From a former Stadia user; latency was never an issue or noticeable with a 4K stream. And I've played quite a bit of fast paced shooters in the platform.
The experience is extremely dependent on location, bandwidth, local setup and availability of close services (in my case, the closest DC was <15ms away according to Stadia telemetry).
His process is impressive, no sketches or hesitation; he just "prints" whatever was in his mind. The scenes are incredibly detailed, with complex perspectives
While I think humanity is bad at predicting what the future will look like (where's my damn flying car..), _my_ prediction is that AI video generation will not be a major disrupter to ad agencies / film makers.
It's a nice experiment, but I really doubt the level of artistic direction required to meet specific customer requirements will ever be replaced by "AI"
Well.. I don't actually think the article claims a 'cure' either.
If I get cancer, have treatment for it and am now in remission, who would think these doctors 'cured' cancer? Everyone would understand _my specific instance_ of cancer is reduced to being not a concern anymore.
Being in remission just means a greatly reduced or almost inexistent dependency on insulin.
I'm not saying the conclusions of the article are correct either (low sample size, short period of time), but it's a far cry from claiming 'LCHF cures T1D'
What? How is a user supposed to show you and reproduce their problem when you're connected over SSH or RDP? Both run in a different session than the user's