Author: Aceso Under Glass

Scott talks about a bit of medical theater, in which companies demand doctors notes to allow people to bring in their own chairs to work; he is often the person in best position to provide that note, despite being a psychiatrist who usually works remotely. He’s knows little about back pain and nothing about chairs. His compromise has been to write out a note saying “[Patient] reports back pain that would be improved by a chair they have chosen.” This shouldn’t work because he’s not adding any additional information to the patient’s complaint, but it does.

I had my own experience with this in my first real-world programming jobs. The overhead lights in my office gave me a headache, so I turned them off. Someone noticed and didn’t think it was fair my roommate be forced to work in the dark (although he told me it was fine). I indeed had to go to the doctor, who dutifully wrote out “Elizabeth says the lights give her headaches…”. Luckily I had fabulous health care insurance and flexible hours at my job, so this was a minor annoyance.

I really disliked the way the company handled it though. Their solution was a lamp, which was fine, but there was a lot of pressure to say either “yes, the problem is entirely solved” or “no, it is definitely not solved.” “Can I take a week to see if I develop a headache?” was not an option. I don’t think I even got to pick out my own lamp.

Several years later I had a similar problem at a different large tech company you have definitely heard of, one known for being an amazing place to work. I hated working in an open office and tried to leverage my post-dental-surgery fragility into a private office or the right to work from home. In retrospect, this was not a reasonable accommodation: despite having offices in 50 countries and running much of the world’s telework infrastructure, that company’s workflow was set up for in person communication. But they couldn’t just say that, so they tried to find ways to meet the letter of my doctor’s note without actually giving anything up. This was a problem because what I really wanted was control over my environment, and that was the exact thing they didn’t want to give me.

I think a thing that’s going on with the notes is that gatekeeping makes sense in some circumstances (substantial changes in job duties), and it was no skin off the decision-maker’s nose to make that a universal rule. “Protection from lawsuits” factors in, but if companies were paying the cost of the note they would draw the line in a different place. As it stands, it’s worth infinite amounts of your time and money to avoid risk of a lawsuit that might cost them either.

Recently I listened to Econtalk’s episode on the doctor-patient relationship, which was primarily about a study showing that hospitalized patients did better when their care was supervised by their primary care physician. The study shows what it shows (although pilot studies are prone to not replicating), but implicit in the discussion was the idea that continuity of care was important because doctors remembering things is better than writing them down. This may be true, and the podcast in fact cover some of the problems with medical records*, but if so that is a profound failure of medical record keeping. Human memory is the last thing you want to trust your life to. I pay extra for the kind of doctors that spend actual time talking to me, and I still have to remind the one that that drug she loves has unacceptable side effects for me, every time I see her. I in no way trust her to remember something as dense and detailed as test results. That’s why she writes them down.

My understanding is that many people do understand this, but fixing medical record keeping involves a huge government entanglement that stifles all progress. So we’re stuck with human memory as the thing keeping us alive.

Note that there are good reasons to have your PCP supervising your hospital stay even if everything is properly written down, which is hospitals are full of specialists who aren’t even trying to coordinate care. You benefit immensely from someone thinking “how are these things going to interact?” In theory this should be done by a hospitalist, but they only have so much time and not doing it is much faster. But let’s not pretend we’ve proved human memory is in any way a good solution to any problem ever.

*EHRs simultaneously have too much information, taking a long time to complete and read, and don’t have space for information off the beaten path.

EDIT 1/1/2019: I have left LRI and no longer recommend donating to them.

A theme on this blog is “there isn’t enough good information.” Anti-aging research is worse than most topics at this, perhaps even worse than nutrition. There are several reasons for this: time/money requirements of research, the FDA’s focus on pathology rather than health, and science as a whole disincentives the kind of replication you need to truly trust a result. But there’s a new organization, the Longevity Research Institute, aiming to fill this gap. (Full disclosure: I’m on the board).

To run a longevity experiment, you have three choices: research something with a short life cycle (limited applicability to humans), research something with a long life cycle but track biomarkers instead of actual lifespan (risks goodharting the biomarker while not affecting or even hurting lifespan), or wait. Even a mouse experiment can take four years for your subjects to die. And the more successful your treatment, the longer (and thus more expensive) the experiment.

This would make lifespan-increasing treatments more expensive, but that’s hardly the end of the world. What could be more worth paying for than life? Unfortunately, the FDA takes another view. To get FDA approval, a drug must show effectiveness treating a specific disease, and aging is not considered a disease. This has been challenged recently, with the TAME trial, but has been a fact of life for a very long time and there’s no guarantee TAME will win the fight.

Finally, science in general is bad at replication. Repeated studies are necessary to have any confidence in a treatment, but glory goes to the original discoverers of a treatment and their funders, not to replicator three of ten. So we end up with a list of promising treatments that no one can trust because no one has any incentive to test them again.

All together, this leads to a lot of low-hanging fruit in longevity research; there’s a ton of promising molecules out there that only need money to be tested. The Longevity Research Institute aims to pick that fruit. It has a very targeted mission: take compounds already shown promising in at least one trial and fund replication trials where you actually wait to see how long the mouse lives. That’s it. Yes, it should worry you that this isn’t already being done.

The LRI is run by Sarah Constantin, who writes the blog this blog wants to be when it grows up. You can see her past work in e.g. treatment-resistant depression, STD transmission, and chronic fatigue syndrome. Because there are very few fixed costs (Sarah’s salary, some legal consulting) and studies are so expensive, there’s lots of room for more funding. You can see the LRI’s roadmap here and an initial list of promising compounds here. LRI’s first trial, of the synthetic pineal peptide epitalon, launched just last month.

For more information, you can follow the blog or Facebook group. If you feel so moved, you can donate to LRI here. I’m happy to talk to anyone considering donating (elizabeth-at-this-domain-name), and for larger amounts Sarah herself is available.

In Robert Kegan’s In Over Our Heads, he says “it is a poor school whose favorite students are the ones it does not have to teach.” (page 171). I get what he’s going for, but having lived it, I can also say it’s a poor school whose least favorite students are the ones it does not have to teach. It tends to lead to things like resentfully insisting you do have something to teach the student and forcing them to go through the motions of learning it, or standing in their way when they try to do something they’ll actually learn from. Or an intrusive focus on their personal life.

I feel like the whole concept of favorites is leading people astray here, which is a very Kegan thing to say.

When you hear “supply chain”, your first thought is probably not “let’s base a leisure activity around that.” But the world is so big and game development is so cheap now that that this does not stop there being more video games than I can play that cater to my taste in complicated supply chains. If you happen to share those tastes, here are some games to try:

Colonization: Built off the Civilization 4 engine. You build cities near resources on the map, which you can harvest using units of various skill levels, which can then be processed into various goods, again by units of various skill levels. Theoretically you need to build liberty so you can declare war against your colonial power, but I always quit before that stage, because why would I spent time on war when there are supply chains to build. You have to move goods between cities (for processing and selling to the mainland) manually, which makes this my low-thought comfort game.

Anno (1404): This is a whole series based on both trading and supply chains. I would say they made it just for me but there is a combat section :(. I’ve only played Anno 1404, supposedly the best of the series. Anno 2070 is set in post-apocolyptic water world, which seems like it’s even more for me, but it has annoying Ubisoft DRM so I’ve never actually played it. Anyway, the mechanics. James Recommends does a better job explaining than I will, but it is a video, so I will try anyway.

You start on an island that has all the resources you need to keep your peons alive, which you provide by building resource extractors (apple farms) and processing plants (cider mills). As you make them happy they evolve into better people, enabling you to tax them more and build new buildings. Which is good, because these better people have more demands for making them happy. Complicated demands.

For example, Patricians want books, which requires a printing house. Books need indigo (which requires a sufficiently large Oriental settlement on an island fertile for indigo) and paper, which requires a paper mill, which requires lumber input, which comes from a lumberjack hut. Oh and paper mills can only be built on rivers, so you might have to find an island just for that. That’s not too bad. But you also have to build the printing house, which requires glass. Glass comes from a glass smelter, which needs quartz (which requires a higher level Oriental settlement and a quartz quarry) and potash, which comes from a glassworks building that runs on trees.

Put more graphically

What is most unique about Anno is that your resources are limited to the island they were farmed/created on, and if you want them on a different island you have to move them with ships.

Unlike every other game I can think of, Anno lacks a staffing mechanic. All buildings come fully staffed, even if there’s no housing on their island (and there are reasons you generally concentrate your housing on two islands). Housing is essentially a tax farm, with which you get money to build more buildings. I think the purpose of combat here is to create something for you to spend money on that is not more buildings, which is fine, but it involves more strategery and uncertainty than I feel like dealing with in my leisure time. Luckily, Anno lets you turn off pirates and rivals completely so I can build my little anthill uninterrupted.

Tropico: I think I’ve played Tropico 1, 3, and 4, which is enough to say they’re all the same and it doesn’t really matter which one you get.* In Tropico you are the Presidente of a banana republic, and have goals like “build a kickass diversified economy” and… I didn’t commit the others to memory. I think one involves a Swiss bank account?

Tropico’s production supply chains are much simpler than Anno’s, barely more complicated than Civilization’s.  Raw material -> processed good -> ship off the island for money. But it does have a nifty education mechanic. Different buildings require employees of different genders and education levels. You can’t simply overeducate your workforce, because being underemployed makes people dissatisfied and inclines them towards regime change. You could rig the election or kill them, but that has other consequences. You can give them makework, but that gets expensive, especially if you give them makework as teachers and professors and create more education.

Tropico has no combat mechanic, unless you count peasant revolt, which I don’t because it can be averted with good enough supply chains.

*I picked up Tropico 5 in the latest Humble Jumbo Bundle, which is around until 4/3, but I’m trying to finish Soma in time to review it while the latest Humble Indie Bundle is up, so I haven’t actually tried it. I assume it’s like the last 4 but with more DLC.