AI Revolution vs. Computing Revolution

I have been reading a lot about computing revolution in 60s/70s. Seeing a lot of parallels to today. A bunch of enterprises spent millions on these IBM/DEC machines but didn't see any ROI. But then a lot of consulting and product companies popped up to help with that. IBM/DEC had crazy high margins during this time (similar to Nvidia/micron etc) but all withered away eventually. In the end the software companies (Microsoft, Google etc.) captured most of the value, what's the equivalent of that for AI?

2 points | by nitinagg 20 hours ago

3 comments

  • silverFork 20 hours ago
    Well the cheap and dinky but yet accessible won over complex and expensive. Then they all learned to be cheap and then extract the cash some other way; ads became the money generator over functionality; they started to look at whatever industry that can make money using internet and sucking it all.

    Ai premise is that it will be replacing costly labor or maybe something else will emerge but whatever that shows any new way of making money. The same old companies are all looking like hawks to steal any business that will make money out of Ai. Who has the most cash, will buy it or copy it. They will all so the same thing and my guess is that they will compete for the same and add new industries to extract cash from them.

  • Wormaus 18 hours ago
    Of course they will. They are for-profit companies. It’s all about being the 'best', the 'biggest', and the most 'famous'. Whoever can control their environment gets to set the rules. And that’s exactly what’s used to become even 'bigger', 'better', and more 'prominent'. Everything else is just a means to an end.
  • platevoltage 13 hours ago
    Much like PCs made these big mainframes impractical, LLMs will probably end up mostly on-device, and medium/large organizations will run their own inference in-house. That might work out for some hardware companies, but I think there will be a big loss on these data centers.