AI 'Noose' Tightening?

[RKW]

Patrick Boyle did a good run down on this a few days ago.

But you know, they still gotta punch out those IPOs before people realise what a turkey this whole thing is as an investment.

The real tragedy of all this, is that really large pools of money flow into the latest scam, instead of to proper research and projects to benefit all.

You may be right that there's an AI investment bubble. But that doesn't mean AI itself is a scam or that the technology is going away. We've seen this before with transformative technologies: a technology can be genuine while the investment boom surrounding it is excessive. Railroads, electricity and the internet all attracted capital beyond what some individual investments ultimately justified. That didn't make railroads, electricity or the internet scams.

 

[DesertRatt]

Railroads, electricity and the internet

The railroads were massively over-invested in it's infancy and it took 50 years for the revenue to come in and payoff that investment. Likewise, on the Internet, the carriers waaaay over built the fiber network at the inception and it took about 15 years for traffic to fill the fiber network. Good point! As an aside, it looks like the EV Tech was wildly overinvested. Ford wrote off $19B by themselves.

 

[RKW]

The railroads were massively over-invested in it's infancy and it took 50 years for the revenue to come in and payoff that investment. Likewise, on the Internet, the carriers waaaay over built the fiber network at the inception and it took about 15 years for traffic to fill the fiber network. Good point! As an aside, it looks like the EV Tech was wildly overinvested. Ford wrote off $19B by themselves.

The computer industry is another example. The clearest example is the late-1990s technology/dot-com boom. The investment was excessive. But the underlying technology wasn't. Companies bought too many: PCs, servers, telecom systems, fiber-optic networks routers, data-center equipment and software. A lot of investors lost enormous amounts of money. But the result was also that the United States ended up with a tremendous amount of computing and communications infrastructure at dramatically lower prices. That infrastructure subsequently became enormously valuable. So you can have this sequence: Real technological revolution → huge investment → excessive investment → financial collapse → surviving infrastructure becomes extremely productive.

 

[felix_katze]

Railroads, electricity and the internet all attracted capital beyond what some individual investments ultimately justified. That didn't make railroads, electricity or the internet scams.

Go a bit further back. Canals were the big thing, massive amounts of money poured in to the new transformative thing of the day, only for railways to come along and kill them dead.

The scam is because this is all engineered precisely to list IPOs at nonsense values. It's exactly what happens when you have a captured regulatory system that no longer functions.

But this time is different....

Nah it's the same as the silver mines Twain wrote about.

 

[felix_katze]

There's a reason this is happening very specifically in the US, right now, and almost nowhere else.

 

[audiotemp]

There's a reason this is happening very specifically in the US, right now, and almost nowhere else.

I assume you are referring to introduction of the "pitch clock" to MLB.

 

[Mr Boochie]

An interesting Frontline show yesterday called "Scammed." It involved online scamming — including the use of AI and cryptocurrency.

 

[dfunghi]

This link should give you access to the NYT article.

https://www.nytimes.com/2026/09/28/technology/openai-astra-safety.html?unlocked_article_code=1.E1E.w0Jt.r3XD1CHfT1te&smid=url-share

They didn’t intend to read the article. They just wanted to complain. Simply discussing ongoing happenings in the AI arena seem to elicit deep feelings in some that have led to an unpleasant exchange. Very non AK like.

 

[w3jn]

Note to all: I've had to delete a bunch of political posts, and posts that were bigoted and nasty.

Please be civil to each other, and refrain from politics.

 

[DesertRatt]

An interesting Frontline show yesterday called "Scammed." It involved online scamming — including the use of AI and cryptocurrency.

This is to be expected in a declining empire and culture. The market system demands compound annual growth. And, it doesn't care what it must do to get it. From that dilemma, the Internet was born without a scintilla of planning. You reap what you sew.

 

[Mr Boochie]

It's sad about Senator Guffey's son.

 

[louisjames]

They didn’t intend to read the article. They just wanted to complain. Simply discussing ongoing happenings in the AI arena seem to elicit deep feelings in some that have led to an unpleasant exchange. Very non AK like.

Thanks for the heads up and the context. Very sad.

 

[mrswakka]

This is to be expected in a declining empire and culture. The market system demands compound annual growth. And, it doesn't care what it must do to get it. From that dilemma, the Internet was born without a scintilla of planning. You reap what you sow.

Like usual, I asked Alexa, because I know you guys love it. The difference is, the version of Alexa I’m using is different. using the online version, instead of closing the window, I email the page to myself. I’ve had one ongoing conversation for 6 months.
The other day Alexa spent 20 minutes telling me how Amazon prime was a scam. Even though the Steelers play the browns Thursday night…

“

Sep 30 at 4:45 PM
Hiya dude, can you give me a short answer where the payoff on Ai? Explain like I’m slow, like usual
Hey Dave! So you already got the deep dive on this earlier today — here's the short version:


The payoff on AI is real but slow. Think of it like electricity in the 1800s — they ran the wires everywhere, but it took decades before factories actually got faster. Same thing is happening now. Companies are dumping billions in, but 95% of enterprise AI projects aren't showing measurable results within six months.


Where it does pay off right now:

• •
  Narrow, specific tasks — writing, coding assistance, customer service bots
• •
  Individual productivity — one person using AI well moves faster. The problem is that speed doesn't scale when teams try to collaborate with it.
  

The broader "AI changes everything" thesis? Still unproven — and the pattern historically (electricity, computers, internet) is that the real payoff takes decades, not years.



Short version: useful in spots, overhyped overall — which is basically what you've been saying all along.

 

[mrswakka]

And who knows what technology will come up before that kicks in dude?


Exactly -and that's the wildcard nobody can price in. By the time AI's "decade-long payoff" actually arrives, something else might have already lapped it. Quantum computing, brain-computer interfaces, some tech we haven't named yet — history's pretty consistent about this: the thing that actually changes everything usually isn't the thing everyone was betting on. The internet didn't kill TV, it killed Blockbuster. Nobody saw that coming either.









Alexa says: Exactly — and that's the wildcard nobody can price in. By the time AI's "decade-long payoff" actually arrives, something else might have already lapped it. Quantum computing, brain-computer interfaces, some tech we haven't named yet — history's pretty consistent about this: the thing that actually changes everything usually isn't the thing everyone was betting on. The internet didn't kill TV, it killed Blockbuster. Nobody saw that coming either.

 

[felix_katze]

Blockbuster killed Blockbuster, like Kodak killed Kodak, and Yahoo killed Yahoo.

 

[theophile]

Blockbuster killed Blockbuster, like Kodak killed Kodak, and Yahoo killed Yahoo.

So: AI will kill AI?

 

[mrswakka]

So: AI will kill AI?

I can always ask…

 

[Mr Boochie]

So: AI will kill AI?

Some newer technology may come along to supplant AI.

 

[Enophile]

One AI to rule them all, one AI to find them, one AI to bring them all, and in the ether bind them.

We know how this will end; there can be only one.

 

[felix_katze]

So: AI will kill AI?

Those businesses failed because when given the opportunity to radically alter their business, they passed.

The internet didn't kill blockbuster, passing on buying Netflix did.