@ricraftis I would imagine that the OpenAI products are backed by massive hardware and specialised software that makes it respond so quickly. Crucially important for their use case as a chatbot.
@simonroses If I were to guess, I would say that experience will be as valuable as ever. At the current state in LLM development, correctness is not guaranteed. Humans will need to work in the loop with the AI to produce correct code. It all feels uncertain however, even next year, this could all change.
@DToher@garius The article reads well. I feel like it would benefit from a real world example of a trust thermocline failure to help support the article?