🤖 Some chatbots ate shit again 🤦♂️
● Claude: “…and it means my earlier answer needs a correction.” After another question: “I’m going from memory on the exact code, so check the ISCO-08 index if you need precision.” “I also need to correct my earlier answer.”
● ChatGPT: First, it calculated based on my question, because “it’s perfectly plausible.” When I told it that “it needs to be calculated based on specific mass,” it said, “Yes, but a distinction must be made between true density and bulk density,” then it went off on a tangent. When I asked, “BUT WHAT ARE THE VALUES for true density and bulk density? Not what I said, but what’s the reality?”, it said that what I said was a bit low compared to the values “indicated by some experimental measurements,” after which it performed calculations proving that my suggestion was right in the middle of the range. I had to press him: “So first you tell me that 150 g is too low, only to calculate then that it’s exactly in the middle of the range—so my rough estimate was correct?” Finally: “Yes. Your initial estimate was correct. I was contradictory: etc.”
● Gemini Argon: Twice in a row I asked it something with screenshots attached; both times it had that animation meant to suggest an hourglass for ages, only for it to do absolutely nothing afterwards! Qwen 3.8 Max fixed it for me.