@anderslund @engedal Nej for sataan da - det må du ikke. Jeg lod et par lokale instanser af Mistral og OSS-GPT gå bananas på hinanden, og lige nu er de ved at bygge en tidsmaskine i baghaven!

@openrisk sorry to hear that, how much has the traffic load reduced ever since you implemented that feature?

@tommihow many r's are there in "strawberry" ?

@grepe just wow. That is incredibly disrespectful.

@AnnieBuddy @chris @josie_osborne @votetanille @avilewis Indeed ... let's rethink the limits on those Limited Liability Companies.

@Schneier_rss How many times must it be said: This technology is tailor made for fraud & almost nothing else.

@malte also recommended! Wikipedia:Signs of AI writinghttps://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing

@engedal tror ikk der er den store forskel men jeg ved det ikk. Vil selv helst bruge europæiske modeller men kender ikk detaljerne i forskelle. Det kommer forhåbentlig m ngt seriøs benchmarking på netop de punkter du nævner

@Yendolosch @emacsomancer The use of "hacked" in that headline is a bit self-aggrandizing?

@pseudonym We are using AI inexactly the worst ways possible.Caveat: I am a never AI-er, due to the ethical issues surrounding how training data is gathered, the severe ecological and economic impacts, and the fact that deepfakes are objectively making the world a shittier place.But pretend for a second, none of those are a problem anymore. We are still using AI wrong. You don't have it produce a mountain of code and have a human review it. You still use humans to produce the code, and have AI help other humans to review it. AI isn't terribly good at writing code, but it has been shown to be effective at finding a few classes of bugs humans are typically very bad at finding.But that won't allow you to fire people and replace them with monkeys on typewriters, so it'll never happen.

@budududuroiuusing the #noAI tag for this take is basically like using the #vegan tag tell everyone how good you think bacon is

@anderslund Jeg kiggede kort et par af de andre videoer, og de lader til at starte med en person der snakker og bevæger sig imod kameraet, men derefter skifter det (ret brat) over til en anden stemme og andre slags optagelser.Det er samme teknik hele vejen igennem ligner det.Mystisk at det originale toot'er ikke lader til at synes der er noget galt

@irkan Depends. Mistral is fine for general use, especially in English, but ChatGPT has the edge once we switch to Danish. Image generation is way better when using ChatGPT.Both have become pretty limited as long as you are using them for free, so in general we have switched to a local gpt-oss-20b on Ollama here at home, with Mistral and ChatGPT as a backup when the local AI isn't sufficient.