Woah this is a super interesting idea for an application of machine learning. I like to think that I have a good sense of where people use machine learning, but what sets this apart from the large majority of applications is its using machine learning to study another concept that requires predicting averages, rather than specific outcomes.

2. Another way I might choose to frame this discussion has to do with another way to think about uncertainty. Rather than think about where the “real odds” may be, one can think of it in terms of how much you should be willing to adjust your odds. For instance, one can be very certain in a prediction that a coin flip will be heads exactly 50% of the time. Because we can have high certainty, it would take a lot of information to change your opinion. If someone told you “they knew” heads only came up 30% of the time you wouldn’t believe them; and if you flipped a coin 10 times and got 8 heads it wouldn’t really change your opinion on the probability. But let’s say you have a weighted coin, but you don’t have any idea how it is weighted. Now if you flipped the coin 10 times and got 8 heads, you would believe that the coin is probably a little weighted towards the head. Thus, one can have a real understanding of what uncertainty in prediction means through a more application-based explanation.

To add on to that, I think we have in the world a large problem with under-funding research, especially in science. These activities have really large positive externalities, and if we fund them more we could easily accelerate advanced in everything from climate change to fighting pandemics to faster computers.

I think when it comes to the stock market, my bet it has to do a lot with unknown unknowns. It is probably impossible, even for a perfectly intelligent machine or being, to accurately predict where the stock market is going to go just based on past stock market data. I mean you could use it a little, but really only a little. So as a result you need to think about all sort of outside factors from politics to big movies coming out. It's really hard to know and classify all the different possible factors, which is I think why it is hard.

I am also very interested in the politics of climate change. I am specifically interested in a very odd paradox in the Democratic party. In one of the debates when the primary candidates were asked what the most important issue of our time was, by far the most common answer was climate change. Yet we have seen little to no effort to try to make a carbon dividend or some other major climate bill one of the major policy efforts of the Biden administration. I would have loved to get a specific answer to why some of the experts believe this paradox exists.

This has some interesting parallels in machine learning. While often times it isn't viable, quite often the gold standard for machine learning models is if you have lots and lots of historical data about what has happened before. For instance to recommend TV shows, Netflix simply looks at what people with similar watching habits have liked in the past in order to recommend stuff to you.