 · David Senra · Acquisition talks · 2014
Google bought DeepMind in January 2014 for $650 million, of which Demis Hassabis netted $136 million.
“But in January 2014, Demis goes ahead with Google, and this is something he's never regretted. Google bought DeepMind for $650 million. Demis netted $136 million.”
0:39:23 · David Senra · Google acquisition · 2014
Soon after the Google acquisition, DeepMind's annual staff costs reached $260 million, six times its total spending in its first three years.
“Not long after the Google acquisition, DeepMind was paying $260 million in staff costs annually, six times more than its total spending during its first three years of existence.”
0:39:23 · David Senra · DeepMind spending · 2014
After the acquisition, DeepMind staff could enter any Google office, while Google staff were barred from DeepMind's premises.
“People that worked at DeepMind could get into any Google office globally, but people that worked at Google were barred from DeepMind's premises.”
0:41:49 · David Senra · Google arrangement
In an early 2015 email to Elon Musk, Sam Altman said it was almost certainly impossible to stop humanity developing AI and that someone other than Google should do it first.
“I've been thinking a lot about whether it's possible to stop humanity from developing AI. I think the answer is almost definitely not.”
0:42:24 · David Senra · OpenAI origins · 2015
Elon Musk and Sam Altman launched OpenAI as a not-for-profit lab aimed at breaking the Google DeepMind monopoly on AGI.
“Elon and Sam Altman team up to launch OpenAI, a not-for-profit lab explicitly aimed at breaking the Google DeepMind AGI monopoly.”
0:43:12 · David Senra · OpenAI founding · 2015
AlphaGo Zero learned only by playing itself and greatly outperformed the earlier AlphaGo, which had trained on expert human games.
“Learning only from self-play, the system outclassed its predecessor by a mile.”
0:43:42 · David Senra · AlphaGo Zero
DeepMind released its protein-folding results to science so researchers worldwide could use them free.
“the company gifted its results to science, allowing researchers all over the world to make free use of its discovery.”
0:47:48 · David Senra · AlphaFold release
DeepMind's AlphaFold team, perhaps about 20 people at its peak, outperformed hundreds of academic scientists who had worked on protein folding for decades.
“Hundreds of academic scientists had spent decades on protein folding. How was it that the DeepMind's team, numbering perhaps 20 at its peak, had defeated all of them?”
0:48:59 · David Senra · AlphaFold team
ChatGPT reached a million users within five days and 100 million within two months, making it the fastest-growing consumer application ever.
“Within five days, they collected a million users. Within two months, it had amassed an astonishing 100 million, making it the fastest growing consumer application ever.”
0:49:20 · David Senra · ChatGPT
In response to ChatGPT, Google planned to merge Google Brain and DeepMind and build a single chatbot model called Gemini.
“They also planned to merge Google Brain and DeepMind. Not even Google could afford the luxury of duplicate research teams.”
0:51:12 · David Senra · Google DeepMind merger · 2023
At Google DeepMind any team member could propose and test an improvement to the Gemini model, and upgrades that improved performance were added to the shared master code.
“Any team member was welcome to propose an improvement to the model and test it. If the upgrade boosted performance, it was added to the master code on which everyone was building.”
0:52:27 · David Senra · Gemini development