One of the areas unlikely, in my view, to ever be replaced by AI is that of coming up with brilliant ideas. It’s so difficult to define this process because inspiration often has no logical basis other than perhaps, in some cases, being “obvious” once expressed, but not obvious enough to have been expressed by someone without that moment of inspired thinking.
After recent issues with Home Secretries here in the UK, culminating in the recent sacking, a friend & colleague has used ChatGPT4 to summon into being the HomeSecretaryBot™. This is truly inspired idea IMHO. They have then asked HomeSecretaryBot™ to consider certain recent(ish) questions so that we can compare the Bot’s output with what actually happened.
The results are not perfect; some experimentation needs to be undertaken with the initial instructions to make it clearer that decisions rather than just opinions are required. Often the process of defining the initial parameters is a bit hit & miss.
A final observation from me: if you read the dialogue, notice how the human is extremely polite to the AI. This might seem unnecessary, but I think that this is actually really important. People who are nice tend to be polite, gentle, and respectful in their interactions with others – whether the “others” are colleagues, children, animals, or anything with real (or even aparent) sentience. We should treat AIs with this level of respect for our own long-term wellbeing.
The details – including how to Build Your Own – are all in this excellent and fascinating article by my colleague which can be found here.
Gânduri Eclectice - Eclectic Thoughts 