- Kenneth P. Vogel and Christina Jewett for The NYT: A $5 Million Donation From Big Tobacco Preceded F.D.A. Vape Decision. [Note: Gift link ]
There is no definitive evidence linking the new F.D.A. guidance to the lunch, the donation or specific lobbying. But the episode represented a victory for an industry that mostly had been on the defensive for years.
The now former FDA Commissioner Mary Makary quit in protest, and this isn’t the first time lobbying has led to FDA turnover. Yes, lobbying is great again! Say what you will about Makary or his recent subordinate (and, full disclosure, my co-fellow, co-author and friend) Vinay Prasad — as I have — at least they had principles. Those who pay for a STAT+ subscription can get the opposite take from Matthew Herper, who called Makary the worst FDA commissioner in 25 years. [Note: And before you ask what poor Jane E. Henney (FDA commiss 1999–2001) did to him, 25 years is how long Herper has been covering biotech. The headline could have used an “at least” for accuracy. ]
- Nick Bowlin and Katie Campbell for ProPublica and The Frontier: Oily Sludge Is Flooding Their Dream Home. Oklahoma Regulators Say They Can’t Help.
The basement of a brand new house being filled with oil-smelling, oil-appearing sludge and the government agencies are calling it “water”; a personal and bureaucratic nightmare. This is the wider context:
The Frontier and ProPublica’s reporting on oil and gas pollution in Oklahoma over the last year has shown how old oil wells abandoned by the industry pose severe public and environmental health risks. Officially, the state lists 19,000 orphan wells that state regulators are responsible for cleaning up, but the true figure is likely over 300,000, according to federal researchers.
Drill, baby, drill! Preferably through an LLC, so that you can forget about the holes you left behind once the boom busts. This is why I am surprised by otherwise sensible people like Casey Hendmer being so frustrated with lack of drilling [Note: X-post ] in oil-rich California. Could it be that even the smart Californians who would prefer not to live above an abandoned well? I mean, even the ones with just water in them can be scary. Or is it the case of Eden for the rich and stinky sludge for the poors? Let them have oil!
- Chance Townsend for Mashable: Google overhauls its AI subscription tiers, makes them cheaper.
This is one of many reports from Google I/O which focuses on the new prices without mentioning the severely restricted token limits for all tiers. Here is an example of what lower limits mean in practice. I would like to commend Google’s marketing team for this PR sleight of hand: does it count as shrinkflation is the prices have also gone down? And how on Earth would those lower prices help the already abysmally low revenue? Maybe the relative cost of tokens will have increased, but who knows? It’s not like Alphabet is a publicly traded company that should report that kind of information. Good thing AI is its own thing and isn’t affecting anything else around it. Asbestos indeed.
- Dalmeet Singh Chawla for Nature: Researchers who use hallucinated references to face arXiv ban.
There is now a one-year ban from posting on arXiv for all (co-)authors whose preprint have references that LLMs conjured out of thin air, or other signs of passing on LLM-generated content without human review (such as paragraphs starting with “Here is a 200-word summary of…”). [Note: A question to readers more style-minded than me: is stringing these four — an ellipsis, quotation mark, right parenthesis, period — one after another a typographical faux pas? ] Note that this is for physical sciences only, life sciences-minded bioRxiv and medRxiv have not (yet) instituted such rules. Which didn’t stop some life scientists from defending [Note: X-post ] the practice of not checking one’s own references: who has the time? Apparently not people with current or former NIH funding. Having once spent a full day finding the correct reference to back up a non-essential introductory claim in one of my least-cited papers (7 as of today) I empathise with the suggestion that references should be more of a guide than firm fact, but empathy is one thing and truth another and in matters of science I will stand behind the truth because if not then what on Earth are we even doing? Unsurprisingly, Andrew Gelman has a good take on the matter.