The AI Daily Brief: Artificial Intelligence News and Analysis - AI and the Turning Point Moment in Human History

Episode Date: August 12, 2023

A reading of and discussion around Tyler Cowen's "Existential risk, AI, and the inevitable turn in human history." 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 reading a set of pieces by economist and thinker Tyler Cowan on AI, existential risk, and the inevitable turn in human history. The AI breakdown is a daily podcast and video about the most important news and discussions in AI. Go to Breakdown.network for more information about our newsletter, our Discord, and our YouTube channel. Hello, friends, welcome back to the AI breakdown. We are here with another long reads for your weekend. And I felt like since we had done a bunch of coverage over the last couple of

days given that new AIPI survey about how Americans are just terrified of AI, that it might be good to share a counter perspective to some of those concerns, and maybe one from a slightly orthogonal voice, let's say, to the AI industry itself. For that, we turn to Tyler Cowan. Tyler Cowan is a professor of economics at George Mason University. He's a Bloomberg columnist, but he's probably best known for his blog Marginal Revolution. The first piece of Tyler's that we're going to read today was published on March 27th. It is called existential risk, AI, and the inevitable turn in human history. Tyler writes, In several of my books and many of my talks, I take great care to spell out just how special

recent times have been, for most Americans at least, for my entire life and a bit more. There have been two essential features of the basic landscape. One, American hegemony over much of the world, and relative physical safety for Americans. Two, an absence of truly radical technological change. Unless you are very old, old enough to have taken in some of the world. World War II, or were drafted into Korea or Vietnam, probably those features describe your entire life as well. In other words, virtually all of us have been living in a bubble outside of history. Now, circa 2023, at least one of those assumptions is going to unravel, namely

number two. AI represents a truly major, transformational technological advance. Biomedicine might too, but for this post I'll stick to the AI topic as I wish to consider existential risk. number one might unravel soon as well, depending how Ukraine and Taiwan fare. It is fair to say we don't know, nonetheless, number one, also is under increasing strain. Hardly anyone you know, including yourself, is prepared to live in actual moving history. It will panic many of us, disorient the rest of us, and cause great upheavals in our fortunes, both good and bad. In my view, the good will considerably outweigh the bad, at least from losing number two, not number one, but I do understand that the absolute quantity of the bad disruptions will be high. I am reminded

of the advent of the printing press after Gutenberg. Of course, the press brought an immense amount of good, enabling the scientific and industrial revolutions among many other benefits, but it also created writings by Lenin, Hitler, and Mao's read book. It is a moot point whether you can blame those on the printing press, nonetheless the press brought, in combination with some other innovations, a remarkable amount of true moving history. How about the wars of religion in the bloody 17th century to boot? Still, if you were redoing world history, you would have take the printing press in a heartbeat, who needs poverty, squalor, and recurrences of Genghis Khan-like figures. But since we are not used to living in moving history,

and indeed most of us are psychologically unable to truly imagine living in moving history, all of these new AI developments pose a great conundrum. We don't know how to respond psychologically or, for that matter, substantively, and just about all of the responses I am seeing I interpret as copes, whether from the optimists, the pessimists, or the extreme pessimists, e.g. Eliezer. No matter how positive or negative the overall calculus of cost and benefit, AI is very likely to overturn most of our apple cards, most of all for the so-called chattering classes. The reality is that no one at the beginning of the printing press had any real idea of the changes it would bring. No one at the beginning of the fossil fuel era had much of an idea of

the changes it would bring. No one is good at predicting the longer-term or even medium-term outcomes of these radical technological changes. We can do short-term, albeit imperfectly. No one. Not you, not Eliezer, not Sam Altman, and not your next-door neighbor. How well did people predict the final impacts of the printing press? How well did Did people predict the final impacts of fire? We even have an expression playing with fire. Yet it is, on net, a good thing we proceeded with the development of fire. In parentheses, Tyler writes, Fire, you can't do that. Everything will burn. You can kill people with fire. All of them. What if someone yells fire in a crowded theater? So when people predict a high degree of existential risk from

AGI, I don't actually think arguing back on their chosen terms is the correct response. Radical agnosticism is the correct response, where all specific scenarios are pretty unlikely. Nonetheless, I am still for people doing constructive work on the problem of alignment, just as we do with all other technologies to improve them. I have even funded some of this work through emergent ventures. I am a bit distressed each time I read an account of a person arguing himself or arguing herself into existential risk from AI being a major concern. No one can foresee those futures. Once you keep up the arguing, you are also talking yourself into an illusion of predictability. Since it is easier to destroy than create, once you start considering the future in a tabula rasa way,

The longer you talk about it, the more pessimistic you will become. It will be harder and harder to see how everything hangs together, whereas the argument that destruction is imminent is easy by comparison. The case for destruction is so much more readily articulatable. Boom! Yet at some point, your interhekian, or Paparian, has to take over and pull you away from those concerns, especially when you hear a nine-part argument based on eight new conceptual categories that were first discussed on less wrong 11 years ago.

existential risk from AI is indeed a distant possibility, just like every other future you might be trying to imagine. All the possibilities are distant. I cannot stress that enough. The mere fact that AGI risk can be put on a jar with those other distant possibilities simply should not impress you very much. Given this radical uncertainty, you still might ask whether we should halt or slow down AI advances. Would you step into a plane if you had radical uncertainty as to whether it could land safely, I hear some of you saying? I would put it this way. Our previous stasis, as represented by my number one and number two, is going to end anyway. We are going to face that radical uncertainty anyway, and probably pretty soon.

So there is no ongoing stasis option on the table. I find this reframing helps me come to terms with current AI developments. The question is no longer, go ahead? But rather, given that we are going ahead with something, if only chaos and leaving the stasis anyway, do we at least get something for our trouble? And believe me, if we get nothing, yes, we will reenter living history and quite possibly get nothing in return for our trouble. With AI, do we get positives? Absolutely. There can be immense benefits from making intelligence more freely available. It can also help us deal with other existential

risks. Importantly, AI offers the potential promise of extending American hegemony just a bit more, a factor of critical importance, as Americans are right now the AI leaders. And should we wait and get a more Chinese version of the alignment problem? I just don't see the case for that. And no, I really don't think any international cooperation options are on the table. We can't even resurrect WTO or make the UN work or stop the Ukraine war. Besides, what kind of civilization is it that turns away from the challenge of dealing with more intelligence? That has not the self-confidence to confidently confront a big dose of more intelligence. Dare I wonder if such societies might not perish under their current watch, with or without AI? Do you really want to press the button giving us that

kind of American civilization? So we should take the plunge. If someone is obsessively arguing about the details of AI technology today, and the arguments on less wrong from 11 years ago, they won't see this. Don't be suckered into taking their bait. The longer a historical perspective you have, you take, the more obvious this point will be. We should take the plunge. We already have taken the plunge. We designed and tolerated our decentralized society so we could take the plunge. See you all on the other side. So there's a lot to dig into here. Two things I'm going to put to the side for my little take on this. The first is the China version of the alignment problem. That's an argument that we've heard a lot, we've talked a lot about.

And I don't think in the context of this particular piece, it's either Tyler's strongest point or even really what he's trying to argue. So let's hold that aside. The second thing we're going to do is the clear annoyance with the Eleazar less wrong crowd. I don't know what was happening around March 27th when Tyler wrote this. It was right around the time that Eliezer had an essay in Time magazine, so maybe Tyler had seen some advanced copy of that.

But in any case, again, that's a little bit more inside baseball and personal than I think matters for the sake of our conversation here. But there are a couple parts of this that I do want to discuss. Let's talk first about Tyler's example of the printing press. Now, to some extent, he uses the kind of easily dismissable argument that it enabled the writings of Lenin, Hitler, and Mao's Red Book. I think a much more apt place to understand the negative unintended consequences of what is overall an incredibly important

and positive invention was the next century of European upheaval that happened in the wake of the printing press becoming. available everywhere in Western Europe. That was, of course, the violence of the Reformation and the counter-Reformation, a massive, painful struggle in which the end of the high Middle Ages and the Renaissance gave birth to the very beginning of the modern world. Volumes and volumes and volumes have been written about the connection between Gutenberg's invention and Martin Luther's Reformation. But the point is almost immediately there was chaos and social upheaval in that period of moving history to use Tyler's phrase. What's interesting about that example is,

One, the clear reality that, yes, the world is better when information is easier to move around as it was in the case of Gutenberg in books than it was before that. And two, that it's a great example of how, in most cases in history, technology is not necessarily the cause of upheaval, but the enabler of something that was already brewing. In this case, for example, it wasn't the rise of the printing press that made Europe's faithful upset with their Catholic leaders. It was a 150-year process of losing faith in that institution, involving a multi-decade period in which there were three people claiming to be the Pope and each having very serious factions agreeing with them. It wasn't the printing press that made a Medici Pope decide to start selling indulgences on mass that finally triggered Martin Luther. To paraphrase Billy Joel, Gutenberg didn't start the fire. So let's come back to Tyler's broader piece, though. Because ultimately, I'm kind of just extending his analogy here.

I'm not disagreeing with it. In fact, I'm validating it. What I like most about this piece is the framing of moving from a period of outside of history to living within moving history. This is something that I think absolutely is true and is going to be a cause of much upheaval and challenge just as Tyler identifies. However, what's interesting to me is that what Tyler is arguing here, reduced to its most basic form, is that because we can't know the possible outcomes in the future, because humans

are historically bad at knowing what happens in the future when they're living through moving history, that we shouldn't be overly concerned about existential or extinction risk from AI. I would say that most of the people that I know, who I think are the smartest and have the most sophisticated thoughts around AI safety, start from a similar place of not knowing what the future holds, but come to the different conclusion that if that's the case, if we not only are very bad at knowing the future in general, but also kind of have a pattern of not being able to predict how fast advances in this AI field were going to be. And also on top of that, don't really even quite understand what's going on inside these machines, then isn't the most humble response to that

to ask if we really need to scream forward with all possible haste. We are at what I hope are the last embers of the second phase of this conversation. I would characterize the first phase as the AI safety folks screaming into the void. And the second phase as this interstitial or liminal moment that we've been living through, where all of a sudden a lot of people tend to find themselves agreeing that these are real concerns and things that they'd like to see addressed. In that second weird in-between phase, the sides are hyper-essentialized and reductive. It's scream headlong forward as fast as you can, i.e. accelerationism. Or stop now because it already might be too late, on the other hand. What I think phase three might look like

is a little bit more dispassionate splitting the difference. I'll actually use a, for example, from the AI Policy Institute that we profiled a couple days ago. On their website, they talk about how two big policy changes that would make a difference right now would be for regulators to regulate data centers and for models to have to go through some gauntlet of proving they're safe before they're released to the public. Now, obviously, each of those issues has a ton to debate within it, but at least then we're debating specific policies rather than whether to do something or not. Anyways, even if you don't agree, I hope you will have enjoyed Tyler's framing of his issues and his putting it in a historical context and maybe found that at least a little bit useful. I, of course, appreciate you guys listening, spending some time on a weekend with me.

And so until next time, peace.