Thread

1. conception · 1789695698 · focus · HN ↗

   3.8 Max is the most “grounded” model I think - talks generally normal, doesn’t go crazy and start doing things (I see you Gemini), has good design choices and isn’t overly nitpicky. But god it’s slow. And only available from Alibaba. Their token plan is stingy too. If I had to pick the “old reliable boring” LLM, a modern Claude 4.5 if you will, Qwen is my choice. Hopefully they don’t RL it to oblivion.

   1. rubslopes · 1789696735 · focus · HN ↗

      > RL it to oblivion.

      What would that mean in this context?

      1. pennomi · 1789701049 · focus · HN ↗

         Tuning the model so far in the direction of being aggressively useful that it will quickly go off the rails in the name of helpfulness.

         I swear I spend more time telling Claude not to do things than telling it what to do.

         1. mdp2021 · 1789714175 · focus · HN ↗

            > aggressively useful ... in the name of helpfulness

            But is that because of training, or can that be (also? mostly?) an effect of the "system prompt"?

            1. girvo · 1789729117 · focus · HN ↗

               It’s absolutely down to their post-training RL, yeah. It’s where most of its strongest behaviour comes from, with regards to this kind of agentic behaviour
         2. vintermann · 1789716118 · focus · HN ↗

            I guess the agentic coding benchmarks don't have many rewards for stopping and clarifying what the user wants?

            1. disgruntledphd2 · 1789718734 · focus · HN ↗

               They do not, as they're aiming for full replacement rather than augmentation of human users.

               Personally, I think this is a bad idea, but someone's gotta build the Machine God I guess.

Unofficial Hacker News client; not affiliated with Y Combinator.