Pedro Domingos

The optimal carbon tax is a function of the size of the externality, which is being massively overestimated. The economists are being led astray by bad climate models. Got it?

I really liked your book and many of your insights about AI, but when it comes to the costs of climate change, I put more weight on the estimates of people who have studied this question carefully, like the IPCC and Bill Nordhaus. https://report.ipcc.ch/sr15/pdf/sr15_chapter2.pdf …

And FWIW, the externality from burning coal is overwhelmingly bad even before including any cost at all from climate change. Particulate pollution is deadly.

It would be well worth if we had no better choices, but we do (natural gas, for one). Or would you rather go back to pre-industrial times?

You and I agree then, with Nordhaus et al, that (since we do have better choices) we'd be better off without coal electric generation. A Pigouvian tax would help us get to a better outcome where this category of energy production were replaced.

But you don't seem to agree with me or the market that natural gas is better than renewables.

I’d I agree with the market if it included all relevant externalities. I think that is your position too. I think we have different views on the externalities. And while I agree that experts can be wrong, I would double check my analysis if they overwhelmingly disagreed with me

I have, and they don't (that's the impression you get if you read the media instead of talk to the actual experts; again, reading "Unsettled" should be a real eye-opener).