Zev, I wonder if by 'ignorance' you mean it in a scientific context. If so, I am in total agreement for having such as course. In fact, a class like PredictionX would be very nice for a lot of unis to offer (since it places emphasis on known unknowns and unknown unknowns, in addition to other useful topics). PredictionX is available for free on edX so I think it wouldn't take much work for people to compile course materials. Perhaps the only obstacle would be to actually take the initiative to make it part of traditional high school and university curricula.

You made a wonderful point that did not cross my mind when I watched the interview. Namely, that Meade is limited on data unlike Murphy who has tons but an incomplete theory, which in part leads to both needing to rely on interdisciplinary work to make progress on their problems. I agree that we should champion collaboration as it is essential to make stronger predictions.

I believe the main takeaway of the essay for non-quantitatively-inclined readers should be the discussion of why consider uncertainty in the first place. Humans are not perfect and when we attempt to theorize, experiment, or evaluate observable phenomena there is no way our results will hold 100% of the time. However, having a notion of 'error-spread' or uncertainty is crucial to make this idea transparent and thus make better science and even everyday decisions.

I would also like to know the 'behind the scenes' of AI and particularly ML during the prediction process. I feel that despite computers increasing efficiency when it comes to data processing and later giving an output to be interpreted/analyzed by a human, it kind of gets rid of the fun it is to manually formulate, say, a mathematical theory like Newton's gravity.

This is such a lovely post that makes me appreciate the intersection between psychology, physics, and philosophy. I am very confused on whether we have free will and there are great insights from the quantum world and larger scales. Einstein modeled the universe as a set of points representing events in space and time (technically, a 4-dimensional pseudo-Riemannian manifold with a torsion-free connection---super interesting stuff and I recommend those unfamiliar to look into this!). However, these event points are inherent in the mathematical structure meaning all moments in space and time have been laid out and thus it seems we have no say on the spacetime evolution of the universe. At first glance, this can be in conflict with quantum mechanics due to the uncertainty principles but I second Eric's post: "If our exact actions can't be predicted due to inherent uncertainty in the system, would that imply that our actions are in part governed by randomness? If this is the case, then is that really free will either?". Sure, the quantum is probabilistic and the manifold does not appear to be so as we zoom out but both theories seem to be telling us the same thing in different ways, that is the nonexistence of free will. I am still unsure on whether I agree with this but at least there is food for thought.

This is such a wonderful question, Pedro. I agree with your points and I'm an advocate for continuous improvement of AI education. I am curious to know how you would address the subtleties mentioned in your post? There is no harm in brainstorming even if our level of expertise does not match with that of Schneiderman. I believe teachers would need to undergo rigorous training on AI, not learn the technical aspects but the basics on how it works and more importantly its social consequences. I also think the classroom experience should be discussion-based with students proposing, for example, methods to keep Big Tech accountable, the potential harm of deep fakes, or how to be aware of ever-evolving malicious online activity.