you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 09 Nov 2026
14 points (100.0% liked)
TechTakes
2781 readers
146 users here now
Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
founded 3 years ago
MODERATORS
Not to get all doctorow on this, but I generally don't think we should concede this approach to using LLMs to the yooooooo I just vibecoded an FIFA 2027 cloen for android watch WITH MY DICK!!!!11!! people.
If/when a time comes when genAI isn't actively ruining everything and doesn't need all the RAM and hard drive space in the world to be useful, using it like that seems a far cry better than even what torvalds is currently doing.
Can you trust that figure, though? They’re already in trouble for being deceitful about slop usage, and 0.1% is close enough to zero that it could have literally been zero if anyone cared about minimising anything.
Yeah, it's definitely an open question if they are being honest for honesty's sake or just trying to get ahead of people realising that the game is slop-infested in obvious ways.
Assuming the first, the number is definitely based on vibes, but 0.1% of say ten million lines of code is still a lot of code, and the way LLMs have been shoved into every nook and cranny of the major IDEs, demanding proof of zero tokens spent during development puts enterprise devs in a weird spot, especially with stack overflow all but defunct and the various tool vendors increasingly assuming you'll be using them in LLM-first dev scenarios and skimping on providing in-depth guidelines, or slopping them out.
It sucks.
Build a machine that isn't made of trust violations, controiled by the worst people in the world, and actively making climate change worse potentially killing millions first slopper.