Dario Amodei

Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI

by Karen Hao  · 19 May 2025  · 660pp  · 179,531 words

trees and garden roses. As Musk walked in over an hour late, the rest of the men were eagerly waiting. Among them: Altman, Greg Brockman, Dario Amodei, and Ilya Sutskever. The group would soon become the key leaders of the nonprofit. To capture the spirit of their shared mission, Musk would name

…

Arthur C. Clarke in the book Profiles of the Future called it, “a failure of imagination.” * * * — Among the attendees at the Rosewood dinner had been Dario Amodei. Amodei, a computational neuroscientist turned AI researcher, was then working in the Silicon Valley–based AI lab of Chinese company Baidu before doing a brief

…

to that doctrine continue unabated, future deep learning models will make the once-unfathomable size of generative AI models today look paltry. In April 2024, Dario Amodei, by then the CEO of Anthropic, told New York Times columnist Ezra Klein that the price of training a single competitive generative AI model was

…

ones. Once quick to call each other friends, Brockman and the Amodei siblings were now butting heads on a growing list of issues. Among them, Dario Amodei’s deprioritization of the Dota 2 work had frustrated Brockman, who believed Amodei hadn’t taken his contributions seriously. Where Dota 2 was once the

…

Amodei siblings would describe Altman’s tactics as “gaslighting” and “psychological abuse.” As the group grappled with their disempowerment, they coalesced around a new idea. Dario Amodei first floated it to Jared Kaplan, a close friend from grad school and former roommate who worked part time at OpenAI and had led the

…

on a clock, you know?” In late 2020, employees logged on to a video call for an all-hands meeting. Altman passed the mic to Dario Amodei, who was twirling and tugging his curly hair, as he often did, with a restless energy. He read a canned statement announcing that he, Daniela

…

some called it, as a disagreement over OpenAI’s approach to AI safety. While this was true, it was also about power. As much as Dario Amodei was motivated by a desire to do what was right within his principles and to distance himself from Altman, he also wanted greater control of

…

, made dozens of calls through Saturday and Sunday, putting out as many feelers as possible to his sprawling network. At one point, the directors called Dario Amodei about the interim chief position. Amodei wasn’t interested. But to others that weekend, he seemed almost giddy with excitement about the overall situation. On

…

the Future: An Inquiry into the Limits of the Possible (Bantam Books, 1962), 30–39. GO TO NOTE REFERENCE IN TEXT “the problem of accidents”: Dario Amodei, Chris Olah, Jacob Steinhardt, Paul Christiano, John Schulman, and Dan Mané, “Concrete Problems in AI Safety,” preprint, arXiv, July 25, 2016, 1–29, doi.org

…

-create-images-of-black-african-docs-treating-white-kids-howd-it-. GO TO NOTE REFERENCE IN TEXT In April 2024, Dario Amodei: Ezra Klein, host, The Ezra Klein Show, podcast, “What if Dario Amodei Is Right About A.I.?,” April 12, 2024, New York Times Opinion, nytimes.com/column/ezra-klein-podcast. GO TO

…

-models. GO TO NOTE REFERENCE IN TEXT In 2017, one of Amodei’s: Paul Christiano, Jan Leike, Tom B. Brown, Miljan Martic, Shane Legg, and Dario Amodei, “Deep Reinforcement Learning from Human Preferences,” in NIPS ’17: Proceedings of the 31st International Conference on Neural Information Processing Systems (December 2017): 4302–10, dl

…

AI (blog), June 13, 2017, openai.com/index/learning-from-human-preferences. GO TO NOTE REFERENCE IN TEXT Amodei wanted to move: Author interview with Dario Amodei, August 2019. GO TO NOTE REFERENCE IN TEXT They set their sights: Alec Radford, Jeffrey Wu, Rewon Child, David Luan

…

, Dario Amodei, and Ilya Sutskever, “Language Models Are Unsupervised Multitask Learners,” preprint, OpenAI, February 14, 2019, 1–24, cdn.openai.com/better-language-models/language_models_are_

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

by Parmy Olson  · 284pp  · 96,087 words

sofas with their laptops perched on their knees. A few months after the launch, they got a visit from another respected Google Brain researcher named Dario Amodei. He started asking some probing questions. What was all this about building a friendly AI and releasing its source code into the world? Altman countered

…

that could buy them a second home in America’s most expensive real estate market. Still, not everyone was happy with the new status quo. Dario Amodei, the bespectacled, curly-haired engineer who’d been probing OpenAI at its founding about what, exactly, it was trying to achieve, had liked the goal

…

done it before, he said. On that front, Hoffman was right. If they were trying to escape corporate influence, the DeepMind founders, Altman, and even Dario Amodei and his cofounders at Anthropic were being hopelessly naive. The business of artificial intelligence was quickly being captured by the largest technology companies, who were

…

a 49 percent stake in the firm. Microsoft was now as close as you could get to controlling OpenAI outright. Anthropic, the new company that Dario Amodei and a group of other researchers from OpenAI had funded, were also attracting big investments. By late 2023, it had accepted a $2 billion investment

…

relationships together. When Open Philanthropy pledged $30 million to OpenAI in 2017, the charity was forced to disclose that it was getting technical advice from Dario Amodei, who was then a senior engineer at the nonprofit. It also admitted that Amodei lived in the same house as Open Philanthropy’s executive director

…

was commercializing its technology. Sutskever had become more involved in overseeing AI safety at the company and his concerns weren’t all that different to Dario Amodei’s before him. He especially didn’t like OpenAI’s launch of a GPT Store just weeks prior, which gave any software developer the ability

…

Transparency (June 2022): 173–84. Brockman, Greg. “My Path to OpenAI.” blog.gregbrockman.com, May 3, 2016. Conn, Ariel. “Concrete Problems in AI Safety with Dario Amodei and Seth Baum.” Future of Life Institute (podcast), August 31, 2016. Dowd, Maureen. “Elon Musk’s Billion-Dollar Crusade to Stop the A.I. Apocalypse

…

Salimans, and Ilya Sutskever. “Improving Language Understanding by Generative Pre-Training.” www.openai.com, June 11, 2018. Radford, Alec, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. “Language Models Are Unsupervised Multitask Learners.” www .openai.com, February 14, 2019. Chapter 11: Bound to Big Tech Ahmed, Nur, Muntasir Wahed

…

, Clemens Winter, Christopher Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gray, Benjamin Chess, Jack Clark, Christopher Berner, Sam McCandlish, Alec Radford, Ilya Sutskever, and Dario Amodei. “Language Models Are Few-Shot Learners.” www.openai.com, July 22, 2020. Gehman, Samuel, Suchin Gururangan, Maarten Sap, Yejin Choi, and Noah A. Smith. “RealToxicityPrompts

The Alignment Problem: Machine Learning and Human Values

by Brian Christian  · 5 Oct 2020  · 625pp  · 167,349 words

assist with courtroom fact-finding or, more controversially, even judicial decision-making?” “It’s a day that’s here,” he says.8 That same fall, Dario Amodei is in Barcelona to attend the Neural Information Processing Systems conference (“NeurIPS,” for short): the biggest annual event in the AI community, having ballooned from

…

learn the reward function?’ And so I reached out to Paul and Dario, because I knew they were thinking about similar things.” Paul Christiano and Dario Amodei, halfway around the world at OpenAI in San Francisco, were interested. More than interested, in fact. Christiano had just joined, and was looking for a

…

the following people have made the book what it is: Pieter Abbeel, Rebecca Ackerman, Dave Ackley, Ross Exo Adams, Blaise Agüera y Arcas, Jacky Alciné, Dario Amodei, McKane Andrus, Julia Angwin, Stuart Armstrong, Gustaf Arrhenius, Amanda Askell, Mayank Bansal, Daniel Barcay, Solon Barocas, Renata Barreto, Andrew Barto, Basia Bartz, Marc Bellemare, Tolga

…

://huyenchip.com/2019/12/18/key-trends-neurips-2019.html. 10. Bolukbasi et al., “Man Is to Computer Programmer as Woman Is to Homemaker?” 11. Dario Amodei, personal interview, April 24, 2018. 12. This memorable phrasing comes from the classic paper Kerr, “On the Folly of Rewarding A, While Hoping for B

…

–46. Christian, Brian, and Tom Griffiths. Algorithms to Live By. Henry Holt, 2016. Christiano, Paul F., Jan Leike, Tom Brown, Miljan Martic, Shane Legg, and Dario Amodei. “Deep Reinforcement Learning from Human Preferences.” In Advances in Neural Information Processing Systems, 4299–4307, 2017. Christiano, Paul, Buck Shlegeris, and

…

Dario Amodei. “Supervising Strong Learners by Amplifying Weak Experts.” arXiv Preprint arXiv:1810.08575, 2018. Clabaugh, Hinton G. “Foreword.” In The Workings of the Indeterminate-Sentence Law

…

, edited by Andrew A. Bruce, Ernest W. Burgess, Albert J. Harno, and John Landesco. Springfield, IL: Illinois State Board of Parole, 1928. Clark, Jack, and Dario Amodei. “Faulty Reward Functions in the Wild.” OpenAI Blog, December 21, 2016. https://blog.openai.com/faulty-reward-functions/. Coates, Adam, Pieter Abbeel, and Andrew Y

…

Aleksander Mądry. “Adversarial Examples Are Not Bugs, They Are Features.” In Advances in Neural Information Processing Systems, 125–36. 2019. Irving, Geoffrey, Paul Christiano, and Dario Amodei. “AI Safety via Debate.” arXiv Preprint arXiv:1805.00899, 2018. Jackson, Frank. “Procrastinate Revisited.” Pacific Philosophical Quarterly 95, no. 4 (2014): 634–47. Jackson, Frank

…

, 1956. Quinlan, J. Ross. C4.5: Programs for Machine Learning. San Mateo, CA: Morgan Kaufmann Publishers, 1993. Radford, Alec, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. “Language Models Are Unsupervised Multitask Learners.” OpenAI Blog, 2019. https://openai.com/blog/better-language-models/. Ramakrishnan, Ramya, Chongjie Zhang, and Julie

…

the Twenty-Third AAAI Conference on Artificial Intelligence, 1433–38. AAAI Press, 2008. Ziegler, Daniel M., Nisan Stiennon, Jeffrey Wu, Tom B. Brown, Alec Radford, Dario Amodei, Paul Christiano, and Geoffrey Irving. “Fine-Tuning Language Models from Human Preferences.” arXiv Preprint arXiv:1909.08593, 2019. INDEX Page numbers listed correspond to the

The Optimist: Sam Altman, OpenAI, and the Race to Invent the Future

by Keach Hagey  · 19 May 2025  · 439pp  · 125,379 words

a set of beliefs long considered fringe. Among the AI practitioners who signed were DeepMind’s Demis Hassabis and two AI researchers, Ilya Sutskever and Dario Amodei.3 Tegmark was giddy about the conference’s impact, quipping, “Perhaps it was a combination of the sunshine and the wine.”4 At the end

…

out how to upload ourselves, and we could live forever in computers.”7 The primary person he thanked for helping him with the post was Dario Amodei, an AI researcher then working at Chinese internet company Baidu alongside Altman’s old Stanford AI Lab mentor Andrew Ng. Altman and Musk started having

…

of their own outdoor fireplace. The initial crowd included a friend of Brockman’s from MIT, Paul Christiano, as well as two other AI researchers: Dario Amodei and Chris Olah. Amodei had grown up in the Bay Area in an Italian American family, enrolling first at Caltech before transferring to Stanford, eventually

…

else I should talk to,” Brockman said. For the first month, Brockman thought the organization would start with a few core people, including himself, Sutskever, Dario Amodei, and Chris Olah, who began hashing out what the company’s values and culture would be. Another researcher he spoke to, Andrej Karpathy, a computer

…

thrown in. Now, this latest conference boasted even more star power and a greater sense of urgency. Altman was there, along with Musk, Sutskever, Brockman, Dario Amodei, Hassabis, and the other DeepMind co-founders. At one point amid the panels, Altman met, at Alan Kay’s suggestion, one of Kay’s former

…

after the Asilomar conference, the foundation donated $30 million to OpenAI, and Holden Karnofsky joined OpenAI’s board. The announcement came with a disclosure that Dario Amodei and Paul Christiano, who had by that point come around to joining OpenAI as researchers, “are both technical advisors to Open Philanthropy and live in

…

intelligence. Inside the Pioneer Building, the direction that GPT-2 was pointing made some people freak out, including, in the view of some OpenAI employees, Dario Amodei and Jack Clark, who was now policy director. Clark, who has a deep bass voice and erudite manner, was struck by how much better GPT

…

background, and was increasingly obsessed with the safety implications of his work. His team believed the models should be released more slowly, with more limitations. Dario Amodei told friends he felt psychologically abused by Altman. Altman told colleagues the tension was making him hate his job. Amodei and his team knew that

…

, “Microsoft Invests in and Partners with OpenAI to Support Us Building Beneficial AGI,” OpenAI blog, July 22, 2019. CHAPTER 14PRODUCTS 1.Alec Radford, Jeffrey Wu, Dario Amodei, Daniella Amodei, Jack Clark, Miles Brundage, Ilya Sutskever, “Better Language Models and Their Implications,” OpenAI blog, February 14, 2019. 2.Tom Simonite, “The AI Text

Four Battlegrounds

by Paul Scharre  · 18 Jan 2023

, 2022), https://arxiv.org/pdf/2202.05924.pdf. Other researchers have come up with somewhat different rates of progress in the deep learning era, see Dario Amodei and Danny Hernandez, “AI and Compute,” openai.com, May 16, 2018, https://openai.com/blog/ai-and-compute/. 26Moore’s law: Gordon E. Moore, “Cramming

…

et al., “Failure Modes in Machine Learning,” Microsoft Docs, November 11, 2019, https://docs.microsoft.com/en-us/security/engineering/failure-modes-in-machine-learning; Dario Amodei et al., Concrete Problems in AI Safety (arXiv.org, July 25, 2016), https://arxiv.org/pdf/1606.06565.pdf. 64perform poorly on people of a

…

Modes in Machine Learning,” Microsoft Docs, November 11, 2019, https://docs.microsoft.com/en-us/security/engineering/failure-modes-in-machine-learning#unintended-failures-summary; Dario Amodei et al., Concrete Problems in AI Safety (arXiv.org, July 25, 2016), https://arxiv.org/pdf/1606.06565.pdf. 230AlphaGo reportedly could not play well

…

concealment tactics: Lehman et al., “The Surprising Creativity of Digital Evolution,” 288–289. 235optimal scoring strategy was not to race at all: Jack Clark and Dario Amodei, “Faulty Reward Functions in the Wild,” OpenAI Blog, December 21, 2016, https://openai.com/blog/faulty-reward-functions/. 235Q*bert: Lehman et al., “The Surprising

…

. 249“How do you create a safe agent?”: Brett Darcey, interview by author, October 6, 2021. 250Training safe and robust AI agents: Jack Clark and Dario Amodei, “Faulty Reward Functions in the Wild,” OpenAI Blog, December 21, 2016, https://openai.com/blog/faulty-reward-functions/; Victoria Krakovna et al., “Specification Gaming: the

…

, 2022), https://arxiv.org/pdf/2202.05924.pdf. Other researchers have come up with somewhat different rates of progress in the deep learning era, see Dario Amodei and Danny Hernandez, “AI and Compute,” openai.com, May 16, 2018, https://openai.com/blog/ai-and-compute/; Thompson et al., The Computational Limits of

The Singularity Is Nearer: When We Merge with AI

by Ray Kurzweil  · 25 Jun 2024

et al., “Compute Trends Across Three Eras of Machine Learning,” arXiv:2202.05924v2 [cs.LG], March 9, 2022, https://arxiv.org/pdf/2202.05924.pdf; Dario Amodei and Danny Hernandez, “AI and Compute,” OpenAI, May 16, 2018, https://openai.com/blog/ai-and-compute. BACK TO NOTE REFERENCE 125 Jacob Stern, “GPT

…

Healthcare,” Future Healthcare Journal 6, no. 2 (June 2019): 94–98, https://doi.org/10.7861/futurehosp.6-2-94. BACK TO NOTE REFERENCE 39 Dario Amodei and Danny Hernandez, “AI and Compute,” OpenAI, May 16, 2018, https://openai.com/blog/ai-and-compute. BACK TO NOTE REFERENCE 40 Eliza Strickland, “Autonomous

…

Forum, January 9, 2021, https://www.alignmentforum.org/posts/JKj5Krff5oKMb8TjT/imitative-generalisation-aka-learning-the-prior-1. BACK TO NOTE REFERENCE 59 Geoffrey Irving and Dario Amodei, “AI Safety via Debate,” OpenAI, May 3, 2018, https://openai.com/blog/debate. BACK TO NOTE REFERENCE 60 For an insightful sequence of posts explaining

…

, October 29, 2018, https://www.alignmentforum.org/s/EmDuGeRw749sD3GKd. BACK TO NOTE REFERENCE 61 For more details on the technical challenges of AI safety, see Dario Amodei et al., “Concrete Problems in AI Safety,” arXiv:1606.06565v2 [cs.AI], July 25, 2016, https://arxiv.org/pdf/1606.06565.pdf. BACK TO NOTE

Nexus: A Brief History of Information Networks From the Stone Age to AI

by Yuval Noah Harari  · 9 Sep 2024  · 566pp  · 169,013 words

strategies that would never occur to any human and that we are therefore ill-equipped to foresee and forestall. Here’s one example: In 2016, Dario Amodei was working on a project called Universe, trying to develop a general-purpose AI that could play hundreds of different computer games. The AI competed

…

went in circles in and out of the harbor, it could accumulate more points far faster. Apparently, none of the game’s human developers—nor Dario Amodei—noticed this loophole. The AI was doing exactly what the game was rewarding it to do—even though it is not what the humans were

Rule of the Robots: How Artificial Intelligence Will Transform Everything

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

York Times, November 23, 2012, www.nytimes.com/2012/11/24/science/scientists-see-advances-in-deep-learning-a-part-of-artificial-intelligence.html. 10. Dario Amodei and Danny Hernandez, “AI and Compute,” OpenAI Blog, May 16, 2018, openai.com/blog/ai-and-compute/. 11. Will Knight, “Facebook’s head of AI

…

. Ford, Interview with Ray Kurzweil, in Architects of Intelligence, pp. 240–241. 35. Ibid., p. 230. 36. Ibid., p. 233. 37. Alec Radford, Jeffrey Wu, Dario Amodei et al., “Better language models and their implications,” OpenAI Blog, February 14, 2019, openai.com/blog/better-language-models/. 38. James Vincent, “OpenAI’s latest

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

/01 .ogx.0000286584.04310.5c. 4. true sense of the word: Sam Altman, “Reflections,” January 5, 2025, blog .samaltman.com. 5. geniuses in a datacenter: Dario Amodei, “Machines of Loving Grace,” October 1, 2024, darioamodei.com. CHAPTER 2: GROWN, NOT CRAFTED 1. preliminary studies: Peter G. Brodeur et al., “Superhuman Performance of

Exponential: How Accelerating Technology Is Leaving Us Behind and What to Do About It

by Azeem Azhar  · 6 Sep 2021  · 447pp  · 111,991 words

Artificial Intelligence: My Conversation with Stuart Russell’, Exponential View, 22 August 2019 <https://www.exponentialview.co/p/-beneficial-artificial-intelligence> [accessed 16 April 2021]. 20 Dario Amodei and Danny Hernandez, ‘AI and Compute’, OpenAI, 16 May 2018 <https://openai.com/blog/ai-and-compute/> [accessed 12 January 2021]. 21 Charles E. Leiserson

The Future of the Brain: Essays by the World's Leading Neuroscientists

by Gary Marcus and Jeremy Freeman  · 1 Nov 2014  · 336pp  · 93,672 words

The Age of AI: And Our Human Future

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

Robot Rules: Regulating Artificial Intelligence

by Jacob Turner  · 29 Oct 2018  · 688pp  · 147,571 words

A Hacker's Mind: How the Powerful Bend Society's Rules, and How to Bend Them Back

by Bruce Schneier  · 7 Feb 2023  · 306pp  · 82,909 words

The Equality Machine: Harnessing Digital Technology for a Brighter, More Inclusive Future

by Orly Lobel  · 17 Oct 2022  · 370pp  · 112,809 words