▲ 147 ▼ Jensen Huang says Nvidia engineers should use AI tokens worth half their annual salary every year to be fully productive — compares not using AI to using paper and pencil for designing chips (www.tomshardware.com) submitted 5 months ago by throws_lemy@reddthat.com to c/fuck_ai@lemmy.world 34 comments fedilink hide all child comments Nvidia CEO Jensen Huang said that he’ll be deeply alarmed if an engineer getting paid $500,000 a year does not consume at $250,000 worth of AI tokens to get their job done.
[–] Sanctus@anarchist.nexus 50 points 5 months ago (2 children) Spend half your salary on my ass! Or else! permalink fedilink source hideshow 4 child comments replies: [–] MountingSuspicion@reddthat.com 22 points 5 months ago (3 children) They don't seem to be suggesting they use their own money. Just using their salary as a comparison. It'd be like if a trucking company related gas expenditures to employee salaries. If they had an employee that earned 80k but only spent $20 in gas, that would be odd. They seem to be suggesting that in order to prove their working efficiency, they should be spending a certain amount of company money on AI. I don't agree with the assessment, but I do think it should be clear that they are not expecting that money to come out of employee salaries. permalink fedilink source parent hideshow 6 child comments replies: [–] Kirk@startrek.website 34 points 5 months ago (1 child) Reminds me of Elon's comment about programmers value being based on the number of lines of code they'd written. permalink fedilink source parent hideshow 2 child comments replies: [–] 100@fedia.io 16 points 5 months ago who cares about readability, lets use if statements for everything to reach maximum value, permalink fedilink source parent [–] Jhex@lemmy.world 5 points 5 months ago I think you are correct but using $$ as a metric makes zero sense... is AI less "effective" to engineer chips (as per Jensen) if gas prices raise and therefore the cost of cooling down the data centers? I think this shitty attempt to line up tokens with salary is based on their absolutely stupid idea to start paying and trading with tokens as if they were real money permalink fedilink source parent [–] Jarix@lemmy.world 0 points 5 months ago* Should be clear? In this economy?! permalink fedilink source parent [–] Prove_your_argument@piefed.social 0 points 5 months ago* This is an insane take. No paid internal engineer is blowing their own money on this kind of tooling. Nvidia pays for it. The same expectations would not be true for employees for normal companies either, instead the goal would be to have 2 FTEs and 1FTE worth of token expenditure, instead of 4+ FTEs. Same cost, "more productivity" (at least that is the argument of the people making money from selling AI hardware.) permalink fedilink source parent
[–] MountingSuspicion@reddthat.com 22 points 5 months ago (3 children) They don't seem to be suggesting they use their own money. Just using their salary as a comparison. It'd be like if a trucking company related gas expenditures to employee salaries. If they had an employee that earned 80k but only spent $20 in gas, that would be odd. They seem to be suggesting that in order to prove their working efficiency, they should be spending a certain amount of company money on AI. I don't agree with the assessment, but I do think it should be clear that they are not expecting that money to come out of employee salaries. permalink fedilink source parent hideshow 6 child comments replies: [–] Kirk@startrek.website 34 points 5 months ago (1 child) Reminds me of Elon's comment about programmers value being based on the number of lines of code they'd written. permalink fedilink source parent hideshow 2 child comments replies: [–] 100@fedia.io 16 points 5 months ago who cares about readability, lets use if statements for everything to reach maximum value, permalink fedilink source parent [–] Jhex@lemmy.world 5 points 5 months ago I think you are correct but using $$ as a metric makes zero sense... is AI less "effective" to engineer chips (as per Jensen) if gas prices raise and therefore the cost of cooling down the data centers? I think this shitty attempt to line up tokens with salary is based on their absolutely stupid idea to start paying and trading with tokens as if they were real money permalink fedilink source parent [–] Jarix@lemmy.world 0 points 5 months ago* Should be clear? In this economy?! permalink fedilink source parent
[–] Kirk@startrek.website 34 points 5 months ago (1 child) Reminds me of Elon's comment about programmers value being based on the number of lines of code they'd written. permalink fedilink source parent hideshow 2 child comments replies: [–] 100@fedia.io 16 points 5 months ago who cares about readability, lets use if statements for everything to reach maximum value, permalink fedilink source parent
[–] 100@fedia.io 16 points 5 months ago who cares about readability, lets use if statements for everything to reach maximum value, permalink fedilink source parent
[–] Jhex@lemmy.world 5 points 5 months ago I think you are correct but using $$ as a metric makes zero sense... is AI less "effective" to engineer chips (as per Jensen) if gas prices raise and therefore the cost of cooling down the data centers? I think this shitty attempt to line up tokens with salary is based on their absolutely stupid idea to start paying and trading with tokens as if they were real money permalink fedilink source parent
[–] Jarix@lemmy.world 0 points 5 months ago* Should be clear? In this economy?! permalink fedilink source parent
[–] Prove_your_argument@piefed.social 0 points 5 months ago* This is an insane take. No paid internal engineer is blowing their own money on this kind of tooling. Nvidia pays for it. The same expectations would not be true for employees for normal companies either, instead the goal would be to have 2 FTEs and 1FTE worth of token expenditure, instead of 4+ FTEs. Same cost, "more productivity" (at least that is the argument of the people making money from selling AI hardware.) permalink fedilink source parent