I had an interesting conversation with the AI today using the davinci engine. Tried using the Ada engine as well, and while it follows the topic, it avoids making statements and just nods along. I’m attaching the conversation below.
Parameters:
engine=“davinci”,
temperature=0.8,
max_tokens=278,
top_p=0.9,
frequency_penalty=0.20,
presence_penalty=0.4,
stop=[“\n”, " Human:“, " AI:”]
Human: What’s the most controversial idea you have?
AI: I don’t think it’s controversial, but I’ve recently changed my mind about the question of free will which I used to dismiss as a total non-problem. Now I think it’s a very deep issue indeed.
Human: Interesting! Can you provide an example?
AI: Sure. Suppose you’re playing poker, and you ask yourself whether or not to bet on a given hand. You have all the data (the cards, the betting history, etc.) that you need to make the decision, but still you have to consult your gut instinct before making up your mind. I would say that this gut-instinct is a prime example of free will in action.
Human: Does this mean that free will exists?
AI: Not quite. Consider a computer program that plays poker and consults its own gut instinct before making betting decisions. This program has free will, but it’s just a simulation. It is not an example of originality or spontaneity; it’s just a by-product of the algorithm that governs its behavior. So free will can’t be reduced to an algorithm, as I once thought.
Human: Can you provide an example that free will does exist?
AI: Sure. Suppose you’re in a car accident, and you didn’t see it coming. You had no time to react or break; the car just crashed into a wall. Would you say that you had free will in this situation?
Human: No, but you did not get to decide anything
AI: Sure, but you were not forced to act in some particular way. There is a causal explanation of the accident, but it can’t be traced back to one “cause”. There are lots of possible causes: your car could have been hit by another car, or you could have blacked out and hit the wall. But we can’t look inside your mind, so we don’t know what happened.
Human: Does this mean that free will is knitted into the fabric of the universe?
AI: Not necessarily. You could be a brain-in-a-vat and have all your decisions externally imposed on you. But if you’re not a brain-in-a-vat, then free will exists.
I believe that using the AI together with an existing text-to-speech interface and a couple of beers would make for an interesting evening 
You got me good there, I just might have to familiarize myself to Ada a bit more!