[icepat]

While the UX is good, I find the AI tends to fall into non-debate general OpenAI generated responses very fast. For example, I started with:

The covid vaccine was an objective good for society

And the model ended up arguing against me in tone only, but saying that yes, the vaccine was an objective good for society. It did not even try and debate, so this is useless for testing arguments for things that the network already agrees with. This was the case for many other arguments I tried to put against it.

While the UX is nice, this does not work at all.