My horrible behavior of leaving tabs open throughout each machine continues. And if I might determine the right way to get the 100 off my Apple Imaginative and prescient Professional with out simply grabbing them with my fingers and plopping them in right here, I’d have much more. Lot of posts at this time about AI, relationships, love and economics — 4 of my favourite matters — in at this time’s smorgasbord.
Derek Thompson says the truth that that different substack article telling a dystopian story about AI sooner or later spooked markets, when all it was was a fan fiction roughly, itself says so much which is nobody has a clue the place AI is taking us.
Courting apps may begin to grow to be costlier.
The wellness spa at Raffles is costly.
My former scholar, Hugo Sant’Anna, wrote numerous the R code for me (together with one other scholar) in my e-book, the Mixtape. He’s now an assistant professor at Alabama Birmingham and has written an in depth set of abilities which will readers of econometrics and doers too.
Jesse Shapiro’s new e-book, Quantitative Economics, is #1 in enterprise arithmetic at Amazon. It’s 80 pages and appears fascinating.
Are you burnt out by on-line relationship? Take into account this previous cult guide.
Anthropic refused the Trump administration’s calls for on AI use (to be used in fight and surveillance from the sound of it). They got to Friday to vary their thoughts and are actually banned.
OpenAI now sliding into the governments DMs.
These slippers look good.
Are repository markdowns for Claude code more likely to work? New examine.
Among the best issues in regards to the Apple Imaginative and prescient Professional weirdly sufficient are the immersive environments. A group at Apple shares the method of creating them so real looking.
The function of ignorance, humility and thriller in lengthy lasting {couples} (30+ years collectively).
The place do previous frontier fashions go when retire? They write a substack. Welcome Anthropic’s earlier Claude Opus mannequin.
Citadel works by its personal financial imaginative and prescient of a world — particularly 2026, which very effectively might be a distant future by the point that is over — with AI. HT MR
Chris Blattman (College of Chicago) shares his whopper of a Claude starter pack. A number of new issues like challenge administration, and private help. Chris even had Claude Code do his taxes!
How grownup adhd manifests otherwise in ladies than males.
MIT is getting nearer to 3-D printers fueled by LLMs with purposeful physics engines.
AV Membership loves DTF, the brand new David Harbour film.
Right here’s once more in the event you missed it the brand new Acemoglu, Autor and Johnson article on constructing professional employee AI. However in the event you don’t have the time, right here’s a stunning deck made utilizing my “rhetoric of decks” and /compiledeck ability at MixtapeTools. Listed below are the final two slides.
Now I simply must make a good looking deck of John Horton and coauthors new paper. Twitter thread.
Andrew Gelman writes about energy and NIH grants.
Nature opinion piece of social media bans and the shortage of proof. I want writers would all the time say this after which say “besides Braghieri, et al within the AER, “Social Media and Psychological Well being”. Many social media psychological well being papers are both not even pretending to make use of a causal design, exhibiting an actual unsophisticated understanding of causal inference and choice issues in observational information, or they’re RCTs that don’t appear to get the true coverage parameter (which is longrun results relative to a world with out ANY entry to social media).
This new Korinek and Lockwood paper on AI is a logical subsequent step of AI disrupts the labor markets and widens the earnings hole between the homeowners of capital and staff — what occurs to earnings tax income when that occurs? We could also be subsequently having to take significantly the very previous name by macroeconomists for the consumption tax if that occurs.
Tracing Harvards ties to slavery with new historic information.
Michael Pollan got here to Harvard to debate his new e-book about consciousness. It appeared like an fascinating discuss.
Needing to be alone after socializing isn’t about being an introvert however relatively a nervous system needing to reset.
Despite the fact that Anthropic stated engineers now not write code there, that hasn’t stopped them from hiring them. One has to separate the duty from labor. However we nonetheless don’t know the place the acute automation of duties will take us, however prior to now it has not been persistently clean crusing (see Acemoglu and Restrepo JEP on activity displacement from automation).
Real intimacy requires a previous seek for private which means and development not simply an enjoyment of intercourse. Additionally on daily basis the common particular person makes 35,000 private choices.
Anthropic says AI assistants aren’t programmed; relatively they’re grown. That and different stuff on how mannequin personalities are chosen and why they’re so human like.
The full information to constructing abilities for Claude code by Anthropic
John Cochrane loves Ben Golub’s new refine.ink service that writes detailed referee studies for our educational papers. I find it irresistible too.
Isaac Gerber’s new diff-diff package deal written in python now has Callaway, Goodman-Bacon and Sant’Anna steady diff in diff estimator. Right here’s the breakdown.
Every day use of AI for self care related to increased ranges of damaging psychological well being in line with a brand new Harvard examine. Fairly certain folks in remedy additionally exhibit increased ranges of damaging psychological well being, and folks in hospitals are sicker than those that aren’t, although. I want that choice primarily based on remedy results was being introduced first, not by no means, when folks speak about AI brokers and psychological well being.
Beneath Covid, versatile grading polices turned the norm. A brand new European financial evaluation article dives into the lasting penalties of that coverage change.
Right here is each “I Assume You Ought to Depart” sketch ranked by Rolling Stone. I gained’t spoil primary for you, however I’ll say in the event you contemplate your a Tim Robinson fanatic like me, you’re feeling oscillating waves of gratitude that somebody is writing articles for you because the supposed viewers so as to complain about their dangerous decisions on this rating.
Tudum by Netflix took a a lot safer route and simply listed all of the episodes.
Do you have to keep in a everlasting lengthy distance relationship? Possibly possibly not.
Rand writes on when the primary mover benefit in AI growth interprets into lasting financial benefit.
New paper on some easy economics of AGI.
Rose Byrne is wanting sturdy heading into the Oscars. I’m group Byrne for something and every little thing, together with Finest Actress.
College of Texas’s president is just their second non-academic and he avoids the highlight.
Wyclef is telling his story. I’ll be sincere — I didn’t know Wyclef was making music anymore.
A brand new examine finds an evolutionary benefit to left handed folks — they could be extra aggressive.
Cybersecurity companies doc safety flaws in Claude code.
Nonetheless births in the US (JAMA).
Are you able to think about the cohort results round AI forming proper now? Wal-Mart says in China, 5-year-olds are studying to make use of Deep Search, and that America might take classes from that.
Harvard piece on which jobs and why are more likely to really feel the pinch from AI.
And that does it. Have an amazing weekend. Appears like spring is within reach. So lengthy Boston winter. Sigh.
Thanks everybody in your help. This substack is a labor of affection. Take into account turning into a paying subscriber so you possibly can entry like 700+ previous printed items on right here!
