Here are some things I think I am thinking about this week:
1) Google’s VEO 3 is Mind Blowing.
So far AI has been interesting, but not earth shattering. I use ChatGPT dozens of times a day and it is basically refined search. For me personally, I’d argue it’s turned me into a sort of hyper efficient financial advisor. I am kind of a Jack of All Trades, expert in none type of financial advisor, but ChatGPT makes me an expert in almost any financial field as I know enough to get me 80% of the way there and AI gives me access to 100% of the knowledge. I know all of the right prompts to get to the right answers on things because I have most, but not all of the knowledge. For example, I am not a tax expert, but I know enough about taxes and how they relate to personal finance that I can quickly resolve almost any tax related issue and answer questions that might have otherwise relied on an expert. AI is basically my financial associate, except it’s an associate that does all the tangential things in finance that aren’t my precise area of expertise. 6 months ago, if I had a niche financial planning question I’d outsource it to an actual CFP. Today I get the answer on AI and it’s done faster and in most cases with greater efficiency and clarity. That’s a big deal, but again, not earth shattering as it basically just made me a lot more productive and eliminated any need for me to hire a specialist in these specific fields. But this is starting to all change and it’s occurring very very quickly.
I recall seeing AI videos just a year ago and thinking that this isn’t replacing anything too soon. The hands were weird and the mouth was always off. It was obviously fake. But then Google released VEO 3. And it’s earth shattering. The progress within the technology is stunning. And while video is the most obvious improvement this is also happening across all of AI. It’s changing at an exponential rate and this is going to go from having a small impact to a very big impact over time.
Google’s VEO is the first time I’ve looked at AI and said “oh wow, this is different”. It’s almost visually perfect. But it’s so much better than it was just a year ago that it’s hard to believe. And I can only assume we’re sitting in the early part of an exponential curve. The macro impact is going to be seismic.
2) The Macro Outcomes
The long-term impact of AI is going to be enormously disruptive. Here are some things I think will happen:
a. Inflation will collapse. At some point in the next 5-10 years we are going to be producing so much stuff that supply will suffocate demand. Services are about to become very easily supplanted. This has the potential to have a devastating impact on mid-level white collar services jobs. Meanwhile, executives and supervisors will become increasingly valuable as they’re the ones QCing the AI work. Again, AI is the super associate. And as a result, the cost of this work in aggregate is going to slow or decline as the supply of it booms. This will spread slowly into the price of goods as robotics improves. Before long the robots will be producing so much that goods inflation will also collapse. That is probably much further out on the timeline, but not as far as some might assume. The age of abundance is coming.
b. Everything will become a technology firm. Margins are likely to move higher across all of corporate America. This means valuations could actually move even higher and owning capital is going to become increasingly important for financial well-being. The battle between labor and capital is going to swing even further in favor of capital. The efficiencies from AI will accrue mainly to capital in the form of higher margins and profits. Firms that adopt these technologies and lean into them will be the biggest beneficiaries. The poor and middle class will get increasingly left behind as inequality becomes even worse.
c. Governments will get much bigger. As inflation falls governments will have a much greater capacity to spend. And they’ll need to as employment becomes more fragmented and things like employment claims and Social Security become more important. I wouldn’t be surprised to see a Universal Basic Income in my lifetime.
d. Decentralization will become much more valuable. Relying on big firms for employment is going to be increasingly untenable. The whole economy is going to become increasingly decentralized. Entrepreneurs and consultants who can leverage AI and help other people navigate the economy will become more valuable. Having your own brand and company is the economy of the future.
e. Authenticity, integrity and veracity will become much more valuable. It’s going to become very difficult to distinguish fact from fiction in this world. Authenticity and the ability to disseminate fact from fiction will become very valuable as audiences seek out experts who can disentangle an increasingly inauthentic world.
f. The next recession is going to be a labor market shocker. Each of the last few recessions resulted in fewer and fewer manufacturing jobs. This is primarily a technological effect (and no, the tariffs aren’t stopping this train). When the recession hits, firms fire workers and then realize that technology already made many of these jobs obsolete. AI is going to make this much, much worse. The next time we have a recession we will lose millions of jobs and firms will realize that they can be easily replaced with AI. If the government doesn’t respond in a big way the next recession will be similar to the GFC labor recovery in that it will be a very long and slow one.
3) Can you have an AI induced recession?
Speaking of recessions…I’ve always been critical of the idea that technology is bad for the economy. But that’s mostly a rate of change argument. In other words, as long as the rate of technological change occurs slowly then firms and employees can adapt to the change and find other jobs that replace the technology changes. For example, when Ford created the automobile the rate of adoption was very slow because cars were expensive. It wasn’t the type of technology that could be adopted very quickly by millions of firms or millions of consumers. As a result, all the horse drawn carriage drivers were able to adapt and learn other skills as the rate of car adoption was slow. This is what technology usually does. It has a prolonged investment and adoption curve. AI is different because AI itself is going be so abundant. You won’t need the robot right off the bat (although that’s going to be a game changer as well) because you’ll have some sort of handheld device that functions like a personal assistant 24/7 and access to this technology will have low barriers to entry. The mass adoption of this technology has the potential to disrupt the labor market in a manner that we’ve never seen before because the rate of adoption will be so much faster. Yes, it will make us better off in the long-run, but if you think the politics around things like lost manufacturing jobs are bad, just wait until AI comes for the white collar jobs.
But this is an interesting paradox. AI will make us better off in the long-run. But it will come at a cost. A very disproportionate cost. The Nasdaq Bubble is an interesting case study in how this could all play out over time. Many people don’t realize that GDP never went negative in the aftermath of the tech bubble. We lost almost 3MM jobs and the rate of GDP growth slowed, but from 2000 until 2003 RGDP actually averaged 2.25%. That’s a figure that we’ve all become pretty familiar with given that that’s more or less where we’ve been growing for the last few years. And I suspect that’s what an AI induced “recession” could end up looking like. We’ll get either a massive slowdown in the labor market or a decline in the labor market, but it doesn’t have to coincide with a decline GDP because the gains in real productivity could be more meaningful.1
Anyhow, that’s enough from me about AI and recessions. I actually think it’s all much more exciting than anything else. But it’s going to come at a cost. And people who aren’t prepared for these changes will get left in robotic dust or worse, virtual dust. I hope you’re having a great week. As always, stay disciplined!
1 – The fun outlier scenario here is if AI causes a labor market decline that hurts the people who can least afford it. Ie, people with significant household debt. There’s a reasonable argument to be made that AI and debt-laden consumers won’t play nicely and that this could strain indebted households in a way that has a lot of unpredictable negative knock-on effects.