I've been on here long enough to see it with every tech hype cycle in the past ~20 years. Self driving cars, VR, bitcoin, then generically 'crypto', web3, now AI/LLMs, probably a lot I'm missing right now... It's funny how the optimist control the narrative in the beginning then the pessimist start chiming in around the time the hype starts to vaporize. The AI/LLM cynicism is rising right now it seems.
I guess I have a different view on this... I feel like you left out that the past 20 years has also seen enormous success in this industry, even to a historically unprecedented degree.
You listed a number of things that have been super hyped flashes in the pan that never really panned out (or in the case of self driving cars, have taken way longer to pan out than people expected), but you didn't list the things that were super hyped and then became big successful sectors.
I remember when I thought "web 2.0" was overhyped, but now it's just the water we all swim in. I remember when that went from blogs to being social media, and I thought that was way overhyped too, but it turns out it was a big deal. The cloud was overhyped, "big data" was overhyped, SaaS was overhyped.
And it's true that all of these were overhyped! But they also turned into real business sectors.
It's certainly difficult to predict which hype-y things are going to mature into sustainably large markets, and which are going to fade into obscurity, but a model of "things that are hyped are doomed" is not predictive.
My own prediction is that AI tools are both overhyped and also very promising. I have no illusions that this prediction could be totally wrong. And even if it's right, I'm even more uncertain what the successful business models are going to be after the dust settles. We'll see!
> It's certainly difficult to predict which hype-y things are going to mature into sustainably large markets, and which are going to fade into obscurity, but a model of "things that are hyped are doomed" is not predictive
Perhaps this is where we differ. I offered a list of things that, granted IMO, were all hype with little substance. Or perhaps just on a timeline so long that many people got the hype timing wrong.
Your list was mostly things that were obvious winners. They were mostly building steam with real world use cases and trending hard before the buzzwords got associated with the movements (eg. Web 2.0 and saas). These were obvious enhancements to the status quo. I’m not sure they were overhyped as they did very much become the defacto standard for their time. It doesn’t mean they will hold that title for ever, but tremendous economic value falls under those umbrellas. I’d argue intrinsic value too (unlike crypto).
AI/LLM might do that in some regards. But doing it in a way that makes lasting business sense is still tbd. AI eating the world, still very much tbd. So I do think we agree on this point.
There is a happy medium between "AI eating the world" and "AI is a scam," that is generally best described as "AI performing valuable but often boring services within enterprises."
We don't often see that because it gets totally drowned out between cynics and moralist scolds battling vapid hype-bros and pumpers, and also because it's boring.
It turns out there's actually a lot of tedious classification, OCR, entity extraction, and enterprise search to be done.
Amusingly much of the money made in SaaS was also in boring products!
My contention is that if you went back and did a sentiment analysis of the early to mid hype cycle conversations around the things in your list and compared them with the things in my list, there would be very little difference between the two sets of things. There was lots of hype and also lots of skepticism, for both sets of things.
My point is that hype vs. skepticism just isn't the useful metric.
I agree with you that this does not mean that it's impossible to identify which technologies have a "there" there, and which don't. I just don't think the level of hype is a very helpful input into the calculation.
There is a lot of cynicism (I am a cynic) but LLMs are in my opinion already past everything else you mentioned, they're being used by "ordinary" people in their daily lives. Non-tech people are using LLMs (in many cases where they shouldn't, and some where maybe they should) and the rate at which they've become more useful has far surpassed what my very cynical circles of industry friends expected. I still think things are overblown, but "AI" in one form or another is here to stay, for better or for worse (almost certainly worse).
