I was curious about what in the field of climate change according to him is the most important unknown? I understand that is not his direct field, but since he teaches a class on the unknown in science, it seems like there remains a lot of differing opinions regarding the key questions that remain unsolved in science regarding climate change? Also, I am curious regarding the progress over the course of our lifetime in mental health research, are there many unanswered questions where we have no idea how to begin or we are aware of how to answer but it takes time to explore all the edges? I ask this because naturally I see a lot of my friends and also old people struggle more with mental health in the modern world especially depression, anxiety, and trauma than I do with physical illnesses. Link

His class on ignorance sounds absolutely fascinating. As a non-physical science student, whose last signficant experience with the Chemistry, Biology, and Physics was in high school in a memor-intensive class, I found this class to be so important. As a high school student, I knew economics was deeply flawed but I did not see many other fields that was open for reinvention and comfortable in uncertainty — perhaps too comofortable and too arrogant. Nonetheless, I think a large part of what kept me going through many boring economics lectures is because I knew at the end of the journey the questions I will get to ask are all about the unknown: how will the Indian economy grow? how do you lift people out of poverty without creating staggering new levels of inequality? how do you educate a billion people? I am now reflecting that in high school it did often seem to me a lot of science was already known. If I had known several key question remains unexplored and that it is a realistic possibility that I could be studying in pursuit of answering such a question, which I am sure is the case, I would perhaps have remained far more interested in at least one of those three sciences. Link

I am incredibly fascinated by Murphy’s work in trying to predict stress. And her interest in engaging in interdisciplinary work is understandable given the need for a variety of testable theories that she can leverage her immense data to evaluate. However, one must ask whether this is indeed a futile effort. After all, in my reading of the stress literature in Neuroscience, it seems to be that a whole wide variety of possible triggers can heighten stress. And not all stress is equal. It is possible to predict when a trigger might occur, but it seems challenging to live al life where you are avoiding all possible triggers. Surely, the goal of such research should be educating people on their triggers well in advance of when the next trigger could be without any information on when the next trigger will occur but just that they must be aware that this exists and here are some of the best practices in order to handle the trigger when it does come. It is like being a pilot. A pilot can try to use the most advanced weather systems in order to avoid possible thunderstorms, but sometimes it is just enough to know that there will be thunderstorms because the pilot has been trained and the plane has been built in such a way that it can handle the thunderstorm. Link

Personally, the main takeaway from this course apart from appreciating and applying the core framework (the one where inputs, prediction, and evaluation are all mentioned) is the role of theory in our predictions. The modern rise of ML and Big Data have surely empowered Professors like Susan Murphy to answer human problems with data-backed solutions that previoulsy would have been the subject of many experiments of doctors or phycologists usually in societies with lax ethics rules regarding human experimentation. Yet, it is telling that some of the most important work occurs when employing a theory that many smoke in order to relive stress, so how can we provide them an alternative to this method of stress-relief. This is something we have known for a while and any smoke and any non-smoker with smoker friends observes this with little to no effort. And indeed she mentions unlike Meade where Earthquake predictions struggle with limited data and limited, Murphy is working with a lot of data but potentially not enough theory, so her focus is interdisciplinary work in order to test and develop theories that can focus the direction of her data projects as well as ground her work in explanable theories. This in fact might demonstrate why we need interdisciplinary work to begin with. We have a lot to learn from other people in regards to theories their fields have developed and tested that may be relevant in our predictions. Link

I am very curious that her question about how did you manage your technological transformation without blowing yourself up assumes the intelligence that we meet outside of the planet earth will be much more advanced than us technologically, and I am curious why she thinks this is the case? Going back to the previous comment where we do not know what we are looking for, but it is highly possible that the other life form in a different planet like many animals in this planet has different sensory abilities compared to humans, so our animals are much more aware of when natural disasters are going to occur compared to us, but we do not have such an ability. Apart from the sheer curiosity, is this question about surviving technological progress the primary reason for engaging in this search? Link

The most surprising thing that we learn in this video is that in the search for extraterrestrial intelligence a key challenge is we have zero examples of what we are looking for. If I am searching for my mother's phone, and I have no idea how it looks like, I still know it will be a phone of some kind. However, what if she comes from a different planet and their phone actually looks exactly like our hearing aid but much smaller. I could be looking for her phone, but I would not find it. This is especially the case because once you "find" something, you also want to check if that is indeed what you found, but here again, it is hard to know if you were correct that this intelligence because you must find an objective way of verifying that this is indeed true and that might take a very long time given how large the universe is and how far away the potential life form might just be. It is also fascinating how the search for extraterrestrial intelligence exposes how we have still so much to learn about our own planet: they talk about that we only recently learned that lightning moves upward from the top of a thunderhead or that now we know there are at least 150,000 forms of plankton when we previously thought there was only 15,000. Link

A question I was curious about was relating to his comments that people probably will not understand the severity of climate change by appealing to fear or appealing to scientific statistic models with ever-decreasing uncertainty as by then it will be too late. However, they are far more likely to be persuaded to herd behavior like other cities or states or countries are doing and the impact it has on them. My question is what if some of these actions are inherently preventative of some bad future but only works if others do too, so it could be Europe did those things but we would not know till climate change fully came upon us that they survived, as it is more likely that all of us even those who took preventative action will get hurt? I assumed people are more likely to herd behavior when such behavior is considered as the cause for one's success. Developing countries looked to the west and said we need democracy because that is why these societies are successful. Today they look at China and say we need less democracy and that is why China is successful. Thus, I am not so sure how herding can be a persuasive agent for pushing people to approve of large congressional actions as opposed to just public awareness campaigns demonstrating that this is something we must act or else (aka fear!). Link

I really liked his comment that humans are prediction machines trying to figure out what will make them happy in the future. That is an interesting point because a lot of happiness surveys demonstrate that people actually do know what will make them happy but since they do not think about the few things that make them happy on a daily basis then they often simply forget to do it and remain in inertia. Yet, Dan is also accurate in saying that we might think we know what we like or don't like (losing a job, for example), but people also a few months after that event happens might look back and say the best thing that ever happened to me was something a priori you thought was obviously bad. Also, I was surprised that the species he chose that also thinks about the future was not dolphins or chimpanzees as he mentioned earlier, but rather corvids who are able to conceptualize a time that has not yet come based on previous actions by the experimenters. That is truly remarkable. Nonetheless, my favorite quote from this video was "humans exist because they have an appropriate level of denial. And that people will have too much fear about certain things, can't even function in everyday life." Being a student of political economy, some of the literature that we read has high predictive power for the nature of current regimes across the world, and their unsustainable nature should make me quite fearful but I still maintain a denial that we could be wrong so there is no need to get totally scared just yet. Link

I did have several questions for Ben. I understand that he felt quite confident about his 100-year prediction that we will look back at these silly humans at the start of the 21st century who thought the AI will become smarter than us and will be better than us in every way because by then they will truly understand that AI is just a tool and will come to celebrate and appreciate the many differences that are there between the two agents. However, on what grounds is this prediction being made? Is he uncertain that if there is one particular field in AI that is met with an exogenous shock that transforms that industry and solves their critical problem in advancing the ML tools to become more human-like? What if new forms of legal innovation (similar to how the patent system encouraged innovation) fuels a new industry of novel AI applications that enable new forms of responsibility-sharing for such items even if the company that initially designed it does not feel comfortable taking the entirety of the risk? One must also wonder how the job of a pilot over time has become less necessary for the functioning of the plane even with incredibly high complexity and frequency and stakes of the project at hand? Are a lot more jobs going to occur where the tool is supposed to most of the job and the human is there just to oversee what warning labels are occurring?

I truly enjoyed this interview. Ben's insights were truly remarkable. The most prominent and recurring theme was language. As an economist who is deeply invested in understanding the future of innovation and the implications of that on job growth and distribution, this particular care towards language is of huge significance for economist and politicians communicating how their policies are helping workers who are rightly highly insecure about the age of the robots taking their jobs. The emphasis that they are our partners and that machine knows x or y or that it generates the ideas only reinforces this anxiety. I think thus this focus on language is of high significance. Additionally, I found it immensely interesting in understanding his approach to account for the various levels of certainty needed in different applications of this particular 'tool.' Here, he mentioned that for things that are novel (such as giving a promo for a particular item like 'Target knows you are pregnant but this was not something that was given before) then you probably need to have a much higher certainty even complete certainty. However, for items where humans are currently the agents of prediction such as doctors then what matters is who takes responsibility for failure even if on paper it seems like the computer is performing better (he was happy to accept the machine as a tool of competition if the company that designed it will take the liability of the tool failing). Finally, I am immensely curious about the idea of white-boxing ML. My personal experience playing around with ML in Fall 2019 when I took a data science course was marvel at the predictive capabilities but sheer shock in the black-box approach.

This is for Dan Kammen. I was most curious about how do we communicate that there are certain high-consequence but not necessarily low-probability events that we know will occur but we do not know when and we also do not know exactly what the 'exact' impact of such events will be — would a public awareness project focused on steps that could be taken now that would dramatically reduce the consequence of such events be the appropriate framing? Is such a project necessary given perhaps we do not want to make people panic but rather we want them to feel like there is grave danger ahead but it is preventable if we all just work together? As pointed out in the video, maybe the point of modern predictions is just to create legislative action in the necessary direction, and then after a few years, develop some more predictions highlighting the danger of these high-consequence events? After all, perhaps worrying about these events may be like worrying about studying calculus when we still have not mastered basic algebra: our government policies are decades behind on the most basic necessary action. Additionally, as a final question, I would ask is this something that due to the nature of these events we will never know the probability and impact of such events or that sanctioning future study in these directions will potentially enable far more clarity? Link.

This was a highly illuminating conversation. Firstly, I was most interested in performance contracts. As someone from the economics field, I am incredibly interested in new legal innovations in helping develop new market incentives that can transform the nature of our market economy. This particular innovation focuses on how to reward not just the cheapest bid at the auction but provide incentives for one to create durable projects where the rewards over time of such high-quality build are shared with both the buyer and seller. Secondly, collaboration among different disciplines is necessary for tackling an issue as critical as climate change, so it was interesting to listen to him elaborate upon one of the most important challenges of modern collaboration: the leader of the group will dominate the direction of the project, so even if one includes different disciplines in the team, the overwhelming influence may and has been so far from leaders in the STEM field. The two scientists in conversation clearly agreed they need more 'humanists' to lead and provide the direction. Thirdly, I am very interested in how societies verify accurate information. In the modern world, especially since we need a lot of cooperation from society to both trust science and demand their governments to take action based on said science. He talks about how people in the slums of Nairobi are able to verify and spread the word that someone is selling polluted water but they are more than ready to believe blatantly false information they encounter online. This insight must be acted upon and others such as this must be generated in order to understand how to win the PR war against the vested interests hoping to delay and/or block necessary government action. Link.