Something like half of the companies that look overvalued do not look like they’ll benefit much from AI. They look more like they were chosen for safety against risks such as recessions and other near-term risks. I’m thinking of companies such as Apple, Chipotle, Home Depot, Lululemon, Mastercard, and Guidewire.
The “next few decades” is too long a time. Try evaluating what would have happened if you’d invested in companies in 2008 that looked like they would benefit from this decade’s demand for electric vehicles, robocars, or solar. Or what would have happened if you had tried to act in 1986 on Drexler’s forecast (in Engines of Creation) of a global hypertext system taking off within 10 years?
I’m guessing that if I’d tried those, I’d have guessed on Toyota or Honda for electric vehicles; for solar I actually made some money in Evergreen Solar in 2005-6, so I’m guessing it would have been my top choice (with a few others from this list?); I don’t know what I would have found for robocars; and I would have bought Autodesk if I’d realized it owned Xanadu (I’m unsure when I was able to find out that Autodesk had acquired Xanadu).
How well would those have worked? Evergreen Solar became worthless. Some of the more successful solar companies (Canadian Solar, First Solar) are still below their 2008 peak. Honda and Toyota have been unimpressive. Autodesk has done well, but mostly not while it owned Xanadu or during the dot-com boom.
This suggests that expecting an industrial revolution-level change after 2030 is a poor reason for choosing index funds that are loaded with high p/e stocks.
Something like half of the companies that look overvalued do not look like they’ll benefit much from AI. They look more like they were chosen for safety against risks such as recessions and other near-term risks. I’m thinking of companies such as Apple, Chipotle, Home Depot, Lululemon, Mastercard, and Guidewire.
This is a good point. Perhaps it’s worth shorting (or underweighting) specifically these companies.
The “next few decades” is too long a time. Try evaluating what would have happened if you’d invested in companies in 2008 that looked like they would benefit from this decade’s demand for electric vehicles, robocars, or solar.
Surely Tesla would have been on your list, either then or shortly thereafter (they IPO’d in 2010). And that would have been a good bet.
(I thought about buying in at the IPO, but figured I didn’t know anything the rest of the market didn’t. Then when they finally came out with the Model S and it was winning “Car of the Year” awards, and the stock price barely moved, I couldn’t take it anymore and bought in in Jan 2013. I’m glad I did.)
This suggests that expecting an industrial revolution-level change after 2030 is a poor reason for choosing index funds that are loaded with high p/e stocks.
Note that industrial revolution-level change is more significant than the transition to electric cars. Maybe it’s not more significant than a transition to nanomachines? Unfortunately I wasn’t old enough in 1986 to be able to say what the evidence looked like then, and how that compares to the evidence for pending transformative AI today. All I can say is that it looks to me like AI is coming. And I’d be quite surprised if none of Google, OpenAI, Microsoft, Facebook, Amazon, Tesla ends up being a major player.
Maybe that means I should go long some of those stocks (besides Tesla, which I’m already quite long), and short the rest of the S&P 500.
And I’d be quite surprised if none of Google, OpenAI, Microsoft, Facebook, Amazon, Tesla ends up being a major player.
If we get transformative AI in 2028, then I agree. I have some Google shares for AI-related reasons. But if we don’t get it until 2040, then I’d be mildly surprised if any of them will be close enough to the cutting edge to be major players.
if we don’t get it until 2040, then I’d be mildly surprised if any of them will be close enough to the cutting edge to be major players
Haven’t done any rigorous analysis, but it’s my impression that the field of tech giants has stabilized a bit in the last couple of decades. In that, once you become a tech giant now, you don’t drop out anymore. (Compare Microsoft’s trajectory post-Google to IBM’s post-Microsoft.)
I expect there to also be new tech giants by 2040. But I’d be surprised if all the current trillion dollar market cap companies are irrelevant by then.
That’s thinking about them as tech companies, specifically. Rather than evaluating their current AI plays. But I think the fact that the tech cos have so much engineering talent means they’re not likely to just totally miss the AI trend.
Something like half of the companies that look overvalued do not look like they’ll benefit much from AI. They look more like they were chosen for safety against risks such as recessions and other near-term risks. I’m thinking of companies such as Apple, Chipotle, Home Depot, Lululemon, Mastercard, and Guidewire.
The “next few decades” is too long a time. Try evaluating what would have happened if you’d invested in companies in 2008 that looked like they would benefit from this decade’s demand for electric vehicles, robocars, or solar. Or what would have happened if you had tried to act in 1986 on Drexler’s forecast (in Engines of Creation) of a global hypertext system taking off within 10 years?
I’m guessing that if I’d tried those, I’d have guessed on Toyota or Honda for electric vehicles; for solar I actually made some money in Evergreen Solar in 2005-6, so I’m guessing it would have been my top choice (with a few others from this list?); I don’t know what I would have found for robocars; and I would have bought Autodesk if I’d realized it owned Xanadu (I’m unsure when I was able to find out that Autodesk had acquired Xanadu).
How well would those have worked? Evergreen Solar became worthless. Some of the more successful solar companies (Canadian Solar, First Solar) are still below their 2008 peak. Honda and Toyota have been unimpressive. Autodesk has done well, but mostly not while it owned Xanadu or during the dot-com boom.
This suggests that expecting an industrial revolution-level change after 2030 is a poor reason for choosing index funds that are loaded with high p/e stocks.
This is a good point. Perhaps it’s worth shorting (or underweighting) specifically these companies.
Surely Tesla would have been on your list, either then or shortly thereafter (they IPO’d in 2010). And that would have been a good bet.
(I thought about buying in at the IPO, but figured I didn’t know anything the rest of the market didn’t. Then when they finally came out with the Model S and it was winning “Car of the Year” awards, and the stock price barely moved, I couldn’t take it anymore and bought in in Jan 2013. I’m glad I did.)
Note that industrial revolution-level change is more significant than the transition to electric cars. Maybe it’s not more significant than a transition to nanomachines? Unfortunately I wasn’t old enough in 1986 to be able to say what the evidence looked like then, and how that compares to the evidence for pending transformative AI today. All I can say is that it looks to me like AI is coming. And I’d be quite surprised if none of Google, OpenAI, Microsoft, Facebook, Amazon, Tesla ends up being a major player.
Maybe that means I should go long some of those stocks (besides Tesla, which I’m already quite long), and short the rest of the S&P 500.
If we get transformative AI in 2028, then I agree. I have some Google shares for AI-related reasons. But if we don’t get it until 2040, then I’d be mildly surprised if any of them will be close enough to the cutting edge to be major players.
Haven’t done any rigorous analysis, but it’s my impression that the field of tech giants has stabilized a bit in the last couple of decades. In that, once you become a tech giant now, you don’t drop out anymore. (Compare Microsoft’s trajectory post-Google to IBM’s post-Microsoft.)
I expect there to also be new tech giants by 2040. But I’d be surprised if all the current trillion dollar market cap companies are irrelevant by then.
That’s thinking about them as tech companies, specifically. Rather than evaluating their current AI plays. But I think the fact that the tech cos have so much engineering talent means they’re not likely to just totally miss the AI trend.