With higher interest rates there’s less venture capital around to pour into the data centre build out, so tech companies are cutting back on their wage bills to channel money into building data centres.
AI might give some jobs a productivity boost, but it’s more like: 90 people can do the work 100 people did before. It’s not like, oh we don’t need these 10,000 employees in mumbai anymore because AI is developing that product now.
AI companies are switching to token based billing. Money is running out and heavy AI users are going to have to pay closer to the real price. People suddenly having to pay 10 times as much. You can just as well hire an intern for that money.
Honestly, there is little evidence that it actually has (successfully).
Of course, anything at a world wide scale will be true at some percentage; however, I doubt AI has actually replaced the labour of many. Even Meta gave up on pretending this and they are now firing people just to free money to pour on AI, not because AI is actually replacing those people
yes, it’s when you tell boss, “I can’t do that in 3 hours, it’ll take two weeks”, and probably still have some unknown aspects of quality, that we might not want to sign off on. Maybe we can rush it in 1 week, if you’re ok with want want maybe’ 20% unverified.
Boss fucks off to coprolite - gets it “done” in 3 hours. Gives it to someone else to QR. They comes back to me for advice on turd polishing (apparently that’s my SME). So I then waste time helping that person tactfully create a quality report that says it’s seriously defective and will take weeks to rectify to get it up to an acceptable standard - because it tells us nothing about how it got to it’s erroneous output.
Now, we’ve wasted about a day between us, on dog-shite - and we’ve not learned anything useful.
I don’t know what a “gen Z” is though, but whoever they are they should stand up to shite bosses.
IDK how many jobs AI is actually taking?
With higher interest rates there’s less venture capital around to pour into the data centre build out, so tech companies are cutting back on their wage bills to channel money into building data centres.
AI might give some jobs a productivity boost, but it’s more like: 90 people can do the work 100 people did before. It’s not like, oh we don’t need these 10,000 employees in mumbai anymore because AI is developing that product now.
AI companies are switching to token based billing. Money is running out and heavy AI users are going to have to pay closer to the real price. People suddenly having to pay 10 times as much. You can just as well hire an intern for that money.
So you’re saying AI is not replacing jobs at all?
Inference providers have been charging per token forever.
Honestly, there is little evidence that it actually has (successfully).
Of course, anything at a world wide scale will be true at some percentage; however, I doubt AI has actually replaced the labour of many. Even Meta gave up on pretending this and they are now firing people just to free money to pour on AI, not because AI is actually replacing those people
I think we’ll also see increased burnout as employees are expected to produce more with less.
yes, it’s when you tell boss, “I can’t do that in 3 hours, it’ll take two weeks”, and probably still have some unknown aspects of quality, that we might not want to sign off on. Maybe we can rush it in 1 week, if you’re ok with want want maybe’ 20% unverified.
Boss fucks off to coprolite - gets it “done” in 3 hours. Gives it to someone else to QR. They comes back to me for advice on turd polishing (apparently that’s my SME). So I then waste time helping that person tactfully create a quality report that says it’s seriously defective and will take weeks to rectify to get it up to an acceptable standard - because it tells us nothing about how it got to it’s erroneous output.
Now, we’ve wasted about a day between us, on dog-shite - and we’ve not learned anything useful.
I don’t know what a “gen Z” is though, but whoever they are they should stand up to shite bosses.
“Here’s an expensive chatbot that lies exactly 73% of the time. Now stop reviewing and testing code.”