>>
|
No. 39140
[Edit]
>>39123
Agree strongly on the state of technology. In my opinion there have been no meaningful advances since ~2010 or so in the consumer sector. Consider that you a desktop (or even high-end laptop) from then is perfectly usable now. And the state of the developer ecosystem has gotten _worse_ since 2010 in my opinion. My predictions here are very dismal and cynical: greater shift away from "personal" computing into walled-garden as a service, more javascript bullshit, etc.
I really hope there's another tech crash soon, because I'm sick and tired of inanities (web3, iot, etc.) disguised as progress. In fact I bet there will be, because VC money isn't infinite and there'll be a reckoning sooner or later.
I don't expect adtech and the big tech conglomerates to die just yet, at least not within 10 years. They have too much momentum and so far the populace doesn't really seem to mind. There is increasing murmurs of discontent among devs and power users, so I do expect them to lose quite a bit of mindshare and they've already started their decline from their zenith (e.g. google of 2014 is nowhere close to google of 2022, both in terms of internal and external perception). Near the end of 10 years they will be in a prime position to be usurped by some new upstart.
I do expect that some startup will create a search engine that will beat Google at its own game within the next 5 years by catering to powerusers and building off of that momentum. Google is no longer focusing on search quality, and they're instead optimizing for mobile users in order to chase that sweet ad revenue. This will be their achilles heel, and distributed systems have become commoditized enough that Google's tech stack which was once their secret sauce is no longer a sufficient moat to prevent competition.
On the research side, I expect the hype around ML to slow down. There's been some really terrific achievements in the past few years alone, but I'm willing to bet the next breakthrough will come by taking a step back to develop a more unified theory of _why_ these deep neural nets work the way they do, and why some architectures perform better than others. "Geometric deep learning" is something to keep an eye out for here. I suspect there are also strong ties to information theory here.
As for other CS fields (since that's all I really feel qualified to comment about), there is a decent chance of there being a shakeup in the hardware world. There's a lot of renewed interest in chip design, and while sadly most of it is focused on hardware accelerators, I'm betting some of that will nonetheless end up being put into good practice. We're already seeing a switch away from x64, with of course not only apple's m1 but also aws offering graviton and such.
>>39124
I have no idea where all this hype around AR/VR came from. Just because Facebook insists on this "metaverse" doesn't mean there's actually anything behind it. Whatever the result is, I suspect it will be similar to the whole "smartwatch" thing – overpromised and underdelivered. There will be a place for it – AR especially – but it will coexist in the context of the rest of the tech ecosystem.
|