What we’re reading (12/15)
“Bank Groups Sue The Consumer Financial Protection Bureau Over A Proposed Cap On Overdraft Fees” (Associated Press). “Under the finalized rule from the Consumer Financial Protection Bureau that was announced on Thursday, banks will be able to choose from three options: they may charge a flat overdraft fee of $5, they may charge a fee that covers their costs and losses, or they may charge any fee so long as they disclose the terms of the overdraft loan the way they would for any other loan, typically expressed as an annual percentage rate, or APR.”
“Automakers Thrived In The Pandemic. Many Are Now Struggling.” (New York Times). “A few years ago, automakers were celebrating record profits as the pandemic created shortages of new cars, allowing them to raise prices. Now the hangover is setting in. Nissan, the Japanese automaker, is laying off 9,000 employees. Volkswagen is considering closing factories in Germany for the first time. The chief executive of the U.S. and European automaker Stellantis, which owns Jeep, Peugeot, Fiat and other brands, quit after sales tumbled. Even luxury brands like BMW and Mercedes-Benz are struggling. Each carmaker has its own problems, but there are some common threads.”
“The AI Calculation Debate” (Cass Sunstein working paper). “Could AI predict the outcome of a coin flip? Could AI have predicted in (say) 2006 that Barack Hussein Obama would be elected president of the United States in 2008? Could AI have predicted in (say) 2014 that Donald Trump would be elected president of the United States in both 2016 and 2024? Could AI have predicted in (say) 2005 that Taylor Swift would become a worldwide sensation? The answer to all of these questions is "No." AI could not have predicted those things (and no human being could have predicted those things, either). There are some prediction problems on which AI will not do well; the reason lies not in randomness, but in an absence of adequate data. There are disparate challenges here, but all of them are closely connected to the knowledge problem, and in particular to the unfathomably large number of factors that account for some kinds of outcomes and the critical importance of social interactions. In important respects, the Socialist Calculation Debate and the AI Calculation Debate are the same thing.”
“Are LLMs Running Out Of Data?” (Marginal Revolution). “Ilya and many other experts say yes. I would not dare to disagree with them about AI per se, but I cannot say I am entirely convinced. Supply is elastic! That is a time-honored economic truth. So perhaps the future for traditional scaling is brighter than many of the experts currently are suggesting. We outsiders don’t know exactly which sources of data have been fed into the current top models, but surely there is plenty left? And for some price perhaps it can be mobilized. That is without even getting into data generated by mobile high-quality, life-sampling robots, which admittedly are still some number of years away.”
“The Drugs Young Bankers Use To Get Through The Day—And Night” (Wall Street Journal). “Images of Wall Street’s rank-and-file blowing cash on illegal drugs and nightlife are well known, with cocaine a favored drug through the 1980s, as portrayed in ‘The Wolf of Wall Street.’ These days, drugs are more a tool to optimize performance on the job. Especially for entry-level bankers at the analyst and associate level—who work long, tedious hours and fiercely compete for higher-level jobs with big pay days—prescriptions for stimulants such as Adderall and other ADHD drugs have become commonplace.”
