The AI Daily Brief: Artificial Intelligence News and Analysis - How Clashing Egos and Disagreements Shaped the AI Field

Episode Date: December 4, 2023

From Elon Musk's fallout with Larry Page to...well...Elon Musk's fallout with Sam Altman...to Sam Altman's fallout with Dario Amodei to...Sam Altman's fallout with the OpenAI board. The history that shaped the AI space. ABOUT THE AI BREAKDOWN The AI Breakdown helps you understand the most important news and discussions in AI.  Subscribe to The AI Breakdown newsletter: https://theaibreakdown.beehiiv.com/subscribe Subscribe to The AI Breakdown on YouTube: https://www.youtube.com/@TheAIBreakdown Join the community: bit.ly/aibreakdown Learn more: http://breakdown.network/

Today on the AI breakdown, we're getting historical on some of the key moments of friction and tension that led to the landscape of leading AI labs that we have today. The AI breakdown is a daily podcast and video about the most important news and discussions in AI. Go to Breakdown.net network for more information about our YouTube, our Discord, and our newsletter. Welcome back to the AI breakdown. One of the things that is becoming clear is that there is a growing acceptance, an understanding of the history. of the inflection point moment we're living through. And this is something I don't think I have to tell lots of you who have gone and sought out content on this artificial intelligence revolution.

But what's so notable is that things that were formerly just inside baseball in Silicon Valley are now becoming seen as more significant in a way that transcends the technology industry itself. The latest example of this is an absolutely massive piece from over the weekend in the New York Times called ego, fear, and money. how the AI fuse was lit. The people who were most afraid of the risks of artificial intelligence decided they should be the ones to build it. Then distrust fueled a spiraling competition.

Now, unfortunately, I am at the moment traveling for a funeral, and so I had to record this episode before Monday when it actually comes out. But even beyond that, the sheer number of words that the NYT dedicated to this story is evidence of this shift that I was just discussing. And what's more, it contains within it some fairly significant history. So what I'm going to do is go through this story. I'm not going to read it verbatim. I will read a few excerpts, but I'll also share the gist of the narrative, because obviously it's the narrative more than the writing that matters in this case. The story starts at Elon Musk's 44th birthday party all the way back in July 2015, in a Napa Valley resort. At the heart of a story is a conversation between Elon and Google co-founder Larry Page.

The Times writes, AI was the big topic of conversation when Mr. Musk and Mr. Page sat down near a fire pit beside a swimming pool after dinner the first night. The two billionaires had been friends for more than a decade, and Mr. Musk sometimes joked that he occasionally crashed on Mr. Page's sofa after a night playing video games. But the tone that clear night soon turned contentious, as the two debated whether artificial intelligence would ultimately elevate humanity or destroy it. As the discussion stretched into the chilly hours, it grew intense, and some of the more than 30 partiers gathered closer to listen. Mr. Page hampered for more than a decade by an unusual ailment in his vocal cords, described his vision of a digital utopia in a whisper.

Humanity would eventually merge with artificially intelligent machines, he said. One day there would be many kinds of intelligence competing for resources and the best would win. If that happens, Mr. Musk said, we're doomed. The machines will destroy humanity. With a rasp of frustration, Mr. Page insisted his utopia should be pursued. Finally, he called Mr. Musk a speciest, a person who favors humans over the digital life forms of the future. That insult, Mr. Musk said later, was the last straw. Many in the crowd seemed gobsmacked if amused as they dispersed for the night and considered it just one of those esoteric debates that often break out at Silicon Valley parties.

But eight years later, the argument between the two men seems prescient. The question of whether artificial intelligence will elevate the world or destroy it, or at least inflict grave danger, has framed an ongoing debate among Silicon Valley founders, chatbot users, academics, legislators and regulators about whether the technology should be controlled or set free. And then the times comes to the quick of it. They write, at the heart of this competition is a brain-stretching paradox. The people who say they are the most worried about AI are among the people most determined to create it and enjoy its riches. They have justified their ambition with their strong belief that they alone can keep AI from endangering Earth. So as many of you know, this was not some idle conversation, but had pretty dramatic fallout. First of all, Elon and Larry Page are effectively no longer friends. They basically

stopped speaking after that party. Second of all, it was this catalytic event that would lead to the creation of Open AI. As the Times characterizes it, a few weeks later, Mr. Musk dined with Mr. Altman, who was then running a tech incubator, which was, of course, Y Combinator editor's note, and several researchers in a private room at the Rosewood Hotel in Menlo Park, California, a favorite deal-making spot close to the venture capital offices of Sand Hill Road. Now, of course, this fallout between Elon and Larry was not the only fallout in the AI safety story. Elon and Sam Altman also don't talk anymore. Said Altman at one point, there is disagreement, mistrust, egos. The closer people are to being pointed in the same direction, the more contentious the disagreements are. You see this in sex and

religious orders. There are bitter fights between the closest people. Now, of course, a couple weeks ago, just before Thanksgiving, we saw another crazy near implosion in the AI space, which was, of course, the attempt to get Sam Altman out of Open AI. But as the Times points out, that was far from the first big fight with the potential to shape the future of the AI industry. From there, the story moves back all the way to 2010, and a meeting between then-34-year-old Demis Hes Hesabas and leading SFBC, Peter Thiel. From The Times, in 2010, Dr. Hesabas and two colleagues who all lived in Britain, were looking for money to start building artificial general intelligence or AGI, a machine that could do anything the brain could do. At the time,

few people were interested in AI. After a half century of research, the artificial intelligence field had failed to deliver anything remotely close to the human brain. Still, some scientists and thinkers have become fixated on the downsides of AI. Many, like the three young men from Britain, had a connection to Eliezer Yudkowski, an internet philosopher and self-taught AI researcher. Mr. Yudkowski was a leader in a community of people who called themselves rationalists or, in later years, effective altruists. They believed that AI could find a cure for cancer or solve climate change, but they

worried that AI bots might do things their creators had not intended. If the machines became more intelligent than humans, the rationalists argued, the machines could turn on their creators. The Times discusses how Peter Thiel became rich, which is the story we all know, but also writes, he had developed a fascination with the singularity, a trope of science fiction that describes the moment when intelligent technology can no longer be controlled by humanity. With funding from Mr. Teal, Mr. Yutkowski had expanded his AI lab and created an annual conference on the singularity. Years before, one of Dr. Hussabas's two colleagues had met Mr. Yutkowski, and he snagged them speaking spots at the

conference ensuring they'd be invited to Mr. Teal's party. Mr. Yudkowski introduced Dr. Hizabas to Mr. Teal. Dr. Hizabas assumed that lots of people at the party would be trying to squeeze their host for money. His strategy was to arrange another meeting. The next day in Teal's kitchen, he committed to put the first $2.25 million into their startup, which was called DeepMind. Writes the Times, they wholeheartedly believe that because they understood the risks, they were uniquely positioned to protect the world. Said Mustafa Sullyman, one of three DeepMind founders, I don't see this as a contradictory position. There are huge benefits to come from these technologies. The goal is not to eliminate them or pause their development. The goal is,

is to mitigate the downsides. From there, the story goes on to how Demis met Elon Musk. Hasabas had come to tour the SpaceX headquarters, and they were eating lunch afterwards when, as the Times writes, Mr. Musk explained that his plan was to colonize Mars to escape over population and other dangers on Earth. Dr. Hasabas replied that the plan would work so long his superintelligent machines didn't follow and destroy humanity on Mars, too. Mr. Musk was speechless. He hadn't thought about that particular danger. Mr. Musk soon invested in Deep Mind alongside Mr. Teal, so he could be closer to the creation of this technology. From there, the story moves into what Deep Mind built, a system that was able to play classic Atari games like Space Invaders, Pong, and Breakout as an illustration of what could

be possible in the future. And thus we get back to another central actor in our story, Larry Page. This section I'm going to read in a little bit more detail, because it involves Jeffrey Hinton, who has, of course, become one of the central voices in the AI safety conversation that has come to dominate mainstream media this year. The section is called the talent auction, and it begins. In the fall of 2012, Jeffrey Hinton, a 64-year-old professor at the University of Toronto, and two graduate students published a research paper that showed the world what AI could do. They train a neural network to recognize common objects like flowers, dogs, and cats. Scientists were surprised by the accuracy of the technology built by Dr. Hinton and his students.

One who took particular notice was Yu Kai, an AI researcher who had met Dr. Hinton at a research conference, and had recently started working for Baidu, the giant Chinese internet company. Baidu offered Dr. Hinton and his students $12 million to join the company in Beijing. Dr. Hinton turned Baidu down, but the money got his attention. The Cambridge-educated British expatriate had spent most of his career in academia, except for occasional stints at Microsoft and Google, and was not especially driven by money. But he had a neurodivergent child, and the money would mean financial security. We did not know how much we were worth, Dr. Hinton said.

He consulted lawyers and experts on acquisitions and came up with a plan. We would organize an auction, and we would sell ourselves. The auction would take place during an annual AI conference at the Harris Hotel and Casino on Lake Tahoe. Big Tech took notice. Google, Microsoft, Baidu, and other companies were beginning to believe that neural networks were a path to machines that could not only see, but hear, write, talk, and eventually, think. Mr. Page had seen similar technology at Google Brain, his company's AI lab, and he thought Dr. Hinton's research could elevate his scientist's work.

He gave Alan Eustis, Google's senior vice president of engineering, what amounted to a blank check to hire any AI expertise he needed. Mr. Eustace and Jeff Dean, who led the brain lab, flew to Lake Tahoe and took Dr. Hinton and his students out to dinner at a steakhouse inside the hotel the night before the auction. They made the case for coming to work at Google. The next day, Dr. Hinton ran the auction from his hotel room. Because of an old back injury, he rarely sat down.

He turned a trash can upside down on a table, put his laptop on top, and watched the bids roll in over the next two days. Google made an offer. So did Microsoft. DeepMind quickly bowed out as the price went up. The industry giants pushed the bids to $20 million and then $25 million, according to documents detailing the auction. As the price passed $30 million, Microsoft quit, but it rejoined the bidding at $37 million.

Then Microsoft dropped out a second time. Only Baidu and Google were left, and they pushed the bidding to $42 million, $43 million. Finally, at $44 million, Dr. Hinton and his students stopped the auction. The bids were still climbing, but they wanted to work for Google, and the money was staggering. It was an unmistakable sign that deep-pocketed companies

were determined to buy the most talented AI researchers, which was not lost on Dr. Hussabas at Deep Mind. He had always told his employees that DeepMind would remain an independent company. That was, he believed, the best way to ensure its technology didn't turn into something dangerous. But as Big Tech entered the talent race, he decided he had no choice. It was time to sell. So fast forward to the end of 2012, and we've got a situation where both Google and Facebook are trying to get DeepMind to come to them. According to people who were around at the time, the DeepMind founders had two conditions.

One, that DeepMind couldn't be used for military purposes, and that, two, any AGI technology would be overseen by an independent board of technologists and ethicists. Google offered 650 million and agreed to the conditions. Zuckerberg offered even more, but would not agree to the conditions. And thus, DeepMind went to Google. Now, at this point, you may be wondering what happened to that ethics board. Well, that is the subject of the next section. Quote, when Mr. Musk invested in DeepMind, he broke his own informal rule, that he would not invest in any company he didn't run himself. The downsides of his decision were already apparent when, only a month or so after his birthday spat with Mr. Page, he again found himself face to face with his former friend and fellow billionaire. The occasion was the first meeting of DeepMind's

Ethics Board on August 14, 2015. The board had been set up at the insistence of the startup's founders to ensure their technology did no harm after the sale. The members convened in a conference room just outside Mr. Musk's office at SpaceX, with a window looking out onto his rocket factory. But that's where Mr. Musk's control ended. When Google bought DeepMind, it bought the whole thing. Mr. Musk was out. Financially, he had come out ahead, but he was unhappy. Now, the attendees of that first meeting included Larry Page and Sergey Brin, his co-founder at Google, as well as Eric Schmidt, who was then Google's chairman, then was LinkedIn and PayPal founder, Reid Hoffman, as well as Toby Ord, who was an Australian philosopher studying existential risk.

Apparently at that first meeting, Mustafa Sullyman gave a presentation called The Pitchforks Are Coming. He talked about all the different risks he saw, including an explosion of disinformation, the replacement of huge numbers of jobs, the risk that people would accuse Google of stealing their livelihoods, and even a pitch for Google to help provide a universal basic income for people who couldn't work anymore. Elon was reportedly in agreement, but others there were not. Eric Schmidt said he thought the worries were totally overblown, and Larry Page agreed saying AI would create way more jobs than it took away. Eight months after that was when DeepMind's AlphaGo made headlines by beating one of the world's best players at Go,

which was one of these moments where the timeline of AI seemed to get a lot faster than people had previously thought. writes the Times, DeepMind's founders were increasingly worried about what Google would do with their inventions. In 2017, they tried to break away from the company. Google responded by increasing the salaries and stock award packages of the DeepMind founders and their staff. They stayed put. The Ethics Board never had a second meeting. And this brings us to OpenAI.

The Times recounts how, born out of frustration with all of this, Elon along with others like Reed Hoffman and Peter Thiel, had pledged a billion dollars to this new venture which Sam Altman would run. From a capacity perspective, they pulled in Ilya Sutskever from Google, who was, of course, one of the students that had been bought in Dr. Hinton's auction. And this is where they begin recounting the difficulties in structure that were there from the very beginning of this project. Writes the Times, initially Mr. Musk wanted to operate Open AI as a non-profit, free from the economic incentives that were driving Google and other corporations. But by the time Google wowed the tech community with its go stunt, Mr. Musk was changing his mind about how it should be run. He desperately wanted Open AI to invent something that would capture the world's imagination.

and close the gap with Google, but it wasn't getting the job done as a nonprofit. In late 2017, he hatched a plan to rest control of the lab from Mr. Altman and the other founders and transform it into a commercial operation that would join forces with Tesla and rely on supercomputers the car company was developing. When Mr. Altman and others pushed back, Mr. Musk quit and said he would focus on his own AI work at Tesla. In February 2018, he announced his departure to OpenAI staff on the top floor of the startup's offices in a converted truck factory, and when he said that OpenAI needed to move faster, one researcher retorted that Mr. Musk was being reckless. Mr. Musk called the researcher a jackass and stormed out, taking his deep pockets

with him. And this, of course, is where Microsoft enters the story. Quote, OpenAI suddenly needed new financing in a hurry. Mr. Altman flew to Sun Valley for a conference and ran into Satya and Della, a tie-up seemed natural. Mr. Altman and OpenAI had formed a for-profit company under the original nonprofit. They had one billion in fresh capital, and Microsoft had a new way to build artificial intelligence into its vast cloud computing service. But again, we reach another breaking point. Next up, of course, was the break between Open AI and Anthropic. Writes at times, not everyone inside open AI was happy. Dario Amade, a researcher with ties to the effect of altruist community, had been on hand at the Rosewood Hotel when Open AI was born. He was leading the lab's efforts to

build a neural network, called a large language model that could learn from enormous amounts of digital text. By analyzing countless Wikipedia articles, digital books, and message boards, it could generate text on its own. It also had the unfortunate habit of making things up. He was called GPT3 and it was released in the summer of 2020. Ammodai was unhappy with the Microsoft deal because he thought it was taking Open AI in a really commercial direction. He and other researchers went to the board to try to push Mr. Altman out, according to five people familiar with the matter. After they failed, they left. Like DeepMinds founders before them, they worried their new corporate overlords would favor commercial interest over safety. In 2021, a group of about 15 engineers and

scientists created a new lab called Anthropic. The plan, was to build AI the way that effective altruist thought it should be done, with very tight controls. Now, this part of the history is a little bit disputed. Said an Anthropic spokesperson, there was no attempt to remove Sam Altman from OpenAI by the co-founders of Anthropic. The co-founders themselves came to the conclusion that they wished to depart Open AI to start their own company, made this known to Open AI's leadership and over several weeks negotiated an exit on mutually agreeable terms. Using a great convention of telling the narrative through key meetings, the next story in the piece is about a set of meetings

between Altman and other OpenAI leaders like Greg Brockman and Bill Gates. In the first meeting, Gates told them about his skepticism of LLMs. He would stay skeptical, he said, until the technology performed a task that required critical thinking, passing an AP biology test, for instance. Then, of course, the second meeting was five months later, August 24, 2022, where they showed Gates GPT4 and gave that system a multiple choice AP biology test. The first question involved polar molecules, groups of atoms with a positive charge at one end and a negative charge at the other. The system answered correctly and explained its choice. Mr. Brockman said, it was only trained to provide an answer. The conversational nature kind of fell out almost

magically. In other words, it was doing things they hadn't really designed it to do. There were 60 questions. GPT4 only got one answer wrong. Mr. Gates sat up in his chair, his eyes opened wide. In 1980, he had a similar reaction when researchers showed him the graphical user interface that became the basis for the modern personal computer. He thought GPT was that revolutionary. Now that is where the New York Times story ends, but obviously we we can pick up the pieces from there, and even for as much light as the sheds, it barely scratches the surface on how messy and complex and contradictory this space really is. So by way of trying to wrap up, let's start to just pull apart the threads here.

One is this divide that we see now playing out between the EA's and the EACCs of people who are concerned about runaway AI and people who think that those concerns are way overblown. This is one fault line and dividing line between some of the parties involved. However, among the group of people who are concerned, there are also big fractures, which in this story is exemplified by Anthropic Split from OpenAI, but of course is a sub-theme running through basically all the conversations around AI safety that happen. A third really interesting theme is this, if I don't do it right, someone else will do it wrong kind of thinking. You can feel it basically permeating all of these teams from DeepMind to Open AI to Musk. And so even among the people who have these concerns about the future, they've ended up being the biggest progenitors of the AI arms race in many ways. Now, of course, part of that is because looming over them

are their big tech benefactors and patrons, the only companies with the scale of compute and the resources needed to actually allow these labs to grow. The numbers to compete in this space are simply outside the scale of even extremely well-funded venture capitalists. There's just too much money needed. And so the story so far is one of constant compromise and little acquiescences to financial realities that lead people away from the principles with which they started. Now, one of my favorite movies of all time, and a movie that I think is wildly underrated is SLC Punk. And there's a line at the end where Stivo, the lead punk, has decided to go to Harvard rather than waste his life away as a nihilist, because, as he puts it, you can do a hell of a lot

more damage inside the system than outside. I don't think, in other words, even for those who do have these deep concerns around the future of AI, but being inside these labs that basically have all the power is a bad approach. Trying to wrench them to the good from the inside is a coherent strategy. Of course, ultimately the question will be how much those internal forces can compete ultimately with the incredible external pressures. Anyways, we will wrap this recent historical journey here. I think it's fascinating and it's clearly relevant for today. So I appreciate you hanging out and listening.

We'll be back tomorrow with our normal brief then main episode format. Until then, peace.