It still can't actually reason, LLMs are still fundamentally madlib generators that produce output that statistically looks like reasoning.
And if it is trained on both sides of the airfoil fallacy it doesn't "know" that it is a fallacy or not, it'll just regurgitate one or the other side of the argument based on if the output better fits your prompt in its training set.
And if it is trained on both sides of the airfoil fallacy it doesn't "know" that it is a fallacy or not, it'll just regurgitate one or the other side of the argument based on if the output better fits your prompt in its training set.