6 ms·
The explanation is definitely unsatisfying. If I had to guess it was an issue with quantization. OpenAI moves so quickly with their product and doesn't seem to
by laborcontract 3y ago
The explanation is definitely unsatisfying. If I had to guess it was an issue with quantization.
OpenAI moves so quickly with their product and doesn't seem to be shy about quickly pushing changes to prod. There are too many times to count that, using ChatGPT, I've seen the frontend changing before my eyes or seen responses change as backend modifications are pushed.
On one hand it's refreshing to see their lack of aversion to push changes. On the other hand, it leads to probably one of the most unpredictable experiences I've had using any consumer product.
- p10_user 3y agoDoes the API offer "fixed" models that don't change? I was under the impression they did.
- Rastonbury 3y agoI find that the most annoying and borderline unethical, hiding changes and nerfs behind the fact the output is probalistic especially when I'm paying the same amount for the product and others are paying for and using it in production. As many have mentioned, GPT-4 ability to code is not the same as 6-8 months ago