Being a programmer will make me a better doctor | Hacker News Reader

Being a programmer will make me a better doctor

scopeblog.stanford.edu

45 pointschmaynard41 comments

The thing I've learned dealing with professionals in other fields is that they just don't care as much as you do. The economics of their profession simply don't allow it.

In software if you spend 8 hours fixing a bug that bug will be fixed for all future users of your software. Someone spending 8 hours to diagnose a medical condition will surely be meaningful to you but won't help the rest of the people in the waiting room. Further most people do not have exotic illnesses or needs so rarely is such detailed diagnosis necessary.

If you want someone to fix your car right rather than just replace random parts, or to give you detailed medical attention be prepared to search a long time to find someone willing to do so.

> The economics of their profession simply don't allow it.

Indeed, the FT wrote this morning that in this situation concierge medicine is booming in NY. If interested: https://www.ft.com/content/09b48bce-67fd-11ea-a3c9-1fe6fedcc...

I feel the same way about taxes, medicine, and real-estate: there are professionals out there who are more baseline-skilled than I am, but there's absolutely nobody on earth who cares about MY problems as much as I do.

In the average professional-personal interaction, they are getting... 5% of the margin from their intervention? 10 at most? Whereas 90% of the benefit goes to me. So there's a VERY quick diminishing return to their time investment, in a way that is not true for me. Example: real estate agent. They get 3% of the cut, I get 97%. Should they spend 2x as much time to increase the sale price 20%? No, that's not efficient for them, but it IS worth it for me.

If you're dealing with a professional who is MUCH more knowledgeable than you, it's worth it to take their advice, because they can do a better job. But at the margin, it's often worth investing the time to do deep research.

Nobody will ever care about your problems as much as you do, and it shows.

It must depend on the industry. My experience is the opposite.

I've had programming jobs where every single day consisted of me running into a bug and wanting to fix it, and my manager saying there's no ROI in that and I need to add new features instead like The Roadmap says.

After switching to the physical world, there's definitely an understanding that while we shouldn't be slow, we should take as long as needed to do the job right. If we screw up, people could die. As they say: "OSHA regulations are written in blood."

What about your experience with medical professionals, mechanics, etc.? It's all relative.

This is one of the frustrating things about what I have when I go to the doctor.

As a developer, I generally want a lot of information in order to pose theories about what is going wrong. They always seem really annoyed by this.

Mainly because it is genuinely exhausting for any medical practitioner. That lots of patients "enjoy" googling symptoms and coming up with far-fetched self-diagnoses is a given. But couple that with the perceived intellectual superiority of (software) engineers and you get a recipe for disaster. It's the equivalent of a doctor leaning over your shoulder while you're coding and telling you to remove random keywords.

I don't think it's as bad as removing random keywords, but given how limited resources are in medicine, I can see how a provider wouldn't want to explain why every individual's hypothesis is likely incorrect.

On the flip side, misdiagnoses are surprisingly common, and I think it's worthwhile for any provider to take a closer look if a patient has concerns.

If anything, I think this illustrates how much we need to reorganize and improve medicine. It's not like medicine is alone in this respect either, many sectors are inefficient, but when medicine is life altering and can be life or death, it's pretty high on the list IMO.

Yeah I don't buy this.

Like any field I think there is a spectrum of quality and there are some really great doctors that know a lot, some really bad ones, and a lot of mediocre ones.

I've had a doctor (in the bay area) tell me that I should smoke a cigarette instead of having coffee if I'm having trouble sleeping, but want to keep working on something. Another talk positively about the butter coffee guy. I think the main reason they don't talk about a lot of options is probably time constraint and the common case being right most of the time. This means if you're actually not a common case you're probably better off investing your own time to try and figure things out too.

I like this article though, I think there is some similarity of style in troubleshooting software and disease diagnosis (just very different things to reason about).

I think it’s because there’s a lot of ambiguity and “best guesses” in medical science and there’s no equivalent to programming documentation or manpages for medical treatment. Building code is for all intents and purposes a pretty objective and repeatable field of study. Whereas I would liken a medical treatment more to penetration testing because you’re trying to get an established logical system to accept new input/logic (medicine/procedures) rather than trying to build a logical system from scratch (or with building blocks)

Everybody feels the same way - mechanics feel exactly the same. IT tech support people feel the same.

And, you know, it seems like an increasing number of medical professionals just defer to whatever the patient wants anyway. What's the point of saying you're the expert if you won't be the expert?

One of my pet hates is that I have not been able to crack how to make doctors talk about probabilities quantitatively. It just seems to be impossible.

So far I have three theories why this is:

1. Med school teaches people that as they are just a little above God, they are always correct and do not need to care about probabilites.

2. MD's just do not understand probability and it is not taught in med school

3. Most patients have no clue about probability, can't stand uncertainty and they just want "the truth" from the doctor, who then calibrate their behavior according to that.

I kind of hope it is 3. but occasionally feel like it is 1 or 2.

I read When Breath Becomes Air, [0] the story of a physician-scientist's own battle with cancer, and one of the things he writes about was his frustration at getting treated like a patient when he was trying to get hard numbers for his survival probabilities.

It seems the consensus in the medical community is that there's very little good that can come from giving odds to a patient. What are you going to do with that information?

[0] https://www.goodreads.com/book/show/30241171-when-breath-bec...

> What are you going to do with that information?

Yep. Why on earth should I be given the means to make informed decisions about my body and life? What an outrageous idea.

(sarcasm included)

It's a very memorable experience when a doctor thinks he is helping by telling a person they are recovering from a terminal illness and they and their loved ones take him at his word and then go nuts trying to figure out why symptoms are getting worse.

That's the slippery slope you end up on when you say odds don't matter.

Or none of the above. So, you can apply statistics to medicine in the form of incidence of a disease, and likelihood ratios to increase or decrease the chance of disease in diagnosis. The problem is, the data for these is so difficult to interpret. Likelihood ratios in papers often have huge confidence intervals (like anywhere from 2-15), making their use problematic. Then often, they are quoted only for very specific patient groups, and applying them out of these groups invalidates the rule. Then things like prognosis are so variable, taking quantitively becomes a bit pointless. It’s especially true in cases where treatment or investigations are changing constantly, such as cancer. For diagnosis, the history is often more useful. Humans are story tellers, and the description of the problem varies hugely between patients, as does the significance of certain symptoms or descriptions. It’s an art to work out what is the real symptom and what isn’t. To work out the real picture isn’t always easy either.

Ultimately for things where there is relatively good evidence to practice statistics they are made into algorithms (see MDCalc for examples) but again it’s hard to apply them out of the correct patient population. Even these rules seem to poorly apply in many situations.

Then, real hard statistics rarely change management. You think of the reasonable causes and risk reduce as far as possible. Knowing the actual probability doesn’t really change what you can do. Same kind of goes for prognosis. Most patients don’t really know what to do with exact numbers (especially when they vary so much between different patients).

So, ultimately this is why doctors don’t use statistics regularly.

Sorry, but that is to me a textbook example of what I think as case 2.

It is not about hard statistics. It is your job to translate all those uncertainties and sources of knowledge to (subjective) probabilities. If you do not do that, you are doing things wrong. I do not see how there is anything to even discuss here. Probabilities are the current best tool we have to work with uncertainties.

I do not hire a doctor to just tell me what will be done. I hire a doctor as a consultant to give advice what he thinks is wise given his knowledge and then I do the decision what I do with my body and life. And the consultant should damn well give the best available information he has, just not think that I am so dumb that I can't decide myself.[1]

(Again, sorry for the rant, I have been way too frustrated on this specific issue earlier in my life)

[1] Obviously, most of the time I do agree with doctor's recommendations. But if I am one day having a cancer and a doctor refuses to estimate the odds for me living another month, year and decade, strong verbal abuse is given.

>...I kind of hope it is 3. but occasionally feel like it is 1 or 2.

Unfortunately #2 appears to be an issue for many doctors:

https://blogs.cornell.edu/info2040/2014/11/12/doctors-dont-k...

I have family who are MDs and I have been working with MDs for over 5 years. I have always felt the same, but usually MDs don't agree that much - feeling special is part of the meme I guess.

I do feel, however, that the converse would not work as well: most MDs I met do not practice the "abstract" thinking that is required to be a successful programmer. But that's just my experience.

They sound more akin to kernel programmers.

"Based on the patient's past history of seasonal allergies and the high pollen count that day, most signs seemed to indicate that there was nothing to worry about. But--in order to rule out the unlikely "edge case" that the patient had contracted an acute viral infection that could become serious when his immune system was suppressed by chemotherapy--my advisor made sure to run tests for several common bugs before starting treatment."

And that's the problem: thinking like this as a programmer is cheap and fast; as a doctor it is slow and very expensive. I for one would prefer a very cheap doctor that didn't think like this most of the time. Unfortunately, massive malpractice settlements will always hinder the goal of cheap healthcare.

I don't think most programmers I know would have a very winning bedside manner.

The latest generation of doctors don't have it either

To a patient on life support.. did try turning it off and on? How's the healthcheck?

after death, to nurse: did you check if it's plugged in?

Neither have many doctors I've interacted with. The God-complexes in some of those people is just incredible.

Ditto for some programmers who exhibit mind-boggling levels of Dunning-Kruger Syndrome.

I'm just saying - poor bedside manner is a known problem, work on improving that, not bringing in another discipline that for some reason or other seems to have problems with sociability.

Eh, as a doctor with an engineering background who does a lot of programming projects on the side and feels confident in both roles, they’re very different skill sets and I wouldn’t say it makes me a better doctor. If you go into a non-patient facing area like radiology or pathogy you can leverage it into improved toolchains with AI assistance, but you also don’t have to be a programmer to incorporate technology into your practice.

The huge, and frustrating, difference is that software is (almost always) deterministic and (in principle) can be fully understood. With enough time and money, I am positive that I can fix any bug you throw at me. Human illness on the other hand sadly often leaves one without a cure.

Another example of the “everyone should learn to code” frenzy.

Programming, like spoken language, is one way to express yourself. A lot of value is lost because otherwise apt people, with profound skill or potential in a given vertical like medicine, cannot express themselves in digital terms. Moreover, non-technical people use Excel in amazing ways. It would be better if they knew how to program. The volume and distribution of technical labor is becoming more and more of a bottleneck to innovation. Every year that we don’t systematically introduce people to programming is another year where we forsake large amounts of innovation, and the economic value lost compounds, just like investments compound. There must be a large national effort to introduce young people and adults to programming.

There is a curse there though.

I've found fluency with how the computer actually works creates a great deal of friction with regards to trying to communicate with other people.

I'm so tuned to there always being an underlying cause that even my day to day communication takes on aspects of subconscious Neuro- linguistic programming; I end up trying to constantly tease apart or divine the nature of the underlying semantic/experiential construction of relatively high level communication done with a new person. It's gotten so bad, I can't really comfortably engage in small talk, because to me, it feels like I'm fuzzing the other person and being fuzzed by them in return to the point I end up getting frustrated at the complete failure to set up meaningful communication.

What this ends up looking like, is that with someone I know I can get into high throughput, deep conversations. If I don't know someone though, I end up having to noodle around and pry to desperately scrabble together enough information to put forth a credible attempt at meaningful communication where we both come out saying the same thing about the same thing to a high enough degree of confidence that I"m comfortable we formed a shared understanding.

Also, any person that outright frustrates that calibration phase, or demonstrates excessive volatility in terms of being able to reliably recall things (see compulsive deceivers/gaslighters/liars) quickly finds themselves on a low priority tier for further communication in many instances. I don't like doing that, but every time I relent in the practice, I tend to get hurt.

Let's say, arbitrarily, that you automatically connect with 20% of people, 50% of people take some work, and the other 20% aren't worth it.

The 20% that aren't worth it can be further split into people who are very different from yourself, wallflowers, and antisocial individuals. If you are not eager to please, you will identify the last 20% easily. The trick, though, is that you can't automatically dismiss them. You must assess their importance to work or personal goals, and tread lightly if they are important. With the 50% who take some work, I've found that de facto talking to them like we're already friends helps unlock great conversations. I consistently take the small risk of seeming eccentric in order to get past small talk in my first conversation with strangers. I don't even do that to form lasting friendships. It allows me to make "temporary friends" wherever I go, making for a more pleasant day if nothing else. It also helps me have a stronger sense of identity. I feel more "generic" when I talk about generic things.

Natural language and non verbal communication are difficult. I have found that it can be especially difficult for some technically minded people such as yourself, while some people seem to do it effortlessly. I give this advice freely to my friends: you're a hacker, so hack it.

There are an abundance of resources out there for learning social skills, I'd recommend selecting just a few and memorizing them using spaced repetition. These resources will give you social exercises, like asking how a cashier's day is going before doing business with them. I'd recommend memorizing each of the exercises as well, because you will never know when or where you will be able to do an exercise, despite the book(s) telling you it is critical to do each exercise in order, and only to continue reading after having done the exercise -- disregard the author in these cases.

This won't make you a hot shot sales guy or a presidential candidate right away. Think of it as learning to program for the first time. In the beginning the language, syntax, memory allocation, and data types seem like "the" thing to know in order to code, and later on we laugh, because we were missing the forest for the trees. But the thing is, knowing those things are prerequisites to being able to build towards the larger picture of the system as a whole. When you memorize social skills and nuance, that knowledge will initially lie dormant in your mind, and that is an incredibly powerful place for it to be, do not feel guilty for not putting the tools to work right away. Then on some morning you might be in line about to order a muffin, and you might talk to someone in line, or the cashier. You might see some people joking around on a bus. You will start to see the things you've memorized coming to life, and you will have that "oh shit, I can do anything, can't I?" snap moment that comes for many good programmers and engineers.

Some people might look down on you for memorizing social skills, so don't tell them. What they don't realize is that as an engineer, you invested more time in learning technical skills, and as a consequence of that you invested less time learning social skills. Everybody that has a "type A" charismatic personality has memorized/learned it as well, just in a different context.

https://socialpronow.com/blog/books-improve-social-skills/#1

Also, many books will touch on the concept of confidence, or just have it generally implied. The easiest way to get it is through physical exercise. Cardio for clarity and weights for willpower.

> Programming, like spoken language, is one way to express yourself.

I am not a software developer, nor have any interest becoming one. What programming has given me is a Thinking superpower. Thinking as really slow thinking by Kahnemann. That's what I think is the reason more people should have a look at programming. To understand they can think better and way more efficient when using code as a supertool for thinking. Not necessary to produce a software product to sell, but just to enhance their own work/hobby/whatever.

right. it's nice and maybe flattering that he ascribes these things to learning to program. but "[b]reak problems into sub-problems" and "[a]lways check for edge cases" are hardly unique to programming - hence why he found them so transferable (the "[m]emoization" one just seems like the first to me).

still, it's always nice to see an appreciation for our craft from maybe unlikely sources, and there's no reason not to encourage that as much as possible (just don't force it on people)

Replace "programmer" and "doctor" with almost any other professions and it would be just as true. Strategies like "Break problems into sub-problems" is in no way specific to these fields. You might as well say "Being a carpenter will make me a better CEO", and for exactly the same reasons.

I presume that you mean "medical" doctor. I also presume that this is a hypothetical statement. You have not completed your medical training.

Would be very interesting to read about your experiences once you have become a doctor and then putting the two professional experiences into relative context.

Whatever happened to Barack Obama learning how to code?

... but a worst surgeon. Too much of nintendonitis and you are out of the game.

Reply on news.ycombinator.com