Bowie knifes
- 0 Posts
- 5 Comments
doo@sh.itjust.worksto
Technology@lemmy.ml•Luce DFlash: Qwen3.6-27B at up to 2x throughput on a single RTX 3090English
4·7 months agoThe irony. Before llamacpp the only way to run llama was using other and on Nvidia GPUs. Then llamacpp expanded to other models, introduced gguf, added backends to run on GPUs and now we’re taking about running qwen using just python on a single Nvidia. Ouroboros is complete.
doo@sh.itjust.worksto
Technology@lemmy.ml•AI Is still making code worse: A new CMU study confirmsEnglish
23·1 year agoAs I’m slowly evolving my own flavour of spec driven development, I’m starting to think about the generated code as a secondary artefact where main quality criteria is that it’s doing what it needs to and it’s covered with tests.
I guess my current analogy is that I don’t care about how readable or dry is the assembly code generated by compiler.
I have the specifications and the working code with tests. I can always regenerate it if I need to.
But. I still read the produced code, steer the design and correct the obvious blunders. No vibes.
doo@sh.itjust.worksto
Technology@lemmy.ml•AI Is Destroying a Generation of StudentsEnglish
121·1 year agoI would say that we’ve just discovered that our approach to test knowledge and critical thinking was secondary to the means we used.
Apparently, essay writing doesn’t require either of those.
Which means we have to think how we teach kids. They will and should use helper tools, but just like we teach multiplication table before allowing calculators, well have to establish what students have to learn before they can use the LLMs in their studying.

Mathematically love and hate are equal. Just with different signs.