[Mistletoe]

Has there ever been a period o time where people saw a bubble coming and that we were in one, but it just inexorably refused to pop/drug out this long? This isn’t a rhetorical question, I’m wondering how this period compares to other irrational periods of the economy like railroad fever etc.

[neaden]

Not at all, there is a famous saying (often attributed to Keynes but as far as I can tell he never said) “Markets can remain irrational longer than you can remain solvent.”

[guzfip]

“But can the markets remain solvent longer than I can remain rational” is the real question.

[cmiles8]

It’s not been that long really. The dot com bubble was called a bubble for a while before it finally imploded. And just like now folks were in massive denial that it was a bubble.

One of the challenges here is that a lot of folks simply weren’t around then and haven’t seen what happens when everything implodes overnight. Those that have experienced it know what that looks like and know it will happen again.

[ifwinterco]

What makes this time weird though is back then (and in most bubbles) the bubble was creating jobs - supposedly (I'm too young) anyone who could write a bit of HTML could get hired in the dot com boom.

Whereas this is a very weird bubble where it creates big pumps in some equity prices but apart from a tiny number of people who are directly involved in AI research etc. it's not created any jobs, in fact by creating uncertainty it's probably caused fewer jobs to be created.

What that means for labour market dynamics when it pops I really don't know

[bigbuppo]

It's no coincidence that daytrading ascended with the dotcom era.

[bandrami]

In my lifetime: S&Ls, .com's, housing, and NFTs.

[banannaise]

Bubbles don't pop overnight. In the aftermath of any collapse, you can generally see a pretty clear pattern of red flags (and attempts to minimize them or cover them up). Some parties notice earlier than others, but the realization is generally a much more gradual process than the collapse.

[df2sdf]

Its obvious its coming soon simply because the mismatch between hype and expectations is far too large now. Hype is necessary to lure in investment but now its reached its peak imo.

The question is, is it a big bang or a slow release?

[ceejayoz]

NFTs lasted a lot longer than they should have.

[chasd00]

yes they did, Salesforce even came out with an "NFT Cloud" product.

[dragonwriter]

> Has there ever been a period o time where people saw a bubble coming and that we were in one, but it just inexorably refused to pop/drug out this long?

The housing bubble that peaked in 2006 was raised as an issue at least as early as 2000 and became a big topic of conversation in mid-to-late 2002, which is comparable to if we had started talking about an AI bubble roughly simultaneously with the release of GPT-3 and it had become a topic of wide concern shortly after the release of GPT-3.5.

So, in short, not only “this long”, but much longer.

[simianwords]

The bubble bursting has almost become eschatological for deniers. Keep praying that it will happen. It won't.

[ifwinterco]

LLMs can be a massively transformative and valuable technology, and this would still be a bubble, there's no reason those two things can't be true at once.

In fact "pure" bubbles where the focus item is of literally no value (tulips, NFTs) are quite rare. Much more common are the bubbles based on an actual real transformative innovation (canals, railways, radio, internet, LLMs).

Railways did absolutely transform how travel worked in the UK, while simultaneously almost everyone who invested in them lost their shirts

[greenchair]

Why the sudden pivot to agents and subagents? One shotting not good enough? Was a tech plateau reached? Or maybe one related to cost?

[banannaise]

That's also what NFT hypebros said.