Uncertainty in Science -Stuart Firestein The most interesting thing I took away was thinking about our human perception of time and how it varies from person to person. It's crazy to think of all the factors that impact our perception of time by either slowing or speeding it up like for example pain certainly seems to slow it down. It was also very interesting thinking about life before clocks and how people coordinated on a large scale. There's no wonder that Stuart claims clockwork helped enabled the scientific revolution because it allowed for the mass coordination of people. Thinking about a typical modern-day, without a clock almost everything would be messed up. No alarm clock, no sense of meeting times, no idea when lunch or dinner occurs. A question I would've asked is how important is it for us to understand the mechanisms behind artificial intelligence?? Professor Goodman explains that a lot of AI we use gives us good recommendations without us really understanding the mechanisms behind the AI that made it generate these results. I think this is very powerful in helping us create new leads and breakthroughs but I think it can also get us in trouble. If we continue down this path enough without understanding specific AI results, it could create a future where we don't know exactly what we're building. This is specifically dangerous with AI because we need to constantly monitor its progress and make sure it goes down a path that we can control or a path that will in some way help humans. https://www.labxchange.org/library/pathway/lx-pathway:53ffe9d1-bc3b-4730-abb3-d95f5ab5f954/items/lx-pb:53ffe9d1-bc3b-4730-abb3-d95f5ab5f954:lx_simulation:9041b2ca?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction

Predicting Health, and Earthquakes -Susan Murphy & Brendan Meade The most surprising bit of information was this idea of hypothesis-free science. From very early on everyone learns the integral part a hypothesis plays within research and really any science experiment. You really weren’t even allowed to start a lab without first laying out a hypothesis so this was very interesting to see it backwards. It’s interesting that because we now have so much data and access to powerful computing that we can run almost black box algorithms to find interesting trends within datasets. We don’t need to start with an idea of what could happen but instead we can try and let the data tell us what’s going on. This also helps get rid of possible biases to confirm or deny the hypothesis before the study even starts. https://www.labxchange.org/library/pathway/lx-pathway:825945a0-367c-45dc-82b7-3d160c6e6f7a/items/lx-pb:825945a0-367c-45dc-82b7-3d160c6e6f7a:lx_simulation:5ad35586?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction Philosophy & Prediction -Ned Hall One question I would’ve asked is about Bayesian probability and this a-priori assumption we automatically bring into and update throughout any probabilistic event. I would just ask if these assumptions assume that humans are rational and therefore update e probabilities in a rational way?? I find this interesting because a lot of behavioral Econ like expected utility assumes humans understand probabilities and will act rational. However, as Daniel Kahneman points out in thinking fast and slow, humans actually don’t always act rational and often times chose opposing results because of previous states or biases. https://www.labxchange.org/library/pathway/lx-pathway:53ffe9d1-bc3b-4730-abb3-d95f5ab5f954/items/lx-pb:53ffe9d1-bc3b-4730-abb3-d95f5ab5f954:lx_simulation:8bf7271d?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction

The Search for Extraterrestrial Intelligence -Jill Tarter The most interesting bit of information was this tradeoff of human vs. robot space travel. Obviously, we need humans in order to troubleshoot and set up various missions on these space expeditions but our fragile bodies make it very difficult and dangerous. As Jill says, this limits space exploration to a few humans that're often accompanied by various types of robots or machines. It reminds me of the movie Interstellar where the team is accompanied by two human-sized robots. It makes you wonder how these dynamics will continue to change as we make technological breakthroughs and can more heavily rely on robots to do our high-risk missions. https://www.labxchange.org/library/pathway/lx-pathway:34dd3b2c-3aec-460a-817f-da4af2ed1577/items/lx-pb:34dd3b2c-3aec-460a-817f-da4af2ed1577:lx_simulation:e9099212?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction Prediction in Astrophysics -Avi Loeb One question I would've asked is how can we fix the current scientific research/discovery system in order to better align incentives? I think this is very important because ultimately incentives drive everything. If people are more concerned with their self-image, they're less likely to put forth research that may not be perfect. However, as we saw with the path to Newon, past work is incredibly important to make true breakthroughs and it doesn't matter if that work is completely right. By publishing research instead of waiting for perfection, it allows the collection of human minds to work together and more efficiently make discoveries. https://www.labxchange.org/library/pathway/lx-pathway:34dd3b2c-3aec-460a-817f-da4af2ed1577/items/lx-pb:34dd3b2c-3aec-460a-817f-da4af2ed1577:lx_simulation:1a066234?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction

Prediction and Psychology -Dan Gilbert The most interesting bit of information I found was this idea of the future being all there is to care about. Because the past can't be changed and the present is an infinitesimally small moment, that leaves us constantly thinking about the future. Whether that future is in 10 minutes or 10 years, it's constantly on our minds. As Dan Gilbert mentions, we consistently think about the future because it brings us joy and allows us to dream up possible scenarios. I think this brings up a tough tradeoff because we enjoy thinking about the future but we also need to live and act in the present. Without our actions in the present, we can't truly be happy or achieve anything we dream possible about the future. Overall, I think the ability to manage this tradeoff between the future and the present is key to living a good life. https://www.labxchange.org/library/pathway/lx-pathway:53ffe9d1-bc3b-4730-abb3-d95f5ab5f954/items/lx-pb:53ffe9d1-bc3b-4730-abb3-d95f5ab5f954:lx_simulation:5e3f229f?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction Behavioral Economics -David Laibson An unanswered question that I would've asked is first of all how can we better predict long-tail outcomes?? and also how can we better account for them in our modeling? As David Laibson mentioned, these long-tail outcomes are very unlikely but have massive impacts whether that's good or bad. Despite these massive outcomes, it seems that humans and our modern prediction systems focus more on the normal distribution of possible outcomes and in some cases, we've even begun to expect them (stock market returning ~8% per year). These assumptions and models become very dangerous over long periods of time due to the inevitability of long-tail events. Overall, I think if we can better account for these long-tail events, we can dramatically improve our models and predictions. https://www.labxchange.org/library/pathway/lx-pathway:b5121779-9f49-49db-93d9-80d5d67dadb3/items/lx-pb:b5121779-9f49-49db-93d9-80d5d67dadb3:lx_simulation:f10b9110?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction

Personal Genomics -George Church The most interesting bit of information I found was the idea of Moore's law in Biology. Moore's law is easily understandable looking at computers and other electronics over the years but its impact on Biology is less observable. It's very interesting that biological development and Moore's law somewhat tracked each other through time until genomic reading/writing increased by 5 fold due to multiplexing. I find this fascinating because the pace of Moore's law is fast yet our genomic sequencing ability is advancing at a faster pace. It makes you realize that the biological pace of innovation over the next 5-10 years will be unbelievable. https://www.labxchange.org/library/pathway/lx-pathway:0b417a9e-6227-44a8-a887-dbfdf44e37e3/items/lx-pb:0b417a9e-6227-44a8-a887-dbfdf44e37e3:lx_simulation:f6d09171?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction Epidemiology -Megan Murray I would have asked how we can better find genomic correlations across human disease traits? Megan brings up an interesting point about diabetes and TB correlation and I think it would be incredibly beneficial to look for correlations across any disease and any gene. Especially as our genomic sequencing ability continues to advance, I think this could play a massive role in slowing or even preventing the impact of certain diseases. It'd be interesting to know the diseases we already have correlations for and then also in ways we could start to add data and look for correlations in others. https://www.labxchange.org/library/pathway/lx-pathway:0b417a9e-6227-44a8-a887-dbfdf44e37e3/items/lx-pb:0b417a9e-6227-44a8-a887-dbfdf44e37e3:lx_simulation:7f50189c?source=%2Flibrary%2Fclusters%2Flx-cluster%3AModernPrediction

The Future of Energy and the Earth -Dan Kammen The most surprising bit of information I learned was about the idea of overconfidence when it comes to predictions from modeling. As Dan said in the video, the people who modeled were 50% overconfident in their predictions. This was due to their investment in the models and their desire to be right in their predictions. Using this info, they started taking model forecasts and doubling their confidence intervals. I find this very interesting because, despite the massive investment of time and energy that goes into creating perfect models, the human idea of overconfidence still creeps in and impacts the overall forecasts. It is a good reminder that we need to remain humble in our predictions and not get too carried away by certain models. https://canvas.harvard.edu/courses/84859/pages/modern-predictions-and-ai Climate Change -Gina McCarthy An unasked question I would've asked is that if renewable energy is the best way forward in order to solve climate issues while also improving society, how do we get from where we are now to a fully sustainable future? I would ask this because it's clear that renewable energy is needed and it's definitely good for society however, we still need to accomplish a lot before it could support our entire energy needs. We still have issues with the storage and transportation of clean energy, we need a drastic increase in lithium and other rare earth metal production and overall we still have a ton of infrastructure to build-out. Knowing of all these issues, how can we quickly transfer to sustainable energy while not limiting our energy demands. I think that this is an incredibly important question that will be crucial to our plan forward with renewables. https://canvas.harvard.edu/courses/84859/pages/modern-predictions-and-ai