This was quite an interesting read and I loved the Chinese experiment!
You phrase it entirely well: LLMs currently are like having an intern! Itās āsomeoneā you need to cross-check, work with, teach! An intern can be very good on a Monday but tired on a Friday and the outputs get worse š
Every prompt is not final, itās an iteration. You need to keep improving it otherwise it just sounds like a text that is bland/ confusing or might be entirely hallucinating.
Thanks Francisca! I love using the Chinese Room experiment to disarm people who fear monger about the current levels of AI. It also sounds like prompt engineering will now change with the new o1 models rights?
This was quite an interesting read and I loved the Chinese experiment!
You phrase it entirely well: LLMs currently are like having an intern! Itās āsomeoneā you need to cross-check, work with, teach! An intern can be very good on a Monday but tired on a Friday and the outputs get worse š
Every prompt is not final, itās an iteration. You need to keep improving it otherwise it just sounds like a text that is bland/ confusing or might be entirely hallucinating.
Great read! Loved this!
Thanks Francisca! I love using the Chinese Room experiment to disarm people who fear monger about the current levels of AI. It also sounds like prompt engineering will now change with the new o1 models rights?