Don't Believe the Lull: AI Is Screaming Forward

0:00Today on the AI Breakdown, we're exploring an essay from Tyler Cowan about why there is a lull in consumer attention on artificial intelligence, but not in terms of AI progress. The AI Breakdown is a daily podcast and video about the most important news and discussions in AI. Go to Breakdown.net for more information about our Discord, our YouTube, and our newsletter.

0:24Welcome back to The Breakdown. For this weekend long read, we are actually doing the thing where we will start with the thesis and some excerpts from one piece, but then we'll go read a couple other pieces to either validate or challenge the points that it was trying to make. Now, the kickoff piece comes from Tyler Cohen. It was published in Bloomberg originally, and it is called The AI Hype Has Subsided, But the Revolution Continues. It explores some of the themes that have been common threads on this show for the last couple months, including the contrast between, on the one hand, the most intense part of the hype cycle post-ChatGPT coming back down sort of to Earth, or at least down a little bit from the lofty heights where it was, while at the same time seeing absolutely no evidence that AI itself is slowing down.

1:08And indeed, that is a bit of the point that Tyler is making. The subheader reads, don't be deceived. Artificial intelligence will be just as ubiquitous and transformative as advertised. The piece begins, we seem to be in what I can only call an AI lull. The initial excitement about ChatGPT, which started in January, has receded. Google searches for ChatGPT peaked in April and are now down significantly, as is customer engagement with ChatGPT. Cloud2, a new and wonderful model released in July by Anthropic, hasn't captured the public imagination the way ChatGPT did. Do not be deceived. While the hype and marketing may have died down, at least on the retail side, the AI revolution marches on, and it will be just as transformative as advertised.

1:49Next, Tyler talks about how in previous technological revolutions, there have been periods, long periods sometimes, where it was easy to not be excited about the technology. He argued that even before the dot-com crash, much of online commerce was pretty boring. He referenced that famous op-ed where Paul Krugman said that the internet was overrated. And to add a little bit of even longer history to it, Tyler points out that bringing electricity into factories happened over a period of decades, and that the printing press arguably had a bigger impact in the 17th century than it did just after the invention of the press in the mid-15th century.

2:22So one of the interesting parts of Tyler's argument is that having many people viewing AI as, as he puts it, a fascinating novelty but separate from their work might actually be a quote healthy state of affairs for the development of artificial intelligence. The reason why, Tyler writes, we're being granted some time to process the advances in AI and to fathom that large language models really can perform many tasks at high levels. Still, even as the market and individuals are digesting the generative AI movement, Tyler points out that nothing is really slowing down. By way of example, he points to OpenAI's new ChatGPT enterprise offering, the rumors surrounding Google's Gemini, and as he puts it, quote, Perhaps most significant, open-source AI models are advancing at a rapid pace even if most casual users don't realize it.

3:06Many models were released or improved in August with generally better results than might have been expected only a few months ago. Open-source AI creates resources that other programmers can build on much as they can with the open-source operating system of Linux. Perhaps open source AI will never dominate at the retail end of the market, but it will drive a lot of innovation. Ultimately, Tyler comes to the conclusion that it is refreshing, that as he walked through a mall in the nearby area of Northern Virginia, he didn't see people pushing artificial intelligence. As he put it, no posters announcing the coming new AI-powered toys for Christmas, no AI-enhanced shopping app, no AI-assisted beauty salons.

3:40Okay, so Tyler's point is that one, yes, there is a lull in the public attitude towards AI, but that too, that lull has pretty much nothing to do with what's actually going on with the technology. What I'd like to do for the next part of this conversation then is look at a couple other opinion pieces from this week that serve to validate Tyler's point. Specifically, I'm going to look at two areas in which artificial intelligence is having a clear and dramatic impact on conversations of societal importance. The first of those is in and around work and labor unions. On the same day that Tyler published his piece for Bloomberg, Harvard law professor Sharon Block published a piece in The Hill called, We're experiencing a post-inflation AI and climate-driven labor revolution.

4:20Block writes, Labor Day marks the traditional end of summer, and this summer we saw an unprecedented wave of strikes, walkouts, and labor actions. From actors and screenwriters to Starbucks employees to auto workers and delivery drivers, workers throughout the economy are taking this moment to reassess their jobs and negotiate for better pay and working conditions. In the wake of the pandemic, and as we contemplate the future of work, there are undoubtedly very few people working today who aren't asking themselves questions like, is artificial intelligence coming from my job or letting my boss spy on me?

4:49Is this extreme heat making my job dangerous? How can I afford to take care of my family if my wages don't keep up with inflation? Each of these questions, she writes, represents a big transformational challenge, but undergirding them all is a fundamental fact. The economy and the workplace are changing in ways big and small that we are not fully prepared for. One of the responses, Block says, is a resurgence of labor organization. Block points as an obvious example to the writers' strike and the Screen Actors Guild strike in Hollywood. However, Block says that labor organization isn't the only thing that's needed when it comes to the disruption of AI.

5:21She writes, Congress should be at the forefront of addressing the challenges that artificial intelligence poses to the workforce. For example, Uber and Lyft drivers are among the workers most impacted already by algorithmic management. Their bosses are apps, not humans. But because rideshare drivers are treated as independent contractors, they are denied basic workplace protections. Senator Bob Casey, Democrat from Pennsylvania, has introduced a bill to guarantee all workers a human boss. Congress has not yet advanced this legislation, but it should. Now, for the sake of this conversation on this particular podcast, I'm not so much getting into the politics of these particular issues.

5:55But what I'm interested in is the point that AI is significant enough that it's one of the major causal factors driving a resurgence of collective action and union organization. And that's happening regardless of whether people are using ChatGPT as much as they were two months ago. A second area that shows absolutely no signs of slowing down and in fact shows signs of the opposite of speeding up is AI on the military front. Two recent examples of that. The one is another opinion piece, this time from September 6th, by Yulia Lattinina, once again in The Hill, called The Staggering Implications of AI Drone Warfare.

6:28Yulia points out a number of reasons why traditional human-operated drones aren't really a good fit for the Ukraine-Russia war, but that in its place another weapon is arising. They write, So what's left? An AI. A drone GPT. The drone uses GPS, has an industrial or tactical-grade inertial measurement unit capable of navigating in GPS-denied environments, processes various signals such as from nearby cell towers, but above all, it is equipped with a camera and a processor. It takes a snapshot of the target and is able to compare it against a database of blueprints stored in its memory. It's basically a self-homing Tesla with an explosive on board.

6:59So Yulia asks, is this a game-changer? Let me quote to you from a great book by Lawrence H. Keeley, War Before Civilization. The most common form of combat employed in primitive warfare but little used in formal civilized warfare has been small raids or ambushes. That's the essence of war, to kill without being killed, which became almost impossible after the emergence of civilizations and armies. A drone GPT changes back this most fundamental factor. You can once again kill without being killed. You can use a Skynet to do your bidding. Yes, you could previously do it with a Hellfire missile, but a Hellfire costs $150 ,000 a piece.

7:31This stuff can be built for a pittance in a shed. The implications are slow to come, but they are staggering. War, Yulia writes, will never be the same. As if to put a fine point on that, U.S. Deputy Secretary of Defense Kathleen Hicks has delivered two major speeches in the last two weeks or so, all about effectively the integration of AI into the U.S. military. The first was called The Urgency to Innovate and was given on August 28th. In that speech, Hicks said, while DOD always has an imperative to innovate, there's no mistaking why that imperative has taken on more urgency in recent years.

8:00Because the main strategic competitor we face today is different from the rival we faced during the Cold War, a rival who was relatively slow and lumbering compared to the PRC of the present. And while America shed blood and treasure over 20 years of war in Iraq and Afghanistan, the PRC worked with focus and determination to build a modern military, carefully crafting it to blunt the operational advantages we've enjoyed for decades. But the one advantage they can never blunt, steal, or copy no matter how hard they try, because it's embedded in our people, is American ingenuity. Our ability to innovate, change the game, and in the military sphere, to imagine, create, and master the future character of warfare.

8:33Now, a lot of the speech is about how they've spent the last two years improving the Department of Defense's ability to actually innovate and take advantage of that American strategic differentiation. And while there's lots of different aspects of it, one area of focus is on what Hicks calls a data-driven and AI-empowered military. On that front, Hicks referenced one, we issued data decrees to mandate all DOD data be visible, accessible, understandable, linked, trustworthy, interoperable, and secure. Two, they put data scientists across every combatant command. Three, they say, quote, we stood up DOD's chief digital and artificial intelligence office, which is accelerating our adoption of data, analytics, and AI from the boardroom to the battlefield.

9:10Now, at the end of that speech, Hicks also said, quote, today we're making another big bet, the latest piece of our comprehensive warfighting-centric approach to innovation. It's called the Replicator Initiative. Replicator is meant to help us overcome the PRC's biggest advantage, which is mass. More ships, more missiles, more people. Historically, Hicks continues, even when we mobilized our economy and manufacturing base, rarely have America's war-winning strategies relied solely on matching an adversary ship for ship or shot for shot. Instead, we outmatch adversaries by out-thinking, out-strategizing, and out-maneuvering them.

9:40We augment manufacturing and mobilization with our real comparative advantage, which is the innovation and spirit of our people. To stay ahead, we're going to create a new state-of-the-art, just as America has before, leveraging autonomous systems in all domains, which are less expensive, put fewer people in the line of fire, and can be changed, updated, or improved with substantially shorter lead times. Hicks expanded on the Replicator initiative a few days later on September 6th, but I think the news is best summed up by the Wall Street Journal, who wrote on Wednesday morning, Pentagon plans vast AI fleet to counter China threat.

10:11Defense Department seeks an array of air, land, and sea-based autonomous systems to keep pace with adversaries. Hold aside the niceties and the politicking and the call to American ingenuity. And really what she's talking about is the AI-ification of the American military. Now, again, that is the subject of not one, but probably many AI breakdowns in the future. But the point for today, bringing it back to the original piece, is just to validate Tyler Cowen's thesis. That for as much as it may seem like normies aren't talking about AI as much as they might have been a few months ago, it would be a huge mistake to think that artificial intelligence wasn't every bit as transformative, not just in the future, but right now, as so many have said.

10:51Anyways, friends, that is going to do it for today's episode. I hope you are having a great weekend. I appreciate you listening as always. And until next time, peace.