"Generative AI could be a watershed moment for the PC industry."
Should I point out that the only way from a watershed is downhill?
Of course, I'm just a grumpy old git that finds marketing and management speak profoundly irritating.
Business interest in AI PCs is fizzing, at least according to IDC, even though the analyst admits “use cases have yet to be fully articulated.” CEO Pat Gelsinger presents during Intel Innovation in San Jose Intel slaps forehead, says I got it: AI PCs. Sell them AI PCs READ MORE Such is the hype around generative AI since …
Do you ever find that, sometimes, you have just answered your own question?
> it is difficult to see how AI demand will drive PC upgrades
If only we could find a workload that your current, rather average, PC can not handle unless you upgrade it.
> AI models like Stable Diffusion are too demanding for average PCs
Ah ha! Now we just have to build up a demand for running AI models!
“While uses cases have yet to be fully articulated, interest in the category is already strong. AI PCs promise organizations the ability to personalize the user experience at a deeper level all while being able to preserve data privacy and sovereignty.”
Although I'n not a native speaker, I always thought that my English was quite sufficient. But this to me is just a string of marketing terminology generated by a BS-generator. Other than "no-one knows what to do with it", I see no meaning of theses words.
> While uses cases have yet to be fully articulated
If only people like IDC (and the bosses at HP, Lenovo Intel at al) could see beyond the Big Brand hype machines, like ChatGPT (which are all very much based on offering cloudy "solutions") to actually look at how smaller (much, much smaller) models are already being used. Anything TinyML can do, we can do bigger! And even if the actual models are cheap to run (won't need a new PC) then training can always eat up the cycles.
It is surely a lot easier for them to take ideas for use-cases from already working small-scale systems[1] and pump them up into something that won't *quite* run on your existing dreary old PC but will go like the wind on their "next gen", just a bit bigger, PC. After all, software bloat is already an industry standard!
Guess the real problem is that it would mean all these CEOs having to train themselves on published descriptions about people's projects and then have some ability to mix it all up and generate some answers to match the template "needs a bigger PC, but not so stupidly large the market will just laugh at us".
Hmm, training, public data, generating text, fitting a template - that just sounds so familiar, it is on the tip of my tongue...
[1] ok, this one isn't business-oriented[2], but there is still the whole home PC market to re-invigorate! We just set up a Pi Zero 2 to run the BirdNet model for our back garden - now I'd quite like something a bit bigger, not to run 24 hours a day, but which can pick the individuals out of recordings made on trips away from home - what was that bird we saw iat the local water park?
[2] nor is it guaranteed to have wide appeal at home; but it is just *our* latest use-case. The point is, there are lots of projects, so scatter the promotions[3] and something will resonate with each person.
[3] look, do you want a big, fancy, easy to do, hype marketing campaign around one thing or do you want to sell lots of "different" PCs at all sorts of price points?