4 ms·
100% agree. My impression is that the open-weight models have been drawing close-to-level at coding tasks, while Anthropic and OpenAI have been putting large a
by mft_ 3mo ago
100% agree.
My impression is that the open-weight models have been drawing close-to-level at coding tasks, while Anthropic and OpenAI have been putting large amounts of effort into developing their models' abilities in other domains: legal, biomedical/science, etc. Anthropic (especially?) has also been putting more obvious resource behind optimising their harnesses - from Code to Cowork (which is kinda Code for normies), Design, etc.
- _pdp_ 3mo agoGLM 5.2 has replaced "normie" agentic workflows previously backed by Sonnet and Opus. So I don't know. From my end it seems to me they are perfectly capable of working agenticly.
- mft_ 3mo agoMaybe we have different definitions of 'normie'. I'm talking about people who aren't in IT, and who are maybe just learning to use LLMs for aspects of their daily work. These people only know of the big three models, at best - they very rarely know of the open-weight models, and would even more rarely (given their model access is likely determined at a corporate level) be able to access them.
- _pdp_ 3mo agoThat's my point too. If you take GLM and call it ChatGPT or Claude Opus is anyone going to notice? If you are not into agentic AI I would argue that the model type makes zero difference for day to day use because GLM 5.2 is hitting the benchmarks hard. Now for a specialised use case (narrow fields), say cyber, Mythos is possibly better.
- mft_ 3mo agoGotcha. Agree that it's likely that GLM 5.2 could replace Claude/GPT in many easy-to-mid-level tasks. Given there's a (lot of?) uncertainty around the confidentiality of A/O amongst businesses even just looking at adopting Claude/GPT, it will be interesting to see to what extent the (mostly Chinese) open-weight models start to get broad usage.