Ask HN: Will AI LLMs for coding get smarter and cheaper in the future?
2 points
2 hours ago
| 3 comments
| HN
bdz35
2 hours ago
[-]
We are still in the brute force era. Lot of unnecessary compute is happening and that will keep reducing. But the bigger headache with cheaper and smarter is people start running faster and faster in different directions. So cost and complexity moves to keeping things in sync the larger the team gets.
reply
PaulHoule
2 hours ago
[-]
Yes
reply
verdverm
2 hours ago
[-]
It's already happening, see qwen/gemma sized models in the sub 36B category
reply