November 24, 2025

Monday links from assorted social networks, on science, medicine and game development

  • Tom Forsyth on Mastodon: “Recent discussion about the perils of doors in gamedev reminded me of a bug caused by a door in a game you may have heard of called Half Life 2.” Parallels in biology immediately come to mind.
  • David Roberts on Blue Sky: “In an era filled with tech dipshits who never developed emotionally past the age of 13 & use their wealth to become odious monsters … listen to Steve Wozniak.” We are where we are in big part because there weren’t enough Steve Wonziaks in key industries when it mattered. Or rather, because they by definition bowed out and gave the sociopaths free space to roam.
  • Ruxandra Teslo on X: “We should do smth abt this.” The “this” is the threat of clinical trial infrastructure being flooded by the biotech equivalent of AI slop. And many misguided people think that this is a good thing!
  • Joe Janizek on Substack: The birth of Advanced Radiology. Or: radiology as chess. Radiology and pathology are the few areas of medicine in which AI may be produce immediate benefit.
  • Nassim Taleb on Substack: Medical Mistakes with Probability, 2. Why the benefit of statins in people with barely elevated cholesterol and no other risk factors is grossly overestimated. Note that this constitutes most of the market for statins! My cynical take: Now that they are all out of patent I don’t think anyone would complain about cutting back.

November 23, 2025

It is infrastructure day on the blog today, with two updates:

  1. The Blogroll is now a fresh export from my feed reader and an accurate representation of what I am actually reading. I still need to figure out how to make the very detailed “About” field for each entry actually show up, so stay tuned for that one.
  2. The Now page had its biannual refresh. I will at some point make it a more frequent ritual but best not to expect real-time reading/watching/listening lists.

November 22, 2025

Real-world evidence in support of closed science

When I wrote that opening up science and increasing trust in it are mutually opposed goals, I didn’t imagine the perfect example would come up so soon in both the thing that happened and the commentary about the thing. It is helpful, when interpreting what follows, to keep in mind CS Lewis’s lecture on The Inner Ring with the following two adjustments: there are in fact many rings, concentric, with people ordered in them according to some gradient; and although academia has the secret rings Lewis talks about there are also many public ones with members known, where the innermost ring to a high degree overlaps with Harvard.

The first thing that happened was a segment on 60 Minutes, America’s premier newstainment show, about the current administration’s defunding of Harvard and the implications for science. A few scientists gave interviews, including a bench researcher whose lab studied “different aspects of cancer biology, including tumor heterogeneity, cell-cell interactions, tumor microenvironment, cancer metabolism, drug resistance, and cell signaling.” So, very much a “cancer researcher”, though as far removed from the practical aspects of cancer management as you can imagine. Still, from applying for grants to writing up research results for peer-reviewed journals, scientists have been conditioned to tie whatever they are doing to real-life, practical applications: in the interview Dr Brugge said what she and her post-docs must have written hundreds of times before, that her work has the potential to prevent breast cancer.

There is a legitimate discussion to be had about overblown claims to practicality. The debate has in fact been ongoing for decades now in the editorial pages of various scientific journals. But then someone formerly of Harvard, then Duke, then out of academia completely after a legal dispute, wrote about the issue in light of the segment. This is the second thing that happened.

The article for the most part lists personal observations about the two scientists interviewed for the segment (the second was David Liu, about whom the authors had kinder words). It very much had the sound of someone expelled from the circle grinding an ax with the inner ring. This led to even its salient observations being framed somewhat maliciously. For example:

Universities and their faculty have learned that success in today’s system depends not as much on actually doing science but on marketing the perception of science — framing even routine findings as lifesaving advances. “Cancer” has become a brand, a universal justification for more funding and prestige. The public sees heroism; insiders see dollar signs. One of the strangest features of this ecosystem is how many researchers who do pure basic science — work with no foreseeable medical application — nevertheless frame their research as “curing cancer.”

Which goes from pure speculation to undisputable fact. The need to frame everything as “curing cancer” stems from all the money being allocated to cancer research. It is all about the incentives: Willie Sutton robbed banks because “that’s where the money was” and scientists are no different.

But did I just, even in this gentler framing, compare scientists to bank robbers? See, this is why the debate is best held behind closed doors, lest a politician uses the fact that most research findings are false as an excuse to cut funding. This is what most comments to [Mike

This is the open science dilemma: have the debate out in the open and risk providing ammunition to your enemies? Or do it behind closed doors and risk mistrust? A few decades ago the point was moot as the “enemies” were first powerless hippies, then only slightly more empowered religious zealots. As we all know, the anti-science front has since strengthen. Why that is, well, that is yet another debate. Since one of the reasons is that many scientists openly picked sides, whether out of conviction or out of fear from being ostracized, this is also a debate best held behind closed doors.

Until that happens, we will continue to have dialogues such as this one, The link is to what I think is the final post in the back-and-forth, which I think is the only guaranteed way to show the entire thread, but X truly wasn’t built for sharing these kinds of interactions and is not the best medium for having them. all in support of the beef-industrial complex. Other fields have already wised up: the Internet is dying on the outside but growing on the inside, with important conversations moving to private forums. Which, as I argued, they should.

November 21, 2025

After citing Niko McCarty’s list of 30 biomedical essays yesterday, I had an urge to find each and have a separate post linking to them. Well, good thing I procrastinated because he just came out with an ever longer list (130 and counting) that does have the links. Still no Mansions of straw…, but I’ve just asked asked him to consider adding it so let’s see how the list evolves.

A plug for the Daylight computer

You may have noticed more linked lists on this blog, starting this summer and ever-increasing. This is the direct result of moving my RSS reading from (mostly) NetNewsWire on the phone to (mostly) Feedly on the Daylight tablet. Whatever the cons I thought it had in the beginning, they melted away as the proof is in the output. Interestingly, I hardly ever use the pen, but did pair it with an old (pre-Touch ID) Magic keyboard encased in this handy case/tablet stand and this light-weight pair of devices is all I need on most short trips.

Now, it is not a cheap device! There is currently a 48 hour pre-Black Friday flash sale, and it is still $649 pre-tax. It is also much less versatile than an iPad (no camera and therefore no video calls, and certainly not a good media player although being an Android tablet it does have an official YouTube app, unlike some other better-screened devices. But if you already have a large phone and a laptop, does that middle screen truly need to be a full laptop replacement?

I was also pleasantly surprised by the (heavily customized) Android tablet interface. Things have evolved quite a bit since I briefly owned a Fire tablet, which appropriate to the name I wanted to burn in an effigy. I haven’t owned a Remarkable or a similar e-ink device, but from the refresh rate alone I would guess my reaction would be the same. The plain old LCD technology that Daylight uses Even though, yes, they’ve rebranded it to “e-paper” and say it’s their invention. I don’t know enough about screen technology to comment on whether this is valid, but to me it smells like mostly marketing. was the perfect compromise for my uses, and one I hope more companies would emulate.

November 20, 2025

Thursday Twitter hits, biomedical

🏒 Fun to see a winning team at this venue, for once, and with a massive amount of goals. Will come back.

A hockey arena filled with fans celebrates a win with players on the ice and a scoreboard displaying the victory of the Washington Capitals against the Edmonton Oilers.

November 19, 2025

📺 Common Side Effects (2025) worked for me on at least three levels: as a Mike Judge satire (and I love Mike Judge), as the spiritual successor to Scavengers Reign (ditto), and as a true-in-spirit if not in fact look at the American pharmaceutical industry (and here I could link to half of this blog). With 10 episodes of about 20 minutes each — the runtime of approximately one Irishman — it is well worth seeing at a single go.

November 18, 2025

A brief update to yesterday’s post notes that there are still people who care about the true meaning of epigenetics, and even call themselves theoretical biologists. Note that the Institute for Systems Biology is not some drive-by operation, and indeed is the home of this year’s winner of the Nobel prize in medicine. There may be hope yet. (ᔥJeffrey West, on X)

The juicero of coffee?

I did it: I have found my coffee preciousness threshold. Our local coffee shop changed suppliers and only had the beans depicted below to offer. The very helpful barista even offered to pack me a bag of their own in-house coffee beans, which I declined but should have taken as the warning it was.

A box of Onyx Coffee with a tasting guide is displayed on a table next to a bowl of fruit.

Because the beans were… fine. For a light roast, and particularly for the price. Perhaps even on par with Bird Rock, though I will need to make more than one pot for a real test. But everything about this coffee was over-designed, from the embossed packaging to the transparent plastic bag holding the beans to the “tasting notes” insert tucked into the outside pocket. And just look at that website (and the price)!

I am in fact embarrassed for buying it. Who is this for, and do they also own a Juicero? I choose beer over wine because I am repelled by the (usually faux) sophistication of the wine connoisseurs. I’d better reign in my coffee enthusiasm or else switch to tea.