[acdha]

Exactly - the exact same pitch would have been used with “big data” in the 2010s, and probably something about OLAP a decade before that. It was weird seeing that example without even attempting to explain what makes it AI other than, presumably, giving Gartner a lot more money.

[eesmith]

"Diapers and beer" was the go-to story for data mining of purchasing data in the 1990s. https://tdwi.org/articles/2016/11/15/beer-and-diapers-imposs...

That link shows both how widespread the idea was, and how shaky the evidence was.

[larvaetron]

> the exact same pitch would have been used with “big data” in the 2010s

I'm surprised I don't see this mentioned more often. The tech industry feels like a "more obvious than usual" case of déjà vu these past few years.

[acdha]

I think some of that is age and experience helping recognize the cycle (and the big consulting companies always pushing new things hoping their clients won’t notice their previous claims were off) but there also seems to be an angle around how much money is controlled by a handful of people seeking huge returns. The industry tends to focus on what VCs want and there just isn’t much diversity in that community – the guys who got lucky don't reliably keep having new ideas and having more money than you know what to do with tends to stifle creativity: they’re not forced to deal with criticism, nobody is stressing about their success, and their working experience is increasingly outdated because they’re hearing only from other rich guys who also not only don’t have to do the hard parts themselves but probably have entire teams “green-shifting” things so nobody has to tell their boss that their business isn’t as simple as some Gartner analyst assumed.