The AI bubble feels a lot more focused on increasing worker efficiency/replacing workers while the .com bubble felt more like throwing things at the wall to see what stuck.

People had ideas for what you could do online, but no one really knew what it looked like or how it could work. Since no one knew what form the Internet would take, a lot of companies were formed to throw shit at the wall and see what stuck.

I remember there being a lot of innovation in all kinds of directions. The tech looks old and crusty these days if you look back on it but at the time it was exciting, it seemed like something cool and interesting was happening every week.

I think that’s the difference. Back then it was diverse and small companies could quickly gain market share by doing something new and different.

The current bubble is mostly based around a disconnect between what a LLM can actually do, and the hype, VC money, and established tech giants trying to sell a vision of what it might be able to do with more time and investment. There’s little diversity in the tech this time, it’s just LLMs being stuffed into various roles, some of which have utility and many which don’t. That’s why the bubble will burst soon. For every interesting use, like summarizing code there’s an AI powered fridge nobody asked for and nobody cares about.

The other big difference is that while the dot com bubble had the big players participating, many of which are still around today, there was potential for a startup to come along and eat their lunch. Now we just have a bunch of massive tech corporations who are out of ideas, so they’re pushing AI because that’s all they have left to hype because growth must continue at all costs, even if the average consumer couldn’t care less about ChatGPT or CoPilot.

The dot com bubble was this crazy time where for a brief period a generation didn’t know about this thing that the generation after them was increasingly and rapidly interested in.

So like people use the whole “Wild West” metaphor and it’s really apt. Conglomeration was happening in the 90s with the rise of walmart, home depot, etc and it was becoming clear that opening a retail store was a dying business because the “big guys” were destroying towns. Like it was only a matter of time before a walmart came to your town and shuttered main street

But the internet was different. All these established entrenched companies didn’t care about it, yet. So you could make bank with basic ideas. Like oh, clearchannel owns 70% (now like 95%) of the radio stations in the us. Trying to start anything in that space is foolish, plus there’s all these regulations. But internet radio? Boom, millionaire. Petco and petsmart are rising and putting pet stores out of business. But pets.com? Had prominent national advertising including a float in the macys thanksgiving day parade, a Super Bowl ad, and was listed in nasdaq. Boom, millionaires. Busted after 2 years and now redirects to petsmart tho

Some of them stuck around. Like banking was obviously entrenched with old money but then a couple of rich kids were like what if we use daddies money to do internet banking? Then x.com and PayPal started and now we have elon musk and peter thiels reign of terror

AI seems to have some similarities in that there’s the whole “what if we apply AI to x” thing and VC dummies throw cash at it but it’s not as broad so the bubbles not as big and frankly it’s not as definitively revolutionary. The internet was clearly a game changer. Like anyone with half a brain who used it saw the potential early on; it was a new modality for communication with an unprecedented speed and dearth of information. And after it had matured a few years people started to see how fast it was progressing, especially via stuff like games. We went from doom to quake to final fantasy 8 and everquest in the span of the 90s and anyone paying even a lick of attention saw the potential for things like facetime, netflix, youtube, etc eventually.

But with AI it’s harder to picture. There’s the narrative that it will eventually do stuff and it can do impressive things but for the most part most people’s experience with it is that it’s like having a mediocre employee. Their work is okay but you have to constantly check it because they always make stupid mistakes. They tell you they’ll learn to stop doing that but it’s been several years now and they keep doing it. Just like teslas will self drive in 2018, chatgpt will reach agi any day now, maybe, or maybe it’s an illusion and it’s really just a bunch of if>then statements that are constantly trying to fix themselves but messing up others in the process.

Interesting question.

It’s different in the broad mechanics of the hype and the technology, obviously. AI isn’t really a basic information delivery mechanism like the 00s web was. It’s a lot more than that, thanks to the progress we’ve made since then. Plus, it has a lot more potential to impact artists and creative types. The web promised to propel them. Generative AI is threatening to make them completely obsolete.

But a lot of the psyche/sales hype stuff is very similar. The dot com boom started out all about democratizing the world and giving everyone a voice. Then it quickly became a capitalist free for all of coming up with any and every idea to shoehorn the Internet (and especially the web) into every walk of life. And the venture capital money-grabbing frenzy that inevitably led to. So that second aspect is very similar to where we are at right now with AI. And the first aspect isn’t a million miles away from the frequent assurances that AI will better society.

The technology stack for AI is more stable and well thought out than the nebulous web was. But it clearly still has some pretty big implementation challenges. Not to mention the resource/energy ceiling we are currently fighting against.

The potential of AI to destroy huge swathes of professions is massively different from the early commercial web. Back then the web was widely seen as a new source of careers, not a career killer.

I was alive during the dot com bubble and I don’t think normal people even noticed it. The web was a lot less centralised back in those days and stayed that way for a while. The websites that people were actually using didn’t go anywhere when the bubble burst but there was a lot more website turnover in those days anyway. People were always moving to a better newer service and there were multiple search engines that people used. Then AltaVista turned up and that was the engine to use, until Google turned up and everyone started using that. I still remember when Google was this cool new thing that most people hadn’t heard of.

The pop of the AI bubble will probably be much worse. During the Dot Com era, people were actually generating wealth for a period. With AI, so much has been invested in the concept with the result being very little of value being generated by most end users.

The difference was that there were a lot of unnecessary companies all doing the same shit or often, nothing at all. Whereas with AI it’s legacy companies adding unnecessary shit to their portfolio.

For those of us using the tools, actively, it doesn’t seem to be a bubble.

For a lot of us it’s already showing tangible measurable productivity increases, primarily on boring stuff normally we’d hate doing.

As an example, I use it often to help with documenting my code, it’s really good at summarizing what my code does abd making clear, legible, professional documentation for all my code.

That sorta stuff would normally take me hours and hours to do, now it takes about 1.

I still proof read it, but a lot of my typing and formatting and humming and hawwing is gone.

There’s a lotta shit like that out there getting streamlined more and more every month that goes by.

I think it’s maybe 50/50 bubble and actual value. Lots of garbage “products” vaporware out there by people on the bandwagon.

But also a lot of the tools truly are useful to folks.

The AI bubble is a lot dumber.

In the dotcom era things were easy. You could set up a website and make a business out of it in no time. Small effort. If it didn’t perform, you deleted it and tried the next one.

No “Big Tech” that was ruling and owning everything, like today. The existing telco’s were actually big, but they had no clue. Lawmakers had no clue either.

One difference is that we know one was definitely a bubble thanks to hindsight, whereas the other is only hoped to be a bubble because lots of people want it to turn out to be one. It remains to be seen whether it actually is.

When you see a pets.ai that will talk to your let’s, you know we’ve reached .com level saturation.

I think they’re quite comparable. Lots of companies earned a ton of investment funding for having the word “AI” in their name, and they will collapse over the next couple years. But much like the Dot Com bubble, the big ones will survive, possibly with only a few years to recover back whatever they lose during a pop.