Oddly, I am at least tangentially familiar with all 3 of the research areas presented here, and they suffer from 2 very different problems.

AI has political and policy ramifications, but they don't hinge on the underlying mathematics of neural networks. No one's on a megaphone chanting: "What do we want? Backpropagation! When do we want it? NOW!" In climatology and economics however, the policy recommendations ate the underlying field. The line between researcher and activist wasn't just blurred but obliterated.

The public didn't conclude that "Hansen or the IPCC is [part of] a young field with tentative conclusions" from historical ignorance. They concluded that because (within their lifetimes) too many researchers (Hansen included) made wild, unfounded claims that proved false too often. Too many media outlets took IPCC worst case scenarios as likely outcomes. Wild claims get grants and sell newspapers, but they also undermine the science.

Austrians ran the economic show for far too many decades. Again, actual research was subordinated to a political ideology. In econ in particular, the researchers seemed to see themselves almost as oracles or Platonic philosopher kings.

Historical ignorance is not the core problem here. Post Cold War, too many academics forgot that scientists work best when they're "on call" for policymakers, not when they're "making the call" on policy directly.