Frankly, for investing, AGI shouldn’t even be on your radar. AGI can only result in three outcomes - extinction, complete totalitarianism, or some sort of communist system. Most likely any AGI would just kill us, probably by accident. But even if they manage to control it, then what? You quickly end up with one or a handful of AI companies owning literally the entire economy, producing everything. At that point capitalism is done. Either we allow the new oligarchy to exist indefinitely (totalitarianism) or we nationalize the companies and go full Star Trek.

What these AI companies miss is that true AGI is incompatible with capitalism. Capitalism doesn’t work when only 1-3 companies make all the goods and services in the economy.

Your point on 2008 is valid though. There’s a lot we just swept under the rug instead of dealing with properly.

You can’t eliminate the risk, but you can ameliorate it. Yes, sectors like retail, construction, and even healthcare will take a hit during the AI bust. After a crash, there’s much less money floating around. People are out of work. Asset prices are down. People spend less. So every sector gets hit.

But there is a difference between a hit and a collapse. If you’re invested in OpenAI, and the bubble bursts? Your investment is going to zero. The company is going to go bankrupt. Shareholders will be completely wiped out. If you’re invested in say…Home Depot, well yes, in a crash, the stock will go down. With millions laid off, people will have less money to spend on home renovations, so demand for the Depot’s products will decline. Their stock price will take a large hit.

But, if you’re smart, this won’t hurt you. Eventually things will recover, the bubble will clear out, and demand for building products will return. At that point the Home Depot stock price will recover. As long as you don’t sell during the crash, you won’t see any losses.

That is the key difference. An investment in any sector right now risks a decline and then recovery a few years down the road. An investment in the AI sector risks a complete collapse to zero.

NASDAQ is still buying. S&P500 isn’t. But it’s still a major concern if you’re holding a total market index fund or similar.

What do you think about value mutual funds like these?

https://www.schwab.com/research/mutual-funds/tools/schwab-funds/index-funds/fundamental

I’m moving most of my US stock holdings to these funds. It’s still US market exposure, so there will be some risk from the AI bubble. But the key thing is that these indices are not trying to track the Dow Jones, the NASDAQ, etc. They don’t actively manage funds - they don’t pick winners and losers. They are a passive investment blindly applying uniform rules - just like index funds. However, they don’t weight by market cap. Instead they weight stocks in the index based on fundamental values - actual sales, retained earnings, book value, etc. The actual share price of the companies isn’t factored in at all. And you can’t even get into one of these indices until your company is profitable.

The expense ratios are higher than I’m used to. The low-cost index funds I’ve traditionally used have expense ratios more like 0.01-0.02%. This is 0.25%. This is still well below the 1% most managed funds use. But that is a downside of this.

With how tightly the whole global financial system is tied together, I’m not under any illusions that moving to these funds will eliminate my exposure to the AI bubble. But I hope at least to ameliorate it.

Don’t be so smug. The intended victims of this IPO are passive investors who aren’t trying to time the market and are just long-term holding index funds for retirement. That’s exactly who Musk is planning to rob. If you’re taking the typical low-cost index fund strategy, you will be giving approximately 1% of your life savings to Elon Musk.

I do. But I do it mainly to support the creators I follow. I’m not going to sign up for a bunch of patreon or channel subscriptions, but at least they can get their share of the subscription money.

Seriously. My default script when interacting with customer service folks when asked my name is to say my name and then spell out both the first and last names. Because my first name has several variants that people will misspell it with, and the surname is relatively common but still one people struggle with. It’s easier just to spell it out than to fix the inevitable input errors later.

“I would like one cup of water and the entirety of Wikipedia translated back and forth between Navajo and ancient Sumerian Avogadro’s number times.”

FNDX and similar. https://www.schwab.com/research/mutual-funds/tools/schwab-funds/index-funds/fundamental

FNDX and similar.

https://www.schwab.com/research/mutual-funds/tools/schwab-funds/index-funds/fundamental

FNDX and similar.

https://www.schwab.com/research/mutual-funds/tools/schwab-funds/index-funds/fundamental

I’m still invested in some US stocks, but I’m switching my US market exposure to an index fund that weights by actual sales, revenue, and other objective factors, rather than market cap. Companies don’t even get into the index unless they turn a profit first.

Yeah I’m thinking of an automated version of greebling. Except you design the extraneous bits so that they’re only attached to the intended print like print supports - something in a non-critical location, easily torn away.

Then five minutes later, someone figures out how to make a 3d printable gun that bypasses the gun detector on the 3d printer. It’s not like you’re printing a whole gun; you’re printing parts, most of which look nothing like a gun. How hard would it be to design an algorithm that takes a gun part cad file and then adds a bunch of extraneous pieces to it that can be easily removed? Just keep adding extra crap until the system no longer detects it as a gun part.

Ok! :D :D

I’m sure with all the latent mental illness he’ll be great in bed!

There’s a reason we abandoned monarchy. Zuck is effectively the king of meta. Literally. The founding corporate structure is designed so he literally can never be fired. And if you give anyone absolute unaccountable power, it’s only a matter of time until they drive things right off a cliff.

Don’t even wait for the drainage district to do it. What’s Tesla going to do, sue you for blocking their illegal dumping?

I honestly deeply despise articles like this. Articles like this are how we get trapped in forever wars. Fuck the Iran War. I didn’t want it to start, and I don’t want it to continue. But the main reason we get caught in forever wars is that the media lampoons any president that actually does the wise thing and ends the doomed conflict. They portray them as a coward or demonize them for letting troops’ deaths go to waste.

It’s like anything. I can cook my own food. That doesn’t mean I don’t go out to restaurants sometimes. I could use an LLM myself to write a novel. But it wouldn’t be the same novel. And there is some skill in prompt writing. Even then, just the sheer time to generate a novel-length coherent work from small snippets of chat windows is still a large investment of time.

Of course, we know how this will actually go down. AI generated works are going to be much cheaper to produce. Therefore they’ll be me more profitable if they can sell at the same price. Barnes and Noble thus has a strong incentive to not carefully label AI works as AI-generated.

Ideally they would all be in their own section. AI-generated works are only allowed in the part of the store that is labeled as such. That’s the proper way to do this.

But that’s not how it will actually be done. Buried somewhere in the fine print inside the back cover of the book will be a long paragraph, one sentence of which mentions the work is AI generated. Or the inside back cover will have a QR code labeled “notes on this work,” and there will be a hundred page long legal disclaimer that briefly mentions the book is AI generated.