Hey, we’re actually on schedule this week! I won’t congratulate myself too much as the writing and reflection that accompanies each piece are of questionable quality. The formatting of these posts is also something I’m trying to standardize. The editing forms on Mirror are fairly easy to use in the moment but difficult to standardize across posts. Apologies as I iron out these details.
(RE)
The arrow of concern that this author draws regarding the utilization of polygenic risk scores for IVF and the need for great concern doesn't land with me. There is one brief mention of what I believe is this paper's true reason for concern, "These tests demand a broader societal discussion. By nature of their complexity, polygenic risk scores also open the door to evaluating not only disease risk, but traits such as height or intelligence. At present, not enough is known about the genetic contributors to such traits to develop meaningful tests that would allow prospective parents to select embryos. But those data are on the way and the technology is going to move quickly — it is well past time to discuss how far it should go."
This is the final statement in the editorial, but its message is strewn throughout the document. Concerns for the correct counseling of parents about the efficacy of PRS for specific diseases isn't the real issue here. It doesn't sit strongly enough as a concern. Sure, we can't say for sure what the actual probability distribution of your child's physiology w/ respect to these conditions is, so you should just shoot blindly and accept no predictive benefit and select embryos at random. If a parent has to make a decision, give them whatever information we have that is available to act on. Randomness is no better, especially at the point of selecting one of a few embryos.
De Gregorio, D., Inserra, A., Enns, J. P., Markopoulos, A., Pileggi, M., El Rahimy, Y., ... & Gobbi, G. (2022). Repeated lysergic acid diethylamide (LSD) reverses stress-induced anxiety-like behavior, cortical synaptogenesis deficits and serotonergic neurotransmission decline. Neuropsychopharmacology, 1-11.
"Learn to utilize systems that code for you!"
https://www.deepmind.com/blog/competitive-programming-with-alphacode
Li, Y., Choi, D., Chung, J., Kushman, N., Schrittwieser, J., Leblond, R., ... & Vinyals, O. (2022). Competition-Level Code Generation with AlphaCode. arXiv preprint arXiv:2203.07814.
(RA w/ some skimming)
I've spent a small number of hours, some 20-50, working on developing skills for coding in R and Python. At this time, I feel unable to competently write in either of those languages, but feel that I could utilize Stack Overflow or published code packages in R to some success. Seeing the progress in these predictive language algorithms moving into the software world only further inhibits my drive to become self sufficient. Along with the hard fact that my actual daily life will more than likely never require me to write code ever again.
[Credit to Two Minute Papers for the pointer: https://www.youtube.com/channel/UCbfYPyITQ-7l4upoX8nvctg ]
Cobbe, K., Kosaraju, V., Bavarian, M., Hilton, J., Nakano, R., Hesse, C., & Schulman, J. (2021). Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168.
Drori, I., Tran, S., Wang, R., Cheng, N., Liu, K., Tang, L., ... & Strang, G. (2021). A Neural Network Solves and Generates Mathematics Problems by Program Synthesis: Calculus, Differential Equations, Linear Algebra, and More. arXiv preprint arXiv:2112.15594.
(RA w/ Skimming)
This paper is only 7 pages in actual core content with significant supplementary examples and documents. Taken on its face, this is a remarkable accomplishment and holds great promise for schools and students. The ability to generate, grade, and evaluate collegiate-level mathematics questions (and eventually other STEM fields) is a great tool to have at your disposal. Imagine never having to write a new exam or worry that past exam information will be leaked and distributed by students. Imagine always being able to generate new questions that target your specific weaknesses with robust answers you can explore provided in real-time. This excites me greatly, for the more linguistically based fields I wonder if this is more or less helpful. I'd count biology and health sciences in this category and would be very interested to see how its question grading and generation would go.
Blanchflower, D. G., & Bryson, A. (2022). The Female Happiness Paradox (No. w29893). National Bureau of Economic Research.
(RA, with a moderate level of skimming of the whole paper)
Notably, they have a focus on, "wellbeing questions – on calmness, restless sleep., being cheerful, lonely, fearful, helpless, anxious, frustrated, tense and more...." with this focus they find that, "men have higher wellbeing than women. We also find that men are more satisfied than women in regard to broader questions about the state of the economy, democracy and the state of health services and education in their country."
Yet these questions, at least the first set, are focusing on internalizing processes that are already more prone to pathology in females. Gender differences in psychopathologies roughly track males leaning heavier toward externalizing disorders and females towards internalizing disorders (Anti-Social Personality disorder vs Depression as an example) and this could be one partial explanation for the divergence the team is observing. The downward trend they note in well-being is concerning, but I wonder if other social measures of despair such as drug use, suicide, assault, etc... wouldn't catch the population of males who are also seeing decreased well-being though they may still report less helpless and fearful attacks of anxiety.
A quote from the paper goes so far as to say, "Fifth, we show that in all of the various negative affect variables we examine, including anxiety depression, bad mental health days and many more, we confirm that women are always and everywhere more unhappy than men." On average, I'm not sure how much this should falter from our base-rate assumptions for gender differences in psychopathologies.
Garfinkel, I., Sariscsany, L., Ananat, E., Collyer, S. M., Hartley, R. P., Wang, B., & Wimer, C. (2022). The Benefits and Costs of a US Child Allowance (No. w29854). National Bureau of Economic Research.
(RA, with some data skimming)
I remain generally in favor of child tax credit systems. I'm always a bit concerned when reading things like, "The cost of this program is estimated to be 98 Billion dollars, but our models predict a returned value of 900+ Billion dollars." I think part of my apprehension comes from a past reliance on these figures when attempting to convince others of positions I held. If I would use it to nudge you closer to my position, why wouldn't you use it to nudge me?
The Fitzwilliam Endorses the YIMBY Movement
https://www.thefitzwilliam.com/p/yimby?s=r
(RE)
Choice Quote: "Apparently, we’re not in flying cars colonising the galaxy because we wasted all our money on rent."
I suppose I’ll take this time and space to mention that I am broadly pro-YIMBY as well. I live in a city where the negative externalities of NIMBY mindsets and the extreme anti-development mindset of the broad progressive movement are spray painted across our sidewalks and upon the freshly set paint of new higher density building projects. As a member of what is coming to be termed, “supply side liberalism” or “State Capacity Libertarianism” it is difficult to hear the arguments against development so confidently bandied about.
___________________________________________________________________________________________ Photo Credit:
https://www.pexels.com/photo/light-red-art-pattern-11280353/