AlphaFold Concepts

The Infinity Machine: Demis Hassabis, DeepMind, and the Quest for Superintelligence

by Sebastian Mallaby; · 30 Mar 2026 · 607pp · 161,998 words

that DeepMind excelled at. With the search algorithm in place, DeepMind had completed the design of its first serious protein-prediction model, which it called AlphaFold. The “Alpha” name was a bit of a trick: a signal that DeepMind was advancing serenely from one model to the next—from AlphaGo to

…

AlphaZero to AlphaStar to AlphaFold. Given the reality that the protein team had executed a series of pivots, landing on a system consisting of deep learning, some thought the

…

model should have been called DeepFold. But the AlphaFold name was at least partly justified. Even if the self-play part of AlphaGo and AlphaZero could not be applied to proteins, the lineage had

…

been preserved at AlphaFold’s last stage. The search was, as Jumper put it, at least “semi-RL.” It was a lonely vestige of the reinforcement-learning approach

…

his model announced that DeepMind would place twentieth. Happily, another researcher pointed out that the prediction was based on a statistical mistake. Everybody hoped that AlphaFold itself would be less prone to error.[15] CASP got underway in May, with ninety-eight teams participating. DeepMind now proceeded on two tracks: The

…

potential upgrades on the cutting room floor, and now they scooped them up and tested them. But it felt like they had hit a wall. AlphaFold’s accuracy score had plateaued at just under sixty GDT, meaning that it accurately predicted the position of nearly 60 percent of the main atoms

…

fully solving protein folding was too hard: None of those ideas on the cutting room floor were proving to be fruitful. On the other hand, AlphaFold might win CASP that year. Senior wanted to claim victory and wrap up the project. Hassabis objected. He didn’t want to be the best

…

DeepMind might crack the protein problem. * * * • • • The leading contender for the next breakthrough was known as “direct folding.” The idea grew out of an oddity: AlphaFold’s first module, the convolutional network, sometimes seemed to know the shape of a protein even before the search module had discovered it. Without waiting

…

the folding conundrum directly: to compute the exact location of each atom in the final protein structure. The first results from this pivot were lousy—AlphaFold’s GDT score crashed from around sixty to around twenty. But the DeepMind team had faith: They were willing to jump off a cliff and

…

contenders assembled in the seaside sun to hear about their GDT scores. DeepMind’s jitters on the eve of the contest were now permanently erased: AlphaFold bested the other ninety-seven teams at CASP, thanks mostly to its shift from contact maps to distograms.[18] In the contest’s hardest category

…

—which involved “free modeling,” the prediction of structures for which no evolutionary template was known—AlphaFold was most accurate in twenty-five out of forty-three cases; its nearest rival came first in just three of them. The other scientific teams

…

understand biology, you needed more than biological intelligence. Even as he savored victory, Jumper’s attention was elsewhere. For one thing, he knew that the AlphaFold system being celebrated in Cancun would soon be surpassed by the direct-folding version, though of course he didn’t mention this to his rivals

…

Jumper convened his enlarged team and announced a period of exploration. The direct-folding innovation had proved again the value of pivots. But to boost AlphaFold’s GDT score from around sixty to ninety, the team needed further inspiration. Each researcher was invited to show up at the next meeting with

…

a single slide, proposing a blue-sky idea that could transform AlphaFold’s accuracy. For the next three months, an extended hackathon followed. People huddled at whiteboards, traded hypotheses across their desks, and bashed out algorithmic

…

apart from one another on a chain, grasping that these acids would interact in the folded protein structure. By the middle of 2019, the revamped AlphaFold, dubbed AlphaFold 2, was working. At its core was a family of specialized and extremely complex transformers—there was one called the “tetraformer,” which combined four

…

different variations on the standard transformer architecture.[20] As AlphaFold 2 grew more powerful and accurate, DeepMind fed its highest-confidence protein-structure predictions back into the model’s training set. Success fed success. The

…

then he’d add another five to it.”[21] In December 2019, Meyer livened up the protein team’s weekly meetings by playing vintage soundtracks. AlphaFold’s GDT score had reached eighty-four, so he played hits from 1984: Tina Turner’s “What’s Love Got to Do with It,” Frankie

…

he played “U Can’t Touch This,” the iconic hip-hop anthem by MC Hammer, which was released in 1990. It was a jubilant moment. AlphaFold had attained a GDT score of ninety, the accuracy at which X-ray crystallography became obsolete. In May 2020, the next CASP contest started. Over

…

Professor John Moult, the founder and organizer of CASP, received the final scores for that year’s contest. He did a double take: DeepMind’s AlphaFold 2 had scored 92.4, more than 50 percent higher than the best score ever previously recorded. This was the fourteenth CASP competition over which

…

do? How could they know whether DeepMind’s mind-boggling accuracy was legitimate? Together, Moult and Lupas came up with a test: They would ask AlphaFold to predict the configuration of a shape that could not be in its training set because X-ray crystallography had failed to unravel it. There

…

one of these and sent it to DeepMind. DeepMind sent its answer back, and Lupas compared it to his incomplete X-ray mapping. Sure enough, AlphaFold 2’s prediction conformed to the bits of the protein that Lupas had experimentally established; moreover, it predicted the rest of the structure in a

…

“I always hoped I would live to see this day,” Moult said. “But it wasn’t always obvious I was going to make it.”[24] * * * • • • AlphaFold’s breakthrough signaled three kinds of change: for practical discovery, for the scientific establishment, and for the standing of artificial intelligence. In the practical arena

…

, the effects came quickly. As soon as CASP confirmed the accuracy of AlphaFold 2’s predictions, DeepMind cataloged the shapes of all 20,000 proteins in the human proteome, 83 percent of which had not been mapped out

…

“That’s a thing I love about AI,” Hassabis said. “You can have your Christmas lunch while it does something useful.” By the following summer, AlphaFold had plotted 350,000 structures, occurring in everything from yeast to fruit flies. By July 2022, it had folded around 200 million proteins in total

…

made their best discoveries in their late forties, not earlier, and collaborations were growing larger and unwieldier, with some journal papers listing thousands of coauthors. AlphaFold’s triumph served only to deepen the anxiety about this trend. Big pharma had allowed a rank outsider to march onto its turf: What did

…

the worries about mainstream science were the flip side of the excitement about the new science, powered by artificial intelligence. To Hassabis and his followers, AlphaFold’s success signaled a golden era of discovery, touching everything from coding to chemistry. In June 2023, DeepMind announced AlphaDev, a computer science counterpart to

…

in the International Mathematical Olympiad; and a model called GenCast beat the state of the art in weather forecasting. In May 2024, DeepMind rolled out AlphaFold 3. Rather than just divining protein shapes, this iteration predicted the reactions between proteins and other types of molecules.[28] The world appeared to be

…

“I mean, everyone talks about the benefits of language models, but mostly it’s just cheese tomorrow. The clearest benefit from AI so far is AlphaFold. “And I want to go further, as quickly as possible. Actually come up with some breakthrough medicines. Show you can do that in one year

…

rival DeepMind language project. But Hassabis and his lieutenants were skeptical of the potential of large language models, disinclined to follow OpenAI, and preoccupied with AlphaFold, StarCraft II, and governance wrangles with Google. Irving’s arrival tipped the balance at DeepMind. He had spent time inside the belly of the rival

…

have milked the first-mover advantage. “I don’t know if I was ready to pivot earlier because I was in my science phase, doing AlphaFold,” Hassabis said. “The number one thing I wanted to show was that AI could create incredible scientific breakthroughs. It was important for the world to

…

understand that. “But now I have really scratched that itch. AlphaFold is so massive, I’m not sure I can top it. Short of solving physics and the nature of reality, which is my long-term

…

and Oriol Vinyals, Borgeaud, Rae, and a posse of colleagues set about applying the techniques that had worked for DeepMind strike teams from Atari to AlphaFold. First, they embraced a strict unity. All team members poured their energies into improving one single model; no parallel projects were permitted. Next, they

…

2019, DeepMind’s games-playing reinforcement learning systems had set the pace: Atari, AlphaGo, AlphaZero, AlphaStar. Next, with GPT-2 in 2019, and indeed with AlphaFold in 2020, reinforcement learning had been eclipsed: The transformer-based architecture proved so powerful that DeepMind’s machine-based RL (as distinct from the human

…

years or so.[25] The question was how to get around this “data wall.” One established answer was to use AI-generated data. To train AlphaFold, for example, DeepMind had fed the model’s own protein-structure predictions back into its training set. More often, data generated by a large “teacher

…

meeting had been an indication that, with proper organization, success would come fast. As Hassabis had said when describing the case for doubling down on AlphaFold, if ideas are flowing fluidly, that is the signal to push forward. On December 19, 2024, as part of a crush of product launches

…

You could show Astra a broken appliance and ask for tips on fixing it. I asked Hassabis to explain his contribution to these advances. With AlphaFold, Hassabis had dropped in on scientific meetings, shuffled the leadership of the strike team, and discussed the project with John Jumper at two o’clock

…

be postponed. Humanity would benefit from the resulting risk reduction, without giving up much. After all, the most beneficial AI breakthrough to date, DeepMind’s AlphaFold, had required no reinforcement learning. On the stage at Davos, Hassabis brushed Bengio’s idea aside, much as he had rejected the 2023 pause letter

…

not just a vision, Silver and Sutton continued. It was already a reality. By way of illustration, the authors cited AlphaProof, a mathematical counterpart to AlphaFold, and the newest brainchild to emerge from Silver’s team of scientists. AlphaProof incorporated a specialized version of Gemini, so it could take in word

…

of the problem of induction—the invention of machines that could induce patterns in an infinity of data—changed the pace at which science proceeded. AlphaFold heralded infinite discovery, courtesy of infinity machines. “It’s also living life at digital speed,” Hassabis responded. “Maybe it’s kind of like speed-

…

all you could wish for. The average chain of amino acids could theoretically be twisted into 10300 possible forms—trillions upon trillions upon trillions. Yet AlphaFold divined the correct shape of the folded chain, no superpositions necessary. Perhaps the purest contrast between Penrose and Hassabis went back to their common starting

…

DeepMind, January 25, 2022, 39 min., 14 sec., youtube.com/watch?v=ZfJhOTZi0WE. BACK TO NOTE REFERENCE 15 By this point, Jumper had already discarded AlphaFold’s original search algorithm, replacing it with a simpler alternative. BACK TO NOTE REFERENCE 16 Pushmeet Kohli, author interview, June 26, 2023. BACK TO NOTE

…

Race Against Antibiotic Resistance,” Google DeepMind, July 28, 2022, deepmind.google/discover/blog/accelerating-the-race-against-antibiotic-resistance. BACK TO NOTE REFERENCE 26 AlQuraishi, “AlphaFold @ CASP13: ‘What Just Happened?’ ” BACK TO NOTE REFERENCE 27 Outside DeepMind, scientific teams reported huge productivity gains from integrating AI: At one US materials research

…

BACK TO NOTE REFERENCE 10 Rae formed an alliance with London colleagues, including Jonas Adler, Alexander Pritzel, and Sebastian Borgeaud. Adler had worked on the AlphaFold strike team. BACK TO NOTE REFERENCE 11 As of June 2025, no Gemini rival had a context window longer than 200,000 tokens. BACK TO

…

injury (AKI), 178–79, 183–85, 188–90 The Age of Spiritual Machines (Kurzweil), 57–58 Alphabet, 231–32 alpha-beta search, 402n17 AlphaDev, 278 AlphaFold, 382 CASP contest and, 269–73, 276–77 direct folding and, 271–73 success of, 277–79, 313–14 transformers for, 274–75 AlphaGeometry, 278

…

. See also AlphaGo; Gemini; Google acquisition of DeepMind achievements of, xvii–xviii AGI believers and skeptics in, 88 alignment team at, 323 AlphaDev of, 278 AlphaFold and, 269–79, 313–14, 382 AlphaProof and, 378 AlphaStar project of, 225–28 AlphaZero and, 193–200, 212–13, 227, 357 Applied division of

…

achievements of, xvii–xviii affability of, 21–22 on AGI and intuition, 143 on AGI’s potential, 112–15 on AI governance, 255–56 on AlphaFold’s success, 313–14 AlphaGo challenge and, 141 AlphaStar and, 225, 227–28 Altman compared with, 290, 292, 294, 349 ambitions of, xiv–xv, xx

The Singularity Is Nearer: When We Merge with AI

by Ray Kurzweil · 25 Jun 2024

pharmaceutical breakthroughs.[14] This is where the pattern recognition capabilities of AI offer a profound advantage. In 2018 Alphabet’s DeepMind created a program called AlphaFold, which competed against the leading protein-folding predictors, including both human scientists and earlier software-driven approaches.[15] DeepMind did not use the usual method

…

of drawing on a catalog of protein shapes to be used as models. Like AlphaGo Zero, it dispensed with established human knowledge. AlphaFold placed a prominent first out of ninety-eight competing programs, having accurately predicted twenty-five out of forty-three proteins, whereas the second-place competitor

…

experiments, so DeepMind went back to the drawing board and incorporated transformers—the deep-learning technique that powers GPT-3. In 2021 DeepMind publicly released AlphaFold 2, which achieved a truly stunning breakthrough.[17] The AI is now able to achieve nearly experimental-level accuracy for almost any protein it is

…

(January 15, 2020), https://doi.org/10.1038/s41586-019-1923-7. BACK TO NOTE REFERENCE 14 For more detail on how the original AlphaFold achieved great progress on protein folding, see Andrew W. Senior et al., “AlphaFold: Using AI for Scientific Discovery,” DeepMind, January 15, 2020, https://deepmind.com/blog/article

…

/AlphaFold-Using-AI-for-scientific-discovery; Andrew Senior, “AlphaFold: Improved Protein Structure Prediction Using Potentials from Deep Learning,” Institute for Protein Design, YouTube video, August 23, 2019

…

Sample, “Google’s DeepMind Predicts 3D Shapes of Proteins,” Guardian, December 2, 2018, https://www.theguardian.com/science/2018/dec/02/google-deepminds-ai-program-alphafold-predicts-3d-shapes-of-proteins; Matt Reynolds, “DeepMind’s AI Is Getting Closer to Its First Big Real-World Application,” Wired, January 15, 2020, https

…

://www.wired.co.uk/article/deepmind-protein-folding-alphafold. BACK TO NOTE REFERENCE 16 For some more detailed explainers of AlphaFold 2 and the scientific paper describing it, see “AlphaFold: The Making of a Scientific Breakthrough,” DeepMind, YouTube video, November 30, 2020, https://www.youtube

…

.com/watch?v=gg7WjuFs8F4; “DeepMind Solves Protein Folding | AlphaFold 2,” Lex Fridman, YouTube video, December 2, 2020, https://www.youtube.com/watch?v=W7wJDJ56c88; Ewen Callaway, “ ‘It Will Change Everything’: DeepMind’s AI Makes

…

Structures,” Nature 588, no. 7837 (November 30, 2020): 203–4, https://doi.org/10.1038/d41586-020-03348-4; Demis Hassabis, “Putting the Power of AlphaFold into the World’s Hands,” DeepMind, July 22, 2022, https://deepmind.com/blog/article/putting-the-power-of

…

-alphafold-into-the-worlds-hands; John Jumper et al., “Highly Accurate Protein Structure Prediction with AlphaFold,” Nature 596, no. 7873 (July 15, 2021): 583–89, https://doi.org/10.1038/s41586-021-03819-2. BACK

…

://doi.org/10.1038/d41586-021-02265-4. BACK TO NOTE REFERENCE 18 Hassabis, “Putting the Power of AlphaFold into the World’s Hands”; Jumper et al., “Highly Accurate Protein Structure Prediction with AlphaFold.” BACK TO NOTE REFERENCE 19 For relatively simple explainers of these methods, see National Cancer Institute, “CAR T

…

social media, 114–15 alienation, 230 Alphabet. See also Google DeepMind, 41, 42, 47, 50, 196, 238–39 Waymo, 43, 195–96 AlphaCode, 50 AlphaFold, 238–39 AlphaFold 2, 239 AlphaGo, 41–42, 369n AlphaGo Master, 41 AlphaGo Zero, 41–42, 238, 369n AlphaZero, 42 Alquist 3D, 188 Altair 8800, 131, 165

Human Frontiers: The Future of Big Ideas in an Age of Small Thinking

by Michael Bhaskar · 2 Nov 2021

problems of our time. What did just happen? The artificial intelligence company DeepMind, part of the Alphabet group, had been quietly working on software called AlphaFold. DeepMind uses deep learning neural networks, a newly potent technique of machine learning (ML), to predict how proteins fold. These networks aim to mimic the

…

known. Based on its results the neural network keeps learning, getting more accurate the more it guesses to build a highly effective model. At CASP13, AlphaFold was better able to guess the distance and angles between pairs of amino acids. It was a simple, unexpected but powerful approach, supported by world

…

years, humans hadn't thought of. A machine did. Thanks to the program, previously unthinkable moves are now part of the tactical lexicon. AlphaGo, like AlphaFold, jolted the game out of a local maximum. DeepMind is at the forefront of a well-publicised renaissance in AI. (AI itself is a big

…

. New forms of knowledge and perception, quite unlike those of humans, beyond our unaided capabilities, are starting to accelerate the production of ideas. AlphaGo and AlphaFold are signposts to an era where those closest to a particular toolset are best positioned to push back the frontiers of knowledge. Proximity to these

…

hints, this isn't just a matter of inching forward or copying humans; it adds a qualitatively different dimension. If that sounds overdone, consider that AlphaFold wasn't finished in Cancun. Two years later, as I was writing this book, at CASP14 a new program, AlphaFold2, emerged. It performed even better

…

our time. Demis Hassabis himself makes the link explicit, calling AI a sort of general-purpose Hubble space telescope for science.20 Big ideas like AlphaFold and AlphaGo, instances of the big idea of deep learning neural networks, are steadily making a difference at the coalface. To see how AI reshapes

…

2020, available at https://slatestarcodex.com/2018/11/26/is-science-slowing-down-2/ AlQuraishi, Mohammed (2018), ‘AlphaFold @ CASP13: “What just happened?”’, accessed 5 February 2020, available at https://moalquraishi.wordpress.com/2018/12/09/alphafold-casp13-what-just-happened/ AlQuraishi, Mohammed (2020), ‘AlphaFold2 @ CASP14: “It feels like one's child has

…

getting closer to its first big real-world application’, Wired, accessed 5 February 2020, available at https://www.wired.co.uk/article/deepmind-protein-folding-alphafold Ricón, José Luis (2015), ‘Is there R&D spending myopia?’, Nintil, accessed 6 January 2021, available at https://nintil.com/is-there-rd-spending-myopia

…

an Uncertain World, Chichester: John Wiley Senior, Andrew, Jumper, John, Hassabis, Demis, and Kohli, Pushmeet (2020), ‘AlphaFold: Using AI for scientific discovery’, DeepMind, accessed 5 February 2020, available at https://deepmind.com/blog/article/AlphaFold-Using-AI-for-scientific-discovery Shambaugh, Jay, Nunn, Ryan, Breitwieser, Audrey, and Liu, Patrick (2018), The State

…

, 295, 304 algorithms 175, 185, 196, 224, 235, 245 aliens 240–1, 306, 308–9, 337 Allison, Jim 58 Alphabet 193, 225, 265, 294, 295 AlphaFold software 225–6, 227, 228–9, 233 AlphaGo software 226–7, 228, 233 AlQuraishi, Mohammed 225, 226, 229 Amazon 84–5, 214, 272 Amazon Prime

The Coming Wave: Technology, Power, and the Twenty-First Century's Greatest Dilemma

by Mustafa Suleyman · 4 Sep 2023 · 444pp · 117,770 words

, a rank outsider entrant arrived at the competition, with zero track record, and beat ninety-eight established teams. The winning team was DeepMind’s. Called AlphaFold, the project started during a weeklong experimental hackathon in my group at the company back in 2016. It grew to become a landmark moment in

…

: “One of the Biggest Problems in Biology Has Finally Been Solved,” wrote Scientific American. A previously hidden universe of proteins was revealed at staggering speed. AlphaFold was so good that CASP was, like ImageNet, retired. For half a century protein folding had been one of science’s grand challenges, and then

…

space. Parts are too complex to build using conventional tooling and have to be 3-D printed. In chapter 5, we saw what tools like AlphaFold are doing to catalyze biotech. Until recently biotech relied on endless manual lab work: measuring, pipetting, carefully preparing samples. Now simulations speed up the process

…

-insights/the-bio-revolution-innovations-transforming-economies-societies-and-our-lives. GO TO NOTE REFERENCE IN TEXT If you used traditional brute-force computation DeepMind, “AlphaFold: A Solution to a 50-Year-Old Grand Challenge in Biology,” DeepMind Research, Nov. 20, 2020, www.deepmind.com/blog

…

/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology. GO TO NOTE REFERENCE IN TEXT Mohammed AlQuraishi, a well-known researcher Mohammed AlQuraishi, “AlphaFold @ CASP13: ‘What Just Happened?,’ ” Some Thoughts on a Mysterious Universe, Dec. 9

…

, 2018, moalquraishi.wordpress.com/2018/12/09/alphafold-casp13-what-just-happened. GO TO NOTE REFERENCE IN TEXT One headline said it all Tanya Lewis, “One of the Biggest Problems in Biology Has

…

-problems-in-biology-has-finally-been-solved. GO TO NOTE REFERENCE IN TEXT The result has been an explosion Ewen Callaway, “What’s Next for AlphaFold and the AI Protein-Folding Revolution,” Nature, April 13, 2022, www.nature.com/articles/d41586-022-00997-5. GO TO NOTE REFERENCE IN TEXT DeepMind

…

Known Protein,” Financial Times, July 28, 2022, www.ft.com/content/6a088953-66d7-48db-b61c-79005a0a351a; DeepMind, “AlphaFold Reveals the Structure of the Protein Universe,” DeepMind Research, July 28, 2022, www.deepmind.com/blog/alphafold-reveals-the-structure-of-the-protein-universe. GO TO NOTE REFERENCE IN TEXT In 2019, electrodes surgically

…

Incidents Database, 246 air gaps, 241 airlines, 267–68 AlexNet, 58–59 algorithms, 63, 114, 247 Alibaba, 66 Allison, Graham, 123 Alphabet, 128, 256, 257 AlphaFold, 89–90, 109 AlphaGo, 53–54, 113, 117–19, 120 AlphaZero, 54 alternatives, 234 Altos Labs, 85 Amazon, 94–95, 189 Anduril, 166 anthrax, 174

Supremacy: AI, ChatGPT, and the Race That Will Change the World

by Parmy Olson · 284pp · 96,087 words

as much as possible. DeepMind’s had been more academic, publishing research papers about the AlphaGo gaming system and AlphaFold, a novel approach to predicting how proteins fold in the human body. AlphaFold was born out of a hackathon—or a collaborative programming event—at DeepMind in 2016, before turning into one

…

like that. When amino acids in our cells fold up into specific 3D shapes, they become proteins, and badly folded proteins can lead to diseases. AlphaFold was an AI program that predicted what those 3D shapes would look like when they folded up, and DeepMind believed that could help scientists better

…

-sourced its protein folding code to scientists in 2021. At the time of writing, more than one million researchers across the world had accessed the AlphaFold Protein Structure Database, according to DeepMind. But science is a slow process, and while Hassabis could one day still win a Nobel, a major discovery

…

be precisely designed and fully observed. That was how it built AlphaGo, by programming it to play millions of games against itself in simulation, and AlphaFold, which used simulations of protein folding. Training on real-world data—as OpenAI had done by scraping billions of words from the internet—was messy

…

, and Robert West. “Do Llamas Work in English? On the Latent Language of Multilingual Transformers.” www.arxiv.org, February 16, 2024. Chapter 13: Hello, ChatGPT “AlphaFold: The Making of a Scientific Breakthrough.” Google DeepMind’s YouTube channel, November 30, 2020. Andersen, Ross. “Does Sam Altman Know What He’s Creating?” The

…

accelerationists AI Act (Europe) AI Now Institute Airbnb AI Safety Summit Alphabet autonomous units model and China and DeepMind and valuation of See also Google AlphaFold AlphaFold Protein Structure Database AlphaGo Altman, Connie Altman, Jerry Altman, Sam AOL chat rooms and approach to AI of on bias in DALL-E 2 blog

…

coreference resolution Cotra, Ajeya Cruise DALL-E 2 D’Angelo, Adam Dartmouth College Datasheets for Datasets Dayan, Peter Dean, Jeff Deep Blue DeepMind Alphabet and AlphaFold AlphaGo and Applied ChatGPT and culture of secrecy and current state of ethical oversight board and ethics and ethics council and Facebook and formation of

The Age of AI: And Our Human Future

by Henry A Kissinger, Eric Schmidt and Daniel Huttenlocher · 2 Nov 2021 · 194pp · 57,434 words

response, working to understand and explain. A striking example of AI enabling broad-based discovery in the biological and chemical sciences is the development of AlphaFold, which used reinforcement learning to create powerful new models of proteins. Proteins are large, complex molecules that play a central role in the structure, function

…

challenge has been called protein folding. Before 2016, there had not been much progress toward improving the accuracy of protein folding—until a new program, AlphaFold, yielded major progress. As its name implies, AlphaFold was informed by the approach developers took when they taught AlphaZero to play chess. Like AlphaZero

…

reinforcement learning to model proteins without requiring human expertise—in this case, the known protein structures previous approaches relied upon. AlphaFold has more than doubled the accuracy of protein folding from around 40 to around 85 percent, enabling biologists and chemists around the world to revisit

…

old questions they had been unable to answer and to ask new questions about battling pathogens in people, animals, and plants.2 Advances like AlphaFold—impossible without AI—are transcending previous limits in measurement and prediction. The result is changes in how scientists approach what they can learn in order

…

by machines. Advances in AI for scientific discovery have also continued to accelerate. In the summer of 2021, Deep Mind released AlphaFold2, the successor to AlphaFold, which predicts the 3D structure of proteins from their amino acid sequence (see chapter 6). AlphaFold2 and work done at the Baker Lab at the

…

Future, November 17, 2020, https://workofthefuture.mit.edu/research-post/the-work-of-the-future-building-better-jobs-in-an-age-of-intelligent-machines. 2. “AlphaFold: A Solution to a 50-Year-Old Grand Challenge in Biology,” DeepMind blog, November 30, 2020, https://deepmind.com/blog/article

…

/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology. 3. See Walter Lippmann, Public Opinion (New York: Harcourt, Brace and Company, 1922), 11.

Rule of the Robots: How Artificial Intelligence Will Transform Everything

by Martin Ford · 13 Sep 2021 · 288pp · 86,995 words

at board games like Go and chess. But the era of AI being primarily associated with adeptness at games is clearly drawing to a close. AlphaFold’s ability to predict the shape of protein molecules with an accuracy that rivals expensive and time-consuming laboratory measurement using techniques like X-ray

…

drug, DeepMind has instead deployed its technology to gain understanding at a more fundamental level. In late 2018, DeepMind entered an earlier version of its AlphaFold system in a biennial global contest known as the Critical Assessment of Structure Prediction, or CASP. Teams from around the world used a variety of

…

techniques based on both computation and human intuition to attempt to predict the way proteins fold. AlphaFold won the 2018 contest by a wide margin, but even while prevailing, it was able to make the best prediction for only twenty-five of

…

the forty-three protein sequences correctly. In other words, this preliminary version of AlphaFold was not yet accurate enough to be a truly useful research tool.80 The fact that DeepMind was able to refine its technology to the

…

.nature.com/articles/d41586-020-03348-4. 2. Andrew Senior, Demis Hassabis, John Jumper and Pushmeet Kohli, “AlphaFold: Using AI for scientific discovery,” DeepMind Research Blog, January 15, 2020, deepmind.com/blog/article/AlphaFold-Using-AI-for-scientific-discovery. 3. Ian Sample, “Google’s DeepMind predicts 3D shapes of proteins,” The Guardian

…

, December 2, 2018, www.theguardian.com/science/2018/dec/02/google-deepminds-ai-program-alphafold-predicts-3d-shapes-of-proteins. 4. Lyxor Robotics and AI UCITS ETF, stock market ticker ROAI. 5. See, for example: Carl Benedikt Frey and Michael

…

, “DeepMind’s AI is getting closer to its first big real-world application,” Wired, January 15, 2020, www.wired.co.uk/article/deepmind-protein-folding-alphafold. 81. Semantic Scholar website, accessed May 25, 2020, pages.semanticscholar.org/about-us. 82. Ibid. 83. Khari Johnson, “Microsoft, White House, and Allen Institute release

The Future Is Faster Than You Think: How Converging Technologies Are Transforming Business, Industries, and Our Lives

by Peter H. Diamandis and Steven Kotler · 28 Jan 2020 · 501pp · 114,888 words

most likely distance between a protein’s base pairs and the angles of their chemical bonds—aka, the basics of protein folding. They called it AlphaFold. On its first foray into the competition, contestant AIs were given forty-three protein-folding problems to solve

…

. AlphaFold got twenty-five right. The second-place team managed a meager three. If we couple AlphaFold’s progress to Insilico’s GANs and add in the anticipated breakthroughs in quantum computing—another technology being

…

year: Reinhard Renneberg, Biotechnology for Beginners (Academic Press, 2016), p. 281. a biannual competition was created: See: http://predictioncenter.org/. AlphaFold: Read the DeepMind blog about AlphaFold here: https://deepmind.com/blog/alphafold/. Chapter Ten: The Future of Longevity The Nine Horsemen of Our Apocalypse Francis Collins: Dr. Francis Collins shared the stage

…

, Paul, 176 All Nippon Airways (ANA), 26 Alphabet, 46, 89, 162, 235 Project Loon of, 39–40 Verily Life Sciences of, 157 see also Google AlphaFold, 167 AlphaGo, 36 AlphaGo Zero, 36, 37 Alzheimer’s disease, 82, 178 Amarasiriwardena, Gihan, 108–9 Amazon, 4, 21, 47, 100, 107, 108, 114, 119

If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All

by Eliezer Yudkowsky and Nate Soares · 15 Sep 2025 · 215pp · 64,699 words

have to go through a long, drawn-out, incremental process, measured more in months than in hours, to predict… … the sort of protein folds that AlphaFold 3 can easily predict today. Google DeepMind cracked the protein folding problem between the years of 2018 and 2022 (with AIs named

…

AlphaFold 1, AlphaFold 2, and AlphaFold 3). That was the work for which Demis Hassabis, co-founder of DeepMind, won the Nobel Prize in Chemistry. Did I just get lucky with

…

was allowed to deliberately pick the easiest-to-predict proteins capable of building what it needed to build. What came to pass was: The narrow AlphaFold models were able to predict almost all biological protein folds, including the ones that humans considered quite hard to predict. I predicted that it would

…

in hours, to comprehend biochemistry? That’s what people in 2008 predicted about protein folding, a problem that humans found hard and that AlphaFold found easy. And AlphaFold is not a superintelligence; it’s not tens or hundreds of thousands of times faster than humans. An ultrafast mind wouldn’t want to

AI 2041: Ten Visions for Our Future

by Kai-Fu Lee and Qiufan Chen · 13 Sep 2021

Chapter Three: Twin Sparrows Analysis: Natural Language Processing, Self-Supervised Training, GPT-3, AGI and Consciousness, AI Education Chapter Four: Contactless Love Analysis: AI Healthcare, AlphaFold, Robotic Applications, COVID Automation Acceleration Chapter Five: My Haunting Idol Analysis: Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR), Brain-Computer Interface (BCI

…

alerting her of approaching danger. She opened her eyes and reached out for Garcia, ready for a long-overdue plastic-textured embrace. ANALYSIS AI HEALTHCARE, ALPHAFOLD, ROBOTIC APPLICATIONS, COVID AUTOMATION ACCELERATION “Contactless Love” transpires in a society transformed by an ongoing pandemic—in this case, the prospect of the COVID-19

…

. AI can greatly accelerate the speed and reduce the cost of drug and vaccine discovery. For determining protein folding (step 2), in 2020, DeepMind developed AlphaFold 2, which is AI’s greatest achievement for science to date. Proteins are the building blocks of life, yet one aspect of proteins that has

…

to carry out life’s tasks. This is a problem with profound scientific and medical implications and appears well-suited for deep learning. DeepMind’s AlphaFold, trained on a large database of previously discovered 3D protein structures, has demonstrated that it is able to simulate the 3D structure of unseen proteins

…

for each protein. For this reason, traditional methods have solved less than 0.1 percent of all proteins; thus, AlphaFold may offer a way to rapidly grow the number of solved proteins. AlphaFold has been heralded by the biology community as having solved a “fifty-year-old grand challenge in biology.” Once