Finally, the good stuff: ACTION ITEMS.

Outlive offers a wealth of information on improving longevity, but to keep things manageable, we'll focus on three key areas: testing, exercise, and nutrition; along with liberal use of footnotes so we can make it through everything without getting too bogged down.

So lock in and let’s get down to business.

Testing

Before we dive into specific tests, let's address an important caveat: reference ranges. While most lab results include “normal” ranges for each metric, these ranges can be misleading. Here's why:

1. Reference ranges are based on general population distributions.

2. As society has become less healthy overall, these ranges have shifted.1

3. What's considered “normal” may not be optimal for longevity.

Therefore, it's crucial to look beyond whether your results fall within the provided ranges. This also seems like a good place to once again say: I am not a doctor, none of this is medical advice, you should do your own research, etc. Always consult with a healthcare professional to interpret your results in the context of your overall health and longevity goals.

The top five tests

The first thing you run into when you start looking at testing is that it’s an acronym soup of a million tests you can take, most which aren’t immediately obvious what they’re even testing. Outlive doesn’t prioritize the top tests (another example of the overwhelming amount of information in the book), but Attia has talked about them elsewhere, and his top five are:

1. Lipoprotein (a) or Lp(a) - Lp(a) is an indicator of your personal predisposition toward heart disease. It doesn’t change, so you only need to check it once in your lifetime. If it’s elevated, then you need to pay extra attention to your ApoB levels and all the heart disease stuff we talked about earlier. If it’s <14 mg/dL you don’t get a free pass, you just have a “normal” amount of inherent risk.

2. ApoB - We talked about this in part two—this is the preferred way to gauge cardiovascular risk instead of just looking at “bad cholesterol”. ApoB will measure the amount of cardiovascular disease causing particles in your blood. Here’s an example of where reference ranges can be misleading. The test I took said my result of an ApoB of 95 mg/dL was in the normal range, so there was no need to worry, and it does appear to be at about the 50th percentile. But Attia advocates for an optimal range of <65 mg/dL, implying I still have a lot of room to improve.2

3. Oral glucose tolerance test (OGTT) - OGTT measures your insulin response and it appears to involve drinking an absolutely disgusting pure-glucose drink, and then waiting and drawing blood every 30 minutes to check your insulin response over time. If your body can get your blood glucose down to <140 mg/dL after two hours, then you’re normal. If it’s elevated still, that is bad. I can’t speak to this test because I haven’t done it, and it’s not clear if Attia recommends doing this even if you are otherwise young and healthy (he mentions it only once in all of Outlive). You can also get at some of the same data via a continuous glucose monitor, and the HbA1c test, even if OGTT remains the gold standard for assessing pre-diabetes.

4. Alanine transaminase (ALT) - ALT assesses liver health by checking if the ALT enzyme is leaking into the blood. ALT increases when your liver stores fat, which is one sign of metabolic dysfunction. This is commonly included in blood panels. Labcorp says <33 IU/L for women and <45 IU/L is normal, but "the American College of Gastroenterology recently revised it’s guidelines to recommend clinical evaluation for liver disease in men with ALT >33 and women >25. Attia unsurprisingly is even more aggressive, targeting <30 for men, and <19 for women.

5. APOE genotype - The APOE genotype determines an increased risk in developing Alzhemier’s disease risk. It’s one of the few tools we have to detect neurodegenerative diseases. This test is also a one-time thing. I can understand why Attia prioritizes this, but it seems like many of the things you want to do to decrease AD risk (like sleeping well, strength training, eating healthily) are things you want to be doing anyway to fight the other four horsemen, so I’m not sure how critical it is. At the very least, it seems similar to the Lp(a) test in that if you are at higher risk, you should take the recommendations a bit more seriously.

How to get tested

You have a few options to obtain the above tests:

1. Ask your doctor and get referred to a lab

   1. This is easiest if you have easy access to your doctor, but they may be unwilling to order every test you want.

2. Roll your own directly with Quest Diagnostics (e.g., you can order this one, as well as a la carte) or Labcorp or go through something like Own Your Labs

   1. This lets you bypass your doctor and just get the tests you want, but it can be daunting as no one makes it very easy to figure out what the hell you’re doing, and not every test seems to be available to the public, even if doctors can order them.

3. Use a third party (non-insurance based) service like Inside Tracker or Function Health

   1. This is the most expensive option, but they handle a lot of logistics for you and provide nice UX for reviewing the results. These services all tend to use Quest or Labcorp for their blood draws and labwork, so you’re paying a premium on top. The downside for some is they try to upsell you into subscription plans or other related health tracking services. I haven’t tried either of them, but if you can afford it I’d probably just go with this for the convenience if my doctor was unwilling to order for me.

Other useful tests

• Continuous glucose monitor (CGM) - CGMs are devices historically used by diabetics to monitor their blood glucose, but even if you aren't diabetic you can get one to wear to easily monitor your blood glucose from your phone.

  • Getting one requires a prescription for some crazy reason, but once you have one, you can get them for pretty cheap—the Freestyle Libre sensors cost up to $75 for two. They're painless and easy to put on, last for two weeks, and sync to your phone. The manufacturer, Abbott, seems to offer a free trial here. (Update: Apparently the FDA approved the first CGM for over the counter purchase without prescription earlier this year. It’s only for a specific model from Dexcom, but this hopefully means that they’ll all be available OTC soon. Thanks, Prabhu for the tip!)

    

  • You want to a) experiment to see what foods spike your glucose (e.g., for me a banana was surprisingly bad) because everyone is different, and b) see how the timing and order of what you eat affects your glucose response. The goal is to reduce the spikiness of our insulin response. For example, by first eating some protein and complex carbs (like beans), and then waiting ten minutes before eating dessert, I found my blood glucose is much less likely to spike as high. But waiting much longer to have dessert reduced the positive effect of the beans.

  • Stress and lack of sleep also impair your body's ability to dispose of glucose. If you're under a lot of psychological stress, the CGM may be a wakeup call to quit your job.

  • Attia targets 100 mg/dL avg glucose or lower, with std dev of less than 15 mg/dL. This is, to no one’s surprise at this point, extremely aggressive!

• Cancer screening

  • As discussed in part two, the best tool we have to fight cancer is to screen early and often. This is why Attia recommends average risk individuals getting a colonoscopy starting at 40, instead of at 45, and repeating every 2-3 years depending on the findings.3

    Two or three years might seem like a very short window of time to repeat such an involved procedure, but colon cancer has been documented to appear within the span of as little as six months to two years after a normal colonoscopy.

  • Strangely I don’t think he mentions any other specific screening outside colonoscopies in Outlive. Regardless, I guess one takeaway is do not skip any routine screening!

• DEXA scan - A DEXA scan measures your bone density, but also can accurately measure the amount of fat you have, and importantly where it is located. Contrary to most body fat % measurements that attempt to use electrical signals, the DEXA scan uses x-rays to tell how much lean mass, fat, and bone you have in each part of your body. It can tell you how much visceral fat you have, which is the bigger worry, as opposed only total fat. I did mine at Dexafit, but just search for DEXA scans in your area to do one. You can do this once a year to track your visceral fat and bone density longitudinally.

  

Subscribe now

Exercise

Now that we've covered testing, let's move on to the cornerstone of longevity: exercise. If there's one key takeaway from Outlive, it's this: EXERCISE IS THE MOST EFFECTIVE TOOL WE HAVE FOR LONGEVITY. While the tests we discussed are important, they serve little purpose if not paired with a solid exercise routine.

Attia recommends a balanced approach to exercise:

1. 50% cardio, further divided into:

   1. 80% Zone 2 training (moderate intensity)

   2. 20% Zone 5 training (high intensity)

2. 50% strength training

This approach addresses a common question: “Is strength training or cardio more important?” Attia’s answer should surprise no one—all of the above.

Let's break down each component:

Cardio

Cardio is split up by heart rate into different “zones”. Lower zones are lower heart rates. Lower zones burn fat (by calling on fat stores for energy), while higher zones use glucose for fuel.

• Zone 2

  • Working in zone 2 burns fat, and improves the efficiency of our body’s ability to use fat as fuel. It's the most important exercise we can do to fight metabolic syndrome.4 The goal is 3 hours a week, or four 45-minute sessions for most people to see a benefit, but you probably can’t overdo it.

  • There are lots of different ways to measure whether your heart rate is in zone 2, but one easy heuristic is that you can talk in full sentences, but just barely. Another way is to target ~60-70% of your peak heart rate. As you practice, it should get easier, and you should expect to need to bump up your target over time.5

    • Zone 2 is not so taxing that you can't multitask, so you can use the time to listen to podcasts, watch videos, etc. I bought a cheap walking treadmill for my standing desk and use it while playing certain games or watching YouTube, etc. The key is to find a way to fit it into your lifestyle that works for you.

      

• Zone 5

  • In contrast to the relatively chill endurance training of zone 2, zone 5 is intense. Working in zone 5 improves VO2 max, or the maximum rate a person can utilize oxygen. VO2 max is expressed as the volume of oxygen you can use, per kg of bodyweight, per minute. An average 45yr old male will have a VO2 max of around 40 ml/kg/min. An unfit person in 30-40s might only score in the high 20s.

  • VO2 max can always be improved by training, no matter how old you are. Zone 5 training is much harder than zone 2, but the good news is you do not need to do it as often. 1-2 per week is enough. Attia claims it may be the single most powerful marker for longevity.6

  • Do cardio (bike, run, row, etc) for 4 minutes at the maximum pace you sustain for those four minutes (very hard, but not an all out sprint), then ride/jog/etc at an easy pace for 4 minutes of recovery so your heart rate drops back down to <100 bpm. Do this 4-6 times, and cool down.7 You don’t have to exactly follow this prescription—even just training zone 2 will also help improve your VO2 max. If you can fit in some occasional high-intensity cardio, you’re on your way.

Strength

Muscle mass and strength matter (unsurprisingly), especially as you age. Having more muscle mass protects you from plenty of things, but most importantly is correlated with a lower risk of falling, which is a leading cause of death and disability in the elderly.

• It is never too late to start—Attia's mom started at 67! There are lots of studies showing strength training can improve mobility and physical function of all types of people, including those who are already elderly and frail. If you are older and unsure where to start, check out this specific podcast episode on exercise for aging people, or see this footnote for some quick notes to send to your parents.8 Attia aims to lift heavy weights in some form 3-5x/week. (He does a LOT of things 3-5x a week).

• There are a few key movements that map to real-world activities that are useful to train, versus just having enormous biceps (though don’t let me stop you if that’s your goal, Daniel):

  • Pulling—can you carry groceries, lift heavy objects when you are old? All the typical pulling strength movements are good: lat pulldowns, rows, etc.

  • Hip hinge—can you get out of a chair without difficulty? You need to be careful of spinal injury with these, so you want to build up slowly, starting with things like single leg step-ups, and split stance Romanian deadlifts, all starting without weights.

  • Grip strength—can you hold onto railings to avoid falls, or open jars? Hands are our primary interface to the world, and grip strength strongly predicts how likely you are to live.9 If you spend time lifting barbells and dumbbells you will naturally build this. In addition you can perform the farmer's carry.

    

    • Attia’s goal is half our body weight in each hand (so full body weight total!) while walking for 1min for males, and 75% of body weight for females. This is very hard!

    • The dead hang is also a good test as a way to measure improvement.

• For all of these exercises, don't forget the eccentric part of the movement (the release of the muscle after contracting. We tend to focus on the concentric portion such as when we are bringing the dumbbell up in a bicep curl, but ability to slowly release your muscle while holding a weight is valuable especially for the elderly because that's how we slow ourselves from a fall. If we grab onto to a railing to catch ourselves, it's our eccentric strength that reduces the impact.

• Stability

  • Pushing yourself without adequate stability almost inevitably leads to injury. Your body is very good at compensating via "cheats". Attia likes to start with stability before adding any sort of strength training, including some of the assessments like dead hangs and single leg step-ups.

  • Attia spends 1 hour 1-2x/week doing dedicated stability training, with 10-15 min per day on other days, but mostly it's folded into every day training. If we're 80/20ing this, we can probably get away with just making sure we do it as part of our other training. Split-leg deadlifts, lunges, doing curls while standing on each leg, etc. Add in some core exercises, and you're good to go.

  • There's a whole section on how to breathe properly here that I will not go over, among other things like toe yoga.

There’s a lot of detail here, but just remember the goal of 50/50 strength/cardio, and 80/20 zone2/zone5. The rest is all further optimization that you can tailor to things you enjoy and your specific needs. Just start doing something!

Subscribe now

Nutrition

When it comes to nutrition, it's important to recognize that there's no one-size-fits-all approach.10 Unlike exercise, which has more universal principles, nutrition needs vary significantly from person to person. The bloodwork and tests we discussed earlier can help determine your specific nutritional needs.

To simplify this complex topic, consider these key questions:

1. Are you under or overnourished?

2. Are you under or adequately muscled?

3. Are you metabolically healthy?

Most people in modern society fall into the category of being overnourished (consuming too many calories relative to energy expenditure) and undermuscled (insufficient exercise and/or protein intake). This combination often correlates with poor metabolic health.

With this context in mind, let's explore some general nutritional guidelines broken down by macronutrient type across the three types: carbs, protein, and fat. (Technically there are four—alcohol is it’s own category but unlike the first three, it’s nutritionally void and you don’t need it to survive.)

Carbs

• The more refined the carbs (like white bread, regular pasta), the faster and higher the glucose spike. The more fiber or the less processed (like beans), the lesser the impact.

• Attia targets more than 50g of fiber/day. Rice and oatmeal are surprisingly spiky, despite not being refined. Brown rice is only barely better than white rice (for glycemic-ness), though you should test and see for yourself on a CGM.

• Non-starchy vegetables (spinach, broccoli) have no effect. Eat as much as you want.

• High protein and fat have no effect on glucose.

• Everyone knows soda is bad for you, but fruit juice is just as bad for glucose spikes. If you're drinking fruit juice in lieu of Dr Pepper, don't bother. Diet soda does not spike glucose.

Protein

• Protein is the only macronutrient that is essential to our goals—there's no practical requirement for carbs or fats (assuming you are eating enough to survive, which is not hard), but you absolutely can be lacking enough protein.

• The US RDA recommended amount is 0.8 g/kg which is baseline but not good. Studies show elderly people consuming that much and losing muscle mass!

• Attia targets 1.6 g/kg/day as the minimum. Data suggests with normal kidney function, 2.2 g/kg/day is a good place to start.

• Most people do not need to worry about overconsuming. The safe upper limit is 3.7 g/kg/day which would require a LOT of work to do. It is practically impossible to do this by accident.11

Fat

• Fat gets a bad rap, but you need fat. It's a very efficient fuel, and also a building block of many hormones and cell membranes. Dietary fat also tends to leave you feeling fuller than many types of carbs, especially when combined with protein, which helps combat overeating.

• There are three types of fat: saturated fatty acids (SFA), monunsaturated fatty acids (MUFA), and polyunsaturated fatty acids (PUFA). Within PUFA there are omega-3 variants, which can be subdivided into marine sources (EPA, DHA) like salmon, and nonmarine (ALA) sources like nuts/flaxseed. Then there are omega-6 variants.

• Remember that most dietary cholesterol gets passed through your body, so it’s less about the amount of fat versus shifting the fat profile:

  • Attia advises to roughly try to get fat from 50% MUFA sources, reduce SFA to 15-20% and make the rest up with PUFA.

    • In practice this means: olive oil, avocados, nuts (MUFA) > butter, lard (SFA). If you take away one thing on fats, it should be to lower saturated fat consumption.

    • Reduce omega-6 rich oils like soybean, corn, sunflower oils.

    • Increase omega-3 marine oils via salmon and anchovies.

  • You need to monitor a lipid panel (apoB, cholesterol, etc) to track how changes in your fat composition affect your overall metrics. It does not work for everyone, so you must re-check.

None of this should be that surprising, and it’s nice to know that there is not one diet to rule them all. The guidelines are pretty simple: don’t eat too many or too few calories. Consume sufficient protein and essential fats (and cut down on saturated fats). Make sure you have the vitamins and minerals needed.

Coda

If you’ve made it all the way through this interminably long summary, I hope you’ve found it useful. I now understand why the summary for the book was fifteen pages long.

In compiling all of this, I have gone back and forth between “this guy is a madman who can not possibly have time for anything else” and “well at least he’s giving you all the context out there so you can decide for yourself”.

Attia can be polarizing, as this New Yorker piece attests:

Too often, conversations about life extension devolve into unhelpful abstractions and untestable speculation; one appeal of Attia’s advice is that it’s so tangible. Critics can paint his counsel as blindingly obvious or needlessly complex—but he has, at least, inspired large audiences to imagine what a better approach to aging could look like. “There is actually no such thing as atheism,” David Foster Wallace once said. “The only choice we get is what to worship.” In a society that chases money, power, fame, and beauty, there are worse gods than longevity.

Reading the tome that is Outlive, you get the sense that you are destined for failure unless you have insane levels of execution and willpower to do everything Attia suggests.

But as I’ve hopefully made clear, the value of Outlive is as a framework. You don’t have to do it all—perfect is the enemy of good. We aren’t aiming for perfect, we’re aiming for improvement so we can live longer, healthier lives.

I enjoyed Outlive because Attia’s obsession with longevity wasn’t just about extending lifespan, but about making your years enjoyable. In the end how you spend your time on this planet is the ultimate long game. Why not make it a game worth playing?

Thanks to Andrew Stevenson for reading drafts of this post.

All I want to know is where I'm going to die so I'll never go there. —Charlie Munger1

Imagine if you could peek into the future and see exactly what might threaten your health decades from now.

Actually you don’t have to imagine; odds are you know a lot of them already; heart disease, cancer, stroke.2

In Outlive, Attia groups the leading causes into groups he calls the Four Horsemen:

1. Metabolic syndrome or dysfunction

2. Heart disease

3. Cancer

4. Neurodegenerative disease, like Alzheimer’s

It turns out the Four Horsemen are not equally easy to fight. We have much more control over the first two: metabolic syndrome and heart disease. If we can nail those two, it'll go a long way towards fighting all four. That starts with understanding each one in a bit more detail, so let’s dive in.

Metabolic Syndrome: The Silent Threat

Ever heard of metabolic syndrome? If not, you’re not alone. Despite it being number one on Attia’s Four Horsemen list, it’s often overshadowed by its more famous cousins like heart disease and cancer.3

What Is Metabolic Syndrome (aka MetSyn)?

Also known as insulin resistance syndrome, MetSyn is a cluster of conditions that together raise your risk for all other diseases. You are defined as having metabolic syndrome if you have at least three of the following five criteria:

• high blood pressure (>130/85)

• high triglycerides (>150 mg/dL)

• low HDL cholesterol (<40 mg/dL in men or <50 mg/dL in women)

• central adiposity (waist circumference >40 inches in men or >35 in women)

• elevated fasting glucose (>110 mg/dL)

MetSyn is highly correlated with obesity, but they are not the same thing. (Note that it's not one of the five criteria, though central adiposity is clearly highly correlated!) Not everyone who is obese is metabolically unhealthy, and vice versa. Just because you appear skinny does not make you immune to this problem!4

Insulin Resistance

Understanding MetSyn requires understanding how energy storage in the body works. When we eat or drink, glucose is the final form of most carbs after being processed by the body. (This is true whether you are eating sugar or pasta.) When the body senses glucose, insulin is secreted. Insulin moves glucose to where it’s needed (like your muscles lifting weights, or your brain solving chess puzzles), and maintains your blood glucose levels at a roughly steady amount.

When you eat a giant slice of cake, your body responds by pumping out insulin to manage the sudden glucose spike.5 Now, imagine doing this repeatedly, day after day, year after year. Your cells start to tune out insulin's signals, like a teenager ignoring their parent's hundredth request to clean their room.

This is insulin resistance, and it's at the heart of metabolic syndrome.

Insulin resistance is a term that we hear a lot, but what does it really mean? Technically, it means that cells, initially muscle cells, have stopped listening to insulin’s signals, but another way to visualize it is to imagine the cell as a balloon being blown up with air. Eventually, the balloon expands to the point where it gets more difficult to force more air inside. You have to blow harder and harder. This is where insulin comes in, to help facilitate the process of blowing air into the balloon. The pancreas begins to secrete even more insulin, to try to remove excess glucose from the bloodstream and cram it into cells. For the time being it works, and blood glucose levels remain normal, but eventually you reach a limit where the “balloon” (cells) cannot accept any more “air” (glucose).

This is when the trouble shows up on a standard blood test, as fasting blood glucose begins to rise. This means you have high insulin levels and high blood glucose, and your cells are shutting the gates to glucose entry. If things continue in this way, then the pancreas becomes fatigued and less able to mount an insulin response. This is made worse by, you guessed it, the fat now residing in the pancreas itself. You can see the vicious spiral forming here: fat spillover helps initiate insulin resistance, which results in the accumulation of still more fat, eventually impairing our ability to store calories as anything other than fat.

Why Should You Care?

You might be thinking, “I’m not diabetic, so why should I worry?” Here's why: insulin resistance is the silent force behind the other three horsemen. It increases your risk of:

• Cancer (up to 12 times!)

• Alzheimer's disease (5 times)

• Death from cardiovascular disease (almost 6 times)

So, while you're busy worrying about heart attacks or cancer, metabolic syndrome might be setting the stage for all of them.

Diabetes ranks behind things like Alzheimer's as a cause of death in the US, but Peter believes it's because we underaccount for the true impact. People with diabetes are far more vulnerable to other diseases that end up killing you.6

This is why even if you are not at risk of Type 2 diabetes any time soon, metabolic syndrome remains key to longevity. Of all four horsemen, MetSyn receives by far the most coverage in Outlive. I was surprised, since I would not have guessed that it was the number one thing to combat for longevity—heart disease seems to get much more press. The good news is between nutrition and exercise, we have a tremendous set of tools to ensure we stay metabolically healthy, which we will cover in part three.

Subscribe now

Heart Disease: The One Everyone Knows

Heat disease is probably what most people think of first when they think of the most common causes of death. Unlike metabolic syndrome, it needs no introduction. But there are still a few surprising things to learn here.

Cholesterol: The Good, The Bad, and The Misunderstood

When we think of heart disease, we of course think of cholesterol. There are a few misconceptions around heart disease that Outlive discusses.

First, most people have heard of “good” cholesterol (HDLs) and “bad” cholesterol (LDLs), but cholesterol itself is not problematic. The problem is due to the nature of the particle that carries the cholesterol itself. LDL is encased in a protein called apolipoprotein B (or apoB), whereas HDL is wrapped in apoA. Every single lipoprotein that contributes to atherosclerosis (not only LDL, but several others) is wrapped in apoB. This is why when we get to testing, Attia considers apoB a more valuable to test for as an indicator of risk, than the markers we historically measure like LDL and HDL counts.

Second, dietary cholesterol doesn't seem to have much to do with heart disease:

Eating lots of saturated fat can increase levels of atherosclerosis-causing lipoproteins in blood, but most of the actual cholesterol that we consume in our food ends up being excreted out our backsides. The vast majority of the cholesterol in our circulation is actually produced by our own cells.

In 2015, the US advisory committee for nutrition guidelines conceded “cholesterol is not a nutrient of concern for overconsumption.” It's a little confusing because, when talking about nutrition, Attia does talk about how diet affects cholesterol. But it seems like the total cholesterol amount of the foods you eat doesn't matter much, whereas the kinds of cholesterol (saturated fats, polyunsaturated fats, monounsaturated fats) does. We'll get to that in part three.

Not Just For The Elderly

When we think about heart disease, most of us picture an elderly person clutching their chest. But half of major cardiovascular events (heart attack, stroke, etc) occur in men before age 65. A third occur before then for women.

Critically cardiovascular disease is a slowly building process that starts long before you are old, in some cases as early as in your teens! (As a teenager who ate a lot of Chik-fil-A in Texas growing up, I am probably in this category.) But it's also not a one-way street. HDL cholesterol, through better nutrition, can clear away cholesterol clinging to arterial walls.

For many people nutrition is not enough. The good news is that unlike for cancer or neurodegenerative diseases (spoilers), we have excellent drugs to fight heart disease. We have a host of statins that can be prescribed for cholesterol, and notably most people can handle them with minimal side effects. For those people, Attia (per his Medicine 3.0 philosophy) advocates using statins early and often. This is because it's all about total exposure to apoB over time. If the side effects are minimal, and we know that statins can push apoB numbers down, why shouldn't we utilize them earlier to fight back?

This is an aggressive view of health management, and (as far as I can tell) not widely adopted. But who knows, perhaps this will change over the coming years. Until then, there’s plenty we can still do to fight the good fight.

Cancer: The Wild Card of the Four Horsemen

Of all the health topics out there, cancer might be the scariest. It's the bogeyman of diseases, the thing that lurks in our anxieties.

Unfortunately, unlike metabolic dysfunction and heart disease, cancer is an area we have much less control over. Some of us will simply be unlucky.

Screen Early, Screen Often

The main takeaway for cancer from Outlive is that once it is established in the body, we have crappy tools to fight it. Breast cancer and prostate cancer only kill when they become metastatic and spread.

The good news? You can live without breasts or a prostate organs, which means if we catch these cancers early, our chances of survival skyrocket. While we can hope for better treatments in the future, currently our best hope remains to try and catch them as early as possible, when are treatments can be most effective. Because of this, Attia’s strategy for cancer is centered around aggressive screening.

What Can We Actually Do?

Besides screening, there does not seem like there is a lot we can do to prevent cancer. The only risks we have some control over seem to be not smoking (smoking puts you at 10-25x more risk for lung cancer), insulin resistance, and obesity. Remember our old friend metabolic syndrome? Turns out it might be playing for team cancer too. While the evidence for the link between insulin resistance, obesity, and cancer is less strong, it's still just another reason to focus on metabolic health.

Beyond this, there isn't much else to say. Outlive covers a bunch of possible new treatments, which will hopefully bear fruit some day, but that don't really affect those of us without cancer right now.

Share

Neurodegenerative Diseases: The Final Frontier

If you thought our lack of tools to combat cancer was depressing, get ready for more. We do have some genetic screens that can indicate a higher likelihood of developing a disease, but by and large we are worse off here both for screening and treatment than for cancer.

Partly this is because we are generally not that good at understanding this horseman, which limits our ability to come up with effective treatment and prevention plans.

Possible Theories

One theory of the cause of neurodegenerative disease is vascular disease resulting reduced blood flow, for which there is strong circumstantial evidence. If true, then doing all the things we already want to do to fight heart disease will help our brains too.

Another theory is—you guessed it, the recurring scourge—insulin resistance. Having type 2 diabetes greatly increases the risk of developing Alzheimer's disease. (He says "doubles or triples" here but there's no footnote.) If this is the case, doing what's good for metabolic health, will also help the brain.

So while things may look bleak for this horseman, the best weapons we have are ones we were going to deploy anyway against the other horsemen. Yay, I guess? No extra homework!

Attia does mention exercise as the most important tool we have for fighting neurodegenerative diseases, but it also seems to be the most effective thing for fighting, well… everything, as we’ll discover in part three.

tl;dr

To summarize:

• Metabolic syndrome is very, very bad. DO NOT GET DIABETES. It is a source of cascading problems that buffs every other horseman. If the Four Horsemen had a leader, it would be MetSyn.

• Heart disease is also bad, but we have very good drugs in statins to help us.

• We can do a lot to combat both MetSyn and heart disease through exercise and nutrition, which we're about to get into next.

• Cancer and neurodegenerative diseases are less well understood and we have much weaker tools to fight them. For cancer our best bet is early detection. If we succeed at fighting MetSyn and heart disease, we'll incidentally be improving our odds against both of them as well, which is kind of nice because frankly this is a lot already.

Now on to the good stuff: what the hell do we do about all this?

Subscribe now

I was introduced earlier this year to Peter Attia’s book Outlive by my friend Aadil Mamujee, who wrote a post on taking a broader look at his health and inspired me (and Lindsay) to do the same. I am turning 40 this year, and it seemed like a good time to take a look at my overall health. Aadil shared his notes from Outlive with me, but I was inspired to read the entire book. Because I read it on Kindle, I was unprepared for just how large a tome it was—it is effectively a textbook, a fact which I discovered only after buying a physical copy to keep as reference.

Outlive is worth reading in full, but it could benefit from some prioritization and editing. Suggestions like “use a dry sauna at 179F or hotter four times a week to combat cardiovascular disease”—not entirely practical for most people, and certainly less critical—are scattered in the text alongside the important stuff.

The book also frequently jumps from topic to topic. Attia discusses the importance of zone 2 endurance training in the early chapter on metabolic syndrome, but then returns to it hundreds of pages later in the chapter on exercise. If you’ve listened to any of his (typically 2+ hour) podcasts, you’ll know what I’m talking about.1 Even the community crowdsourced summary is fifteen pages long. It’s not entirely his fault; health does appear to be (unsurprisingly, annoyingly) interconnected.

So this is my effort to make Outlive more actionable and organized. I’ve spent hours going through the book and other material online to try and condense things, as well as choosing to include only the bits I think are most widely actionable (e.g., we’re not going to talk about taking rapamycin). Just to be clear, this is my summary which explicitly doesn’t cover everything in the book—I hope it is as useful to you as it has been for me.

Unfortunately even the condensed version would be too long if I tried to cram it into one post so my summary is going to be divided into three(!) parts.

In part one, we’ll talk about the setup: the guiding principles behind Outlive.

In part two, we’ll excitingly cover all the ways you are most likely to die.

Finally, in part three, we’ll discuss what you can do about it.

Subscribe now

The framework

Outlive spends a lot of time establishing underlying principles and framework for approaching health and longevity, before getting into the meat of things. The book’s basic approach is this:

1. We should focus on prevention of disease instead of only on treatment.

2. We should work backwards from what activities we want to do when we’re older (Attia specifically uses centenarians, or 100+ year olds as the target).

3. We should understand what our biggest risks are, and use tests to diagnose which ones we are specifically at greater risk for versus.

4. Finally, we devise a plan to both a) help us achieve our centenarian goals in light of natural fitness decay as we age, and b) mitigate our biggest risks. (Spoiler: it involves exercise.)

Prevention instead of only treatment

A key concept in Outlive is what Attia calls “Medicine 3.0”. The short version is that current medicine—Medicine 2.0—focuses on treating disease, whereas Medicine 3.0 focuses on preventing it.2

With Medicine 2.0, when you discover you have a problem (usually due to some acute event or pain) we jump into action to try and treat the condition. Medicine 2.0 is centered on treatment, and it focuses on short-term risk.

For example, if you are young and eating too much McDonald’s, your ten-year risk of a heart attack is likely lower than a 75-year old who eats a more balanced diet, even if you would end up less healthy than them by age 75 if you kept it up. Because the 75-year old’s risk is higher simply due to age, Medicine 2.0 would rather treat them first.

Medicine 3.0 says if our goal is to have long healthy lives, we should be working to fight and reverse these problems early on, instead of ignoring them until our ten-year risk crosses some arbitrary threshold.

Attia compares this to how we now view smoking:

We know that smoking is causally linked to lung cancer. Should we tell someone to stop smoking only after their ten-year risk of lung cancer reaches a certain threshold?

This sounds absurd, but is actually how we manage our health in many areas. For some reason, our cultural norms around avoiding an unhealthy lifestyle that leads to our most common causes of death have not yet caught up to our views on smoking. We are mostly content to let things be until a red alert sounds in the form of bad blood test results, or worse.

You might wonder whether this topic is really that important to include in an already long book. Attia doesn’t say this, but my view is that you need to buy into this philosophy to bother doing all the work needed to make a difference. Continuing to not smoke is not hard if you have never been a smoker. But taking on the amount of exercise and other activities Attia covers in Outlive takes a lot of time. Like, a lot of time.3

You have to believe it’s worth it, because watching Netflix on the couch is much, much easier.

Share

Working backwards

So maybe you’re sold on the philosophy of prevention. That’s half the battle—as much as we can, we want to prevent getting afflicted with the most deadly diseases. The other half of Attia’s philosophy is figuring out what we want to do with our lives, specifically our lives as elderly people.

Enter what he calls the “Centenarian Decathlon”. That is, literally what are ten activities that are meaningful to you that you want to be able to do when you're a hundred.

People have a conceptual idea of what they will do when they’re old, like play with their grandkids or travel. But fitness declines basically across all metrics as we age. We lose strength. We lose aerobic fitness. We lose muscle mass. It's one thing to imagine these things now when you're young, and another to have the fitness to do them decades from now.

He gives an example of wanting to be able to pick up your grandkids:

Let’s say the kid weighs twenty-five or thirty pounds. That’s basically the same as doing a squat while holding a thirty-pound dumbbell in front of you (i.e., a goblet squat). Can you do that now, at age forty? Most likely. But now let’s look into the future. Over the next thirty or forty years, your muscle strength will decline by about 8 to 17 percent per decade—accelerating as time goes on. So if you want to pick up that thirty-pound grandkid or great-grandkid when you’re eighty, you’re going to have to be able to lift about fifty to fifty-five pounds now. Without hurting yourself. Can you do that?

By identifying your Centenarian Decathlon you can create a plan to help make it happen. Or at least in theory. In reality, throughout Outlive, the prescription seems to be the same regardless of what activities you choose. (We'll get there in part three.) That's okay because most of us want to be some version of “a functional human” and the exercises you need to accomplish that are generalizable, but still it feels like an underdeveloped idea.

Regardless of what your Centenarian Decathlon looks like, working backwards helps us avoid what Attia dubs the “marginal decade”: For many people the last ten years of their life are not great. Medicine 2.0 has gotten better at increasing lifespan, but that doesn’t mean what you’re up to during that extended lifespan is very fun. Outlive proposes the idea of healthspan, which is a subjective (but useful!) measurement of how many healthy years you live, and therefore get to actually enjoy.

Improving your healthspan is sort of like building up a retirement account. You know that later in life you won’t be earning as much and want a nest egg of cash built up you can slowly draw on so you don’t have to go back to work. Given that our fitness also declines steadily as we age, we similarly want to aim to be as healthy as possible now to counteract the drawdown.

And that’s it! Between the ideas of prevention over treatment, working backwards, and the concept of healthspan, you are caught up on roughly 17% of the book. Next up, we dive into the details on the major afflictions out to get us: the Four Horsemen.

Anyways, onwards with Patch Notes #15:

• In a previous Patch Notes I joked about how Sam Altman and Patrick Collison’s complaint about a lack of young founders was… unfounded; it’s just that a lot of them seem to end up committing fraud and going to jail (SBF, etc). Here is The Diff agreeing that average tech founder age seems to be going up, but that one area where there is still room for young founders is hacking:

  The average age of prominent tech founders seems to have gone up a bit in the last decade. There are plenty of very prominent thirty-and-over founders, but fewer wunderkinder than in the early days of social. There is still one area where people in their early twenties can end up running large organizations: the head of the Scattered Spider hacking group was arrested, and turned out to be 22.

  Scattered Spider was the group behind the data ransom attacks at MGM and Caesar’s last year, as well as MailChimp, LastPass, Signal, and Twilio. Pretty impressive! Maybe it says something that smart tech-savvy youth are going down the path of fraud, hacking, and crime instead of founding tech companies. It certainly doesn’t seem like lack of ambition or skill.

• Bumble’s Whitney Wolfe Herd says your dating ‘AI concierge’ will soon date hundreds of other people’s ‘concierges’ for you:

  “If you want to get really out there, there is a world where your [AI] dating concierge could go and date for you with other dating concierge,” she told host Emily Chang. “Truly. And then you don’t have to talk to 600 people. It will scan all of San Fransisco for you and say: ‘These are the three people you really outta meet.'”

  I wrote about AI’s writing and reading to each other awhile ago, and well here we are. There is something kind of odd about this idea though, besides the obvious dystopian vibes. If my AI concierge goes out and talks to other AI concierges and comes back with some matches, at what point is it materially different from any existing matching algorithm today? Dating sites already do some behind the scenes work to find good you potential matches—is this somehow better?

  Subscribe now

• AI-Generated Version of Al Michaels’ Voice Will Deliver NBC’s Paris Olympics Recaps

  The veteran sportscaster will deliver personalized recaps of the media company’s coverage, all of it tailored to fans’ specific interests. Each morning, NBC’s Peacock streaming hub will deliver a video summary of the last day of coverage, built out of clips and a high-quality AI re-creation of Michaels’ voice, which was trained using past appearances on NBC and matches his signature way of speaking.

  
  “When I was approached about this, I was skeptical but obviously curious,” said Michaels — presumably the real one — in a prepared statement.

  In my memory, there is a long-running joke about how they kept Dick Clark in cyrofreeze, and wheeled him out each year for Dick Clark’s Rockin’ New Year’s Eve, which explained how he was somehow always there, and perpetually the same age.1 Anyway with AI now it’s even easier! Maybe they’ll bring him back this year.

• We Need To Rewild The Internet is a long piece lamenting the current state of the tech monopoly-driven internet versus the golden years of the open internet:

  The internet’s 2010s, its boom years, may have been the first glorious harvest that exhausted a one-time bonanza of diversity. The complex web of human interactions that thrived on the internet’s initial technological diversity is now corralled into globe-spanning data-extraction engines making huge fortunes for a tiny few.

  I’m sympathetic to this view—I worked on Google Reader during the heyday of the “blogosphere”, a now hilariously archaic word. The early internet did feel magical and original. That was in part due to how difficult it was to make anything: In a world before the nth JavaScript framework, or GitHub (or git!), or React, or nocode, or Shopify, or AWS, the internet had a sort of homespun, artisanal quality to it.
  
  But creating the tools and platforms to make the internet flourish wasn’t free. The tech giants built things—open source or otherwise—because they believed that as long as the internet grew, they would grow with them. And they were right, it worked! But it wasn’t just a tale of rampant pirate capitalism; we liked what the internet became. Ben Thompson has written extensively about this on his newsletter Stratechery. Here’s a recent piece:

  In short, the analog world was defined by scarcity, which meant distribution of scarce goods was the locus of power; the digital world is defined by abundance, which means discovery of what you actually want to see is the locus of power. The result is that consumers have access to anything, which is to say that nothing is special; everything has been flattened.

  He goes on to give a bunch of examples: Amazon created the “Everything Store” with practically every item on Earth and the capability to deliver it to your doorstep; Google broke down every publication in the world into individual pages; search results didn’t deliver you to the front page of a newspaper or magazine, but rather dropped you onto individual articles. And so on. Most people would not choose to go backwards to the old way!
  
  I agree that something feels amiss with the modern metrics-driven internet. (I wrote about it recently!) But I am always torn when I read articles like this one because it tells a story that we as internet consumers were hoodwinked and “herded into rigid tech plantations” that isn’t entirely honest. Most of us went willingly.

  

• I Cannot:

  Last year, a formal tone that sounded nothing like my speaking voice started to sputter out from my cursor and onto the page: “I cannot think about it now,” “I journeyed back to my abode.” Words elongated, and phrasings—strange ones—appeared. I watched the sentences extend, and noticed they were saying very little, but that they were saying this little in very mannered ways.

  I loved this piece, and am guilty of doing this at times, along with many other writing sins. But the story is about more than just formal writing:

  Like many phenomena noticeable for their formal gestures at nostalgic extremity—the starched high-collared dresses, breakfast cereal made from scratch, the handlebar mustaches—the slightly formal essays point toward, I think, the threat they are meant to oppose: a feeling that things are too much online, that things are too casual and must be elevated.
  […]
  Amidst the tone of graven importance the writers of these essays, maybe there’s not much to say that feels new—or if there is, we are often side-stepping it. In her book Hole Studies, Hilary Plum points out how contemporary essayists, she says, write “I’ve been thinking a lot about . . .” and “then just virtuously mention a subject, not saying one thing of substance about it, moving on before we have to do any work.”

  I have nothing to add except that I really think it is so good and so I will say nothing of substance and instead tell you to just go read the whole thing.

• Feeling Blessed: At the Habsburg convention in Plano by Christopher Hooks:
  There was apparently a Habsburg (yes, those Habsburgs) revivalist convention in Plano, Texas, a few miles from where I grew up. There are so many gems in this piece, including Texas being Habsburg land:

  Tales were told of a time and place when there was a path, whether those paths were “being the Habsburg emperor” or “serving the Habsburg emperor.” Eduard Habsburg, currently the ambassador to the Vatican of Viktor Orbán’s regime in Hungary, noted that Texas was once Habsburg land—through the descendants of Charles V of Spain, who oversaw the boom years of Spanish colonialism. To his mind, he said, it still is. The audience cooed.

  On the push to win sainthood for Emperor Karl I, the last Habsburg:

  After he became emperor, he enlisted the help of German chemical warfare experts for his own army. At the Twelfth Battle of the Isonzo in 1917, his forces sprayed the Italian army with poison gas and smashed them, leading to some three hundred thousand enemy casualties, a defeat so bad it helped pave the way for Mussolini’s rise to power five years later. Karl is perhaps the only current candidate for sainthood who has drowned his enemies in chlorine gas.

• This 2014 post on Japanese zoning laws and the differences vs US zoning shows how Japanese cities end up with a greater diversity of commercial and residential spaces.
  
  A big difference is that in Japan, when an area is zoned for something, say large commercial stores, it automatically gets zoned for “lesser” (as in less disruptive) categories in the same zone, like neighborhood shops, or smaller apartment buildings:

  Same thing for offices, if they want to build offices, they must zone for at least a Category II mid-rise residential zone, which means in practice apartments with at least 3-4 stories. If they want to allow bigger offices, they must also allow bigger residential buildings and stores in the same zone. This "bundling" of uses imposes mixed use developments on cities.

  In the US, we typically zone areas for one specific use like schools or commercial. Building something different requires going through the onerous process of changing the zoning. In Japan, the flexibility to build a variety of things, as long as you are under the zone’s maximum “nuisance level”, results in a more fluid (and efficient) use of space according to supply and demand:

  If there is more land than needed for commercial uses in a commercial zone, for example, then you can still build residential uses there, until commercial promoters actually come to need the space and buy the buildings from current residents.

• A chart from Andrew Van Dam at the Washington Post. If we’re thinking in terms of patch notes for reality, players are not fans of the current patch:

  

  Some of these are debatable, but some are objectively wrong. (The most crime? The most war?) Some of the subjective ones seem suspect too. Really, the worst cuisine or the worst television? But there’s more going on here:

  Marketing researcher Bill Page said that by broadly asking when music, sports or crime were worst, instead of getting ratings for specific years or items, YouGov got answers to a question they didn’t ask.
  

  “When you ask about ‘worst,’ you’re not asking for an actual opinion,” Page said. “You’re asking, ‘Are you predisposed to think things get worse?’”
  

  “There’s plenty of times surveys unintentionally don’t measure what they claim to,” his colleague Zac Anesbury added.
  

  YouGov actually measured what academics call “declinism,” his bigwig colleague Carl Driesener explained. He looked a tiny bit offended when we asked if that was a real term or slang they’d coined on the spot. But in our defense, only a few minutes had passed since they had claimed “cozzie livs” was Australian for “the cost of living crisis.”1

  Thanks for reading Speedrunning Towards Bethlehem! Subscribe for free to receive new posts and support my work.

• Sorry, I only drink wine that was aged listening to the Nier Automata soundtrack now:

  Japanese retailer Onkyo Direct are selling two varieties of official Nier: Automata wine, matured in barrels with speakers attached to them that exclusively pump out the character action games’ soundtrack. The two red wines - named 9S and 2B for Automata’s android heroes - were each force-fed different playlists selected from the game’s soundtrack.

  

• Finally, speaking of music, I recently went to EDC again after my first trip last year. I’ve been mulling over writing something new about it, but in the meantime, here are a few of my favorite sets from this year.

  • Four Tet:

  • Excision—not at all what I expected to enjoy, but… you had to be there, especially for the unhinged visuals:

  • Sub Focus—I sadly missed seeing this live, but it has been one of my favorite sets uploaded, especially since the sunrise set for Worship doesn’t have a full recording:

  • (Updated) WORSHIP—the day I after I posted this, the set finally got uploaded! Probably my favorite set of the weekend:
    

• Still thinking about this piece by on Balatro (which I still haven’t played) that includes this lovely passage:

  The reason gambling is fun is that we are not our reward function. Maybe once we were, maybe once, long ago we were nothing but a little knot of signals that was able to remain stable far out of equilibrium with its environment, a little loop of Bayesian inference that was able to climb the signal gradient towards more and more complex and robust forms of stability. But now that we know that, now that we can look back and see that, we are no longer fully embedded in our reward function, it has become something we can use, something we can play with, instead of simply who we are. With some effort we can partially extract ourselves from the loom of our reward function, look around, and ask: “Where am I?” and “Where do I want to go?”

• If GeoGuessr was too easy, here’s TimeGuessr, where you have to guess both the location and year of specific photos:

• A Pew Research Center survey on on teens and video games shows 85% of teens play games daily, with four-in-ten playing daily:

• A report on TSMC’s troubles building their Arizona fab and the clash between Taiwanese (read: “hardworking”) and American (“lazy”) work culture. Pair with Noah Smith’s follow-up Who’s afraid of East Asian management culture?.
  
  Generally I agree with Noah’s more balanced view—TSMC can’t possibly expect things to go well only communicating in Mandarin and Taiwanese in Arizona—though on the other hand, here’s The Guardian on The soft life: why millennials are quitting the rat race. So maybe there is a grain of truth. I’m certainly uninterested in grinding my face off working these days.

• I’ve gotten into some debates with friends on whether AI is net positive or negative. Here’s a piece on AI audio notching a tally in the negative column:

  In fact, the deleterious effects of A.I. seem much more likely to be felt--and indeed, as the stories of attempted fraud and petty revenge suggest, are already being felt--on a much smaller, more local, generally tawdrier scale, despite the planetary claims of tech execs testifying to Congress. New kinds of targeted scams and new types of local gossip rather than mass propaganda campaigns; less “state-sponsored misinformation” and more “provincial-psychopath-looking-to-settle-a-school-athletics-grudge-sponsored misinformation.”

  I did admittedly like this deepfake of Toby Keith singing a Maoist anthem:

  Subscribe now

• Ne-Yo’s Tiny Desk concert is great (h/t Sebastian):

  I had no idea he was the writer behind so many other massive hits like Rihanna’s Take a Bow, and Beyoncé’s Irreplacable.

• TikTok is obsessed with ... premium-grade industrial glycine from China? (h/t Justin)

  

  @citiesbydianaDon't get CUCKED by inferior glycine suppliers. Donghua Jinlong's Industrial Grade Glycine is the superior choice. 👸🏻🛣️🏭🇨🇳 THIS IS NOT SPONSORED CONTENT OR AN AD THIS IS A PARODY FAN POST. #glycinetok #industrialglycine #brainnourishment #bestglycine #satire #brainrot #donghuajinlong #glycine THIS IS NOT SPONSORED OR BRANDED CONTENT! THIS VIDEO IS SATIRE. I have NO AFFILIATION with Donghua Jinlong, this is just a fan edit. 🥰🥰

  Tiktok failed to load.
  
  Enable 3rd party cookies or use another browser

• I read David Grann’s The Wager awhile back—it’s a true shipwreck and mutiny story and I blew through it in a couple days. I thoroughly enjoyed it and will not spoil anything here, but I do love this bit on the etymology of sayings we take for granted today that all come from British naval terminology:

  During the age of sail, when wind-powered vessels were the only bridge across the vast oceans, nautical language was so pervasive that it was adopted by those on terra firma. To “toe the line” derives from when boys on a ship were forced to stand still for inspection with their toes on a deck seam. To “pipe down” was the boatswain’s whistle for everyone to be quiet at night, and “piping hot” was his call for meals. A “scuttlebutt” was a water cask around which the seamen gossiped while waiting for their rations. A ship was “three sheets to the wind” when the lines to the sails broke and the vessel pitched drunkenly out of control. To “turn a blind eye” became a popular expression after Vice-Admiral Nelson deliberately placed his telescope against his blind eye to ignore his superior’s signal flag to retreat.

  And:

  When ailing seamen were shielded belowdecks from the adverse elements outside, they were said to be “under the weather.”

• SQLite (yes, the database) has a code of ethics governing their interactions with each other and the SQLite community “in accordance with the "instruments of good works" from chapter 4 of The Rule of St. Benedict (hereafter: "The Rule"). This code of ethics has proven its mettle in thousands of diverse communities for over 1,500 years, and has served as a baseline for many civil law codes since the time of Charlemagne.” I really do not know what loving fasting (rule #13) or being in “dread of hell” (rule #45) has to do with SQL databases, but maybe it can’t hurt. (h/t Lee)

• A great visualization of Pareto optimization of Mario Kart 8 cart/driver combinations (also h/t Sebastian)

Okay that’s all for now; if anything particularly resonated with you, let me know, and see you next time!

The Honest Broker

The State of the Culture, 2024

The President delivers a ‘State of the Union’ Speech every year, but that’s a snooze. Just look at your worthy representatives struggling to keep their eyes open. That’s because they’ve heard it all before. We have too. Not much changes in politics. Certainly not the candidates…

Read more

2 years ago · 5228 likes · 554 comments · Ted Gioia

And in particular, this chart:

A friend reached out to ask whether this was just showing the process of modernization; whether film → TikTok was any different from handcrafted pottery → injection-molded mass-produced dinnerware. He wondered whether there was something about these industries that Gioia was decrying, or if it was a more general get off my lawn rant about modernity. While I obviously can’t speak for Gioia, I do have some suspicions.

As it happens, before Ted became a prolific Substacker, he was (and is?) a jazz musician and writer. I own two of his books, How to Listen to Jazz and The Jazz Standards, and recommend them both.1 Unlike EDM, pop, or country, jazz steadfastly refuses to be domesticated and mass-produced.2 Maybe the improvisational nature of the genre makes it inherently resistant to standardization. Or its penchant for really long songs and live sessions.

Jazz is not an easy genre to learn. Playing an instrument well is never easy, but jazz demands all the same technical skills as any other genre while adding on improvisation, significantly more complex chords and harmonies, and rhythmic challenges ranging from syncopation to how-is-he-keeping-track-of-the-beat drum solos (jump to about 4:40). It’s hard to imagine a glut of short-form jazz inundating TikTok because it’s hard to make enough of it.

So at some level, I suspect if you’re a serious jazz musician today, you can’t help but view the present state of culture with dread. Perhaps Ted harbors some mild artisanal distaste of mass-produced culture. I can’t blame him.

But I don’t really think that’s the crux of his complaint. His complaint centers on the evolution towards instant-gratification dopamine release, not mass production. No one—retail addiction edge cases aside—is getting high off buying decorative pottery at Target.3 The problem he is critiquing is the habit-forming addictions enabled and optimized for by modern tech. Meta wants you to keep scrolling on Instagram, and the PMs who work there are explicitly measured against KPIs that support those goals.

It’s the democratization and mass production of content combined with the metrics-driven optimization that is assaulting us today.

I’m reminded of this excellent interview with C. Thi Nguyen, a professor of philosophy who writes about games:

Think Like A Game Designer

C. Thi Nguyen — Game Design as Art, Fish Taco Adventures, and Exploring Game Design Principles in Daily Life (#61)

Get the latest articles delivered to your inbox and contribute to our community by opting for a free or paid subscription. Paying subscribers enjoy an abundance of extra game design content and an additional exclusive newsletter with new lessons, archive access, videos, and more…

Listen now

2 years ago · 4 likes · Justin Gary

The podcast is worth listening to in its entirety, but one part that stuck with me is a discussion on metrics. Nguyen says metrics are “adapted to be comprehensible at scale”; to be used in situations that don’t require high context or background knowledge. He gives an example of it being hard to understand some weird delight you have in playing a particular game, versus a metric like sales data. If you’re a developer, it’s hard to optimize your game for delight!

Subscribe now

This may sound suspiciously like legibility, and it is, but Nguyen is not anti-metrics. He acknowledges that you simply can’t function at scale without metrics. Large scale inherently comes at some cost of context. The PMs on Instagram are not malicious cogs in some infernal dopamine machine. They are like all of us, conscripted into the modern world of apps and optimization that has subsumed us all.

But as an individual, Nguyen warns us that if you get completely captured by metrics, you are “letting the demands of large scale and utterly contextless understanding strike into your soul”.’

…which I think is what Gioia and so many others are reeling from today.

We were unprepared for the dual onslaught of ever more powerful dopamine hits fused with the relentless metrics-driven efficiency of modern technology. And as we endlessly consume content engineered to appease contextless gods, we risk surrendering our taste to the leveling forces of instant gratification. For legibility, uniqueness is a bug, not a feature.

The remedy may look something like learning to play jazz: a long road towards mastery, fueled by your own tastes and desires over public approval, and a refusal to optimized. Jazz’s emphasis on spontaneous improvisation and allergy to rigid formulas might hold a lesson for us that being comprehensible at scale is overrated.

It is not an easy road, but nothing worthwhile ever is.

The name of this Substack, Speedrunning Towards Bethlehem, is nod towards Joan Didion’s Slouching Towards Bethlehem, which is a collection of essays (and titular essay) from 1968 that captures the turbulence of the 60’s. Ever since I read them, nothing has managed to condense the tone of a era quite like Didion. I’m not sure what precisely is missing today, but when I find something that glimmers with a hint of the absurdity of life today, I file it away in a folder called “modernity”. This includes the 2011 Tumblr Yelping with Cormac which features a review in the style of Cormac McCarthy of the gleaming Apple Store in San Francisco:

What were all these people waitin for. He told me it was for a apple phone or some such. I said dont these folks have telephones already? He told me they all had apple phones but it was the older one. I asked him what would happen to the old apple phones. He told me about a fella named Craig had a list and everbody sold their old telephones on it. A telephone sellin list.

Well I told him that all made about as much sense as a horse with two heads and he laughed like that was the funniest thing he ever did hear. Said he was goin to twinkle it. I left before he said anythin else that didnt make no sense and I went to the nearest bar and ordered a double whiskey and sat there drinkin it. I guess I sat there for a long time. Wonderin if when Rome was fallin all the Romans was standin in line waitin to get that new chariot or the like. The barbarians at the gates and them just standin there waitin.

He was ahead of his time—the barbarians would not arrive for another decade. But here, post-2020, the vibes were finally way off, and my “modernity” folder fills up faster every day.

Anyway, here is the latest addition:

Degen communism

Here’s Vitalik Buterin, the creator of Ethereum—which is definitely not a security—proposing degen communism, a hybrid of dynamic chaos and emphasis on social good:

What does the internet of the 2020s - not the "respectable" internet of Substack, not a hypothetical version of Twitter where the bad people and somehow only the bad people are censored, but the real internet as it exists today - fundamentally want? The answer is, it wants chaos. It does not want gentle debates between professionals who "disagree on policy but agree on civics". It wants decisive action and risk, in all its glory. Not a world with genteel respect for principles, where even the loser peacefully accepts defeat because they understand that even if they lose one day they may still win the next, but a world with great warriors who are willing to bet their entire life savings and reputation on one single move that reflects their deepest conviction on what things need to be done. And it wants a world where the brave have the freedom to take such risks.

At the same time, the general welfare of humanity demands a greater focus on the common good. We've seen too many instances of epic collapses, orchestrated by failed machinations of the elites, where the common people end up screwed but the elites remain immune or even benefit.

His view of what makes something “degen” outside of crypto roughly equates to “dynamic”. For example Harberger taxes on intellectual property:

This would work as follows. For a copyright or patent to be valid, whoever owns it must publicly register a value, which we will call that copyright or patent's "exclusivity price". They must then pay 2% of the exclusivity price annually in tax (they can change the exclusivity price at any time). Anyone can pay the owner the exclusivity price, and get an unlimited right to also use (and if they wish sub-license, including to the entire world) that copyright or patent. The original owner would retain the right to use in all cases; others can gain permission to use either by getting the original owner's permission, or by paying the owner the exclusivity price.

I don’t know enough to comment on whether I think his ideas have merit, but I think he’s right that “there is a widespread feeling throughout the Western world that all of our political ideologies are outdated, and are increasingly failing us”. It reminds of what I felt reading Ted Gioia’s piece from last week’s Patch Notes: We can mostly agree that things do not seem to be going quite how we would like, but it’s unclear what anyone is proposing we should do about it. At least Vitalik is trying.1

More on dopamine culture

Speaking of Gioia’s piece, Jonathan Haidt also wrote a recent article on a very related topic around child/teen use of phones, which has a bunch of interesting bits.

On cable news contributing to the rise in safety concerns and associated overprotection:

The changes started slowly in the late 1970s and ’80s, before the arrival of the internet, as many parents in the U.S. grew fearful that their children would be harmed or abducted if left unsupervised. Such crimes have always been extremely rare, but they loomed larger in parents’ minds thanks in part to rising levels of street crime combined with the arrival of cable TV, which enabled round-the-clock coverage of missing-children cases. A general decline in social capital––the degree to which people knew and trusted their neighbors and institutions––exacerbated parental fears.

Teens apparently spend five hours a day on social media platforms, which does sound like a lot, but then US adults are spending five hours a day watching TV, which seems barely better:

The numbers are hard to believe. The most recent Gallup data show that American teens spend about five hours a day just on social-media platforms (including watching videos on TikTok and YouTube). Add in all the other phone- and screen-based activities, and the number rises to somewhere between seven and nine hours a day, on average.

And evidently a majority of students (sampled in a study, caveat emptor) would prefer to live in a world without TikTok or Instagram, but are trapped in the mutually-assured destruction world of social media proliferation and unable to opt out because everyone else is there:

A recent study led by the University of Chicago economist Leonardo Bursztyn captured the dynamics of the social-media trap precisely. The researchers recruited more than 1,000 college students and asked them how much they’d need to be paid to deactivate their accounts on either Instagram or TikTok for four weeks. That’s a standard economist’s question to try to compute the net value of a product to society. On average, students said they’d need to be paid roughly $50 ($59 for TikTok, $47 for Instagram) to deactivate whichever platform they were asked about. Then the experimenters told the students that they were going to try to get most of the others in their school to deactivate that same platform, offering to pay them to do so as well, and asked, Now how much would you have to be paid to deactivate, if most others did so? The answer, on average, was less than zero. In each case, most students were willing to pay to have that happen.

Social media is all about network effects. Most students are only on it because everyone else is too. Most of them would prefer that nobody be on these platforms. Later in the study, students were asked directly, “Would you prefer to live in a world without Instagram [or TikTok]?” A majority of students said yes—58 percent for each app.

Haidt proposes some norms, such as no social media before age 16 at the end of the piece to try and address the issues. He suggests that we can make inroads at the community level to tackle the collective-action problem of everyone wanting out but no one individually being willing to exit, but I have low hopes that there are communities with enough agreement on the issue (and solutions) to cross some critical threshold. And it’s not like communities have hard social boundaries—except those like the Amish, who are presumably fine. If one school district implements a policy, and the neighboring one doesn’t, kids are still going to interact with each other at sports games, the mall, or whatever it is that kids do these days. The real world is messy, and the internet knows no bounds.

One bit I did find amusing was on how newer generations are supposedly more risk averse:

In an interview last May, OpenAI co-founder Sam Altman and Stripe co-founder Patrick Collison noted that, for the first time since the 1970s, none of Silicon Valley’s preeminent entrepreneurs are under 30. “Something has really gone wrong,” Altman said. In a famously young industry, he was baffled by the sudden absence of great founders in their 20s.

I mean I don’t know. The Forbes 30 Under 30 is famously a “pipeline to prison”. Between SBF, Elizabeth Holmes, and Charlie Javice, maybe we have the opposite problem?

Miscellany

• Here’s a really really fantastic video explaining on GPTs work:

• Forget social media, Mambo No. 5 played with bike horns is actually what the internet was made for:

Have a great weekend!

Years ago, one of my friends applied for a vanity license plate MUNCHES in California and was summarily rejected by the DMV, I guess on the grounds that it could be some sexual innuendo? So I was amused to find that there is a Twitter bot that posts actual CA DMV vanity plate applications with comments from the DMV, I guess because it is all public record. The DMV rejects a lot of stuff you would expect like anything related to guns or gangs. Customers can add commentary explaining why their plate is legitimate (like claiming your last name is KOLT), but the DMV is pretty savvy and will use urban dictionary(!) to check things out.

Scrolling through the feed is a weird microcosm of seemingly arbitrary government bureaucratic decision-making: The DMV usually denies drug references, but BLAAZZN is ok. COOKIES are not. Innocuous standalone ‘B’s (which might mean “bitch”) like MANDY B are sometimes denied, sometimes approved. Fart references (GASPASR) are not okay. It is all delightfully Kafka-esque.

Longer reads

• Ted Gioia on the TikTok-ification of everything:

  

  I got into a discussion with a friend about whether AI was going to be net good or bad after last week’s bit on the enshittification of the internet. Ted’s critique of the current state of tech / content platforms is why I think it’s unfair to lay the blame on AI (yet): the cause of the rise of dopamine culture is the same as the one driving enshittification. Platforms profit from engagement, which rewards clickbait-y, dopamine-fueled content. We have been heading down this road for a long time now.
  
  What I’m not sure of is what Ted (or anyone) thinks we should do. Top-down CCP-style control is probably not what he wants. The EU famously deploys regulation with not much innovation to show for it. The US government looks dysfunctional. Maybe California will try something, but (per the DMV and craps rulings above) I’m not sure anyone will be happy about it. It’s not that I disagree with his concerns—half of what I write about is a vague sense of foreboding given the current intersection of tech, culture, and geopolitics—but I have no idea what I’d propose as an alternative. I suppose we will all find out together.

• Pankaj Mishra on the Shoah after Gaza (LRB): A long piece that follows in the wake of Masha Gessen’s In The Shadow of the Holocaust (New Yorker) on the use of the Holocaust in political discourse:1

  But it was only after the Six-Day War in 1967 and the Yom Kippur War in 1973, when Israel seemed existentially threatened by its Arab enemies, that the Shoah came to be broadly conceived, in both Israel and the United States, as the emblem of Jewish vulnerability in an eternally hostile world. Jewish organisations started to deploy the motto ‘Never Again’ to lobby for American policies favourable to Israel. The US, facing humiliating defeat in East Asia, began to see an apparently invincible Israel as a valuable proxy in the Middle East, and began its lavish subvention of the Jewish state. In turn, the narrative, promoted by Israeli leaders and American Zionist groups, that the Shoah was a present and imminent danger to Jews began to serve as a basis for collective self-definition for many Jewish Americans in the 1970s.

  I grew up reading (and still read) a lot of WW2 history, but only more recently have read more history of the post-war era (in the Middle East and elsewhere). There is a lot here, and it goes beyond just recounting the post-war history of the Shoah, and is well worth reading.

• My friend Drew made runner, a non-combat, movement-focused roguelike for the 7DRL challenge where you make a roguelike in seven days. I helped ever so slightly by “asking him probing questions” throughout the design. You can play it here.

  

• The Ship of Theseus is a “thought experiment about whether an object is the same object after having had all of its original components replaced.” The original example asks if you slowly replace all the components (planks, masts, etc) of a ship over time, once all the pieces are no longer original, is it still the same ship?
  
  Ok that’s all you need to know to appreciate this (h/t Lee):

  

• Super Mario Maker 1 was released back in 2015, and let users create custom Mario levels for others to play. As a creator, you had to beat your own level to prove it was beatable in order to upload it, but many levels (around 41,000+ as of February 2023) were so hard that they were never cleared. Nine years later, the Super Mario Maker 1 servers are getting shut down April 8,1 resulting in the deletion of all user-generated maps, and so the community has been racing to beat all remaining “uncleared” maps (maps that have never been beaten by anyone besides the original uploader). Over the last several months, Team 0% has been whittling away at the list, and this week they completed their task.

  

  But since this is gaming, it of course was not without a little drama.
  
  It turns out the very last level the team tried (and failed) to clear for a week (titled “Trimming the Herbs”) was actually an “illegal” level. It was only beaten by its creator when uploading via the use of a bot—a TAS or tool-assisted speedrun that lets a user program a perfectly precise set of inputs. While it’s theoretically possible for a human to perform the same sequence of inputs, it is usually practically impossible to execute consistently, making the level practically impossible.
  
  In the end after 280,000 failed attempts, the original level creator Ahoyo decided to come clean. And so in a sort of anti-climactic fashion, Team 0% had actually finished fully clearing the game a week earlier when the beat what they thought was the penultimate level before Trimming the Herbs.
  
  Here’s a video of the TAS clear Ahoyo uploaded when he created the level:

  The story is not quite over though as even though Trimming the Herbs was revealed to be illegal, players are still trying to beat the level manually for the ultimate cred before the servers shut down.2 (h/t Doug)

• In crypto-land memecoins have been on the rise again for the past couple of months, particularly on Solana. Here is a memecoin, $ALBEMARLE, which appears to be LARPing as an official coin of the small town of Albemarle, North Carolina. Their Twitter/X page features a mix of seemingly real reposts of actual content from the town, and maybe completely made up stuff? The coin has a market cap of $1.8M, and I can not explain any of it except to say that I no longer think any of it is weird, and that we should just get used to it. (h/t Andrew)

• Authors ‘excluded from Hugo awards over China concerns’ (The Guardian):

  Leaked emails from the organisers of the prestigious Hugo awards for science fiction and fantasy suggest several authors were excluded from shortlists last year after they were flagged for comments or works that could be viewed as sensitive in China.
  

  In January the Hugo awards published the statistics behind the 2023 awards, which were held as part of the World Science Fiction Convention (Worldcon) in the Chinese city of Chengdu in October. The data showed that the New York Times bestseller RF Kuang and the young adult author Xiran Jay Zhao were among authors who had received enough nominations to be on the ballot in their respective categories but were deemed “not eligible” by the award’s administrators, without further explanation.

  I’ve only read The Poppy War trilogy by RF Kuang, but based on that series at least, I’m extremely surprised that they would leave her out due to China concerns? The series is basically a fantasy retelling of the Sino-Japanese War with obvious stand-ins for the Japanese (the “Mugenese”) and the British (the “Hesperians”), neither of which are treated favorably in the book. I wonder if the Hugo admins simply omitted her because it dealt with China at all, versus having some particular anti-Chinese message. Either way, not great!

• Here lies the internet, murdered by generative AI:

  The fact that OpenAI was both honestly worried about negative effects, and at the same time didn’t predict the enshittification of the internet they spearheaded, should make us extremely worried they will continue to miss the negative downstream effects of their increasingly intelligent models. They failed to foresee the floating mounds of clickbait garbage, the synthetic info-trash cities, all to collect clicks and eyeballs—even from innocent children who don’t know any better.

  While I think the problems and risks listed here are all very real (and more relevant than p(doom) AGI risks), I think the author’s ire is misguided. We had plenty of clickbait headlines and content farm sites designed optimized for Google SEO long before OpenAI. Enshittification was coined by Cory Doctorow before the AI craze really took hold, and the original piece doesn’t mention AI at all. AI certainly makes it easier to enshittify the web, but the structural causes of things getting… shittier… already exist. The truth is, if you were paying attention, the internet was getting shitty already—maybe AI will force us to reevaluate how things work.
  
  Placing blame for enshittification aside, there seems to be a recent feeling that technology should be inherently good, and that something has gone seriously wrong for us to find ourselves where we are with AI.3 This feels misguided. Technology has always been amoral—we built the first computers to model the hydrogen bomb. The internet started as a defense project. Will AI have costs? Yes. Will AI do nothing “good” for the world? I would not take that bet.

AI Weirdness is a delightful blog that showcases funny AI-related results, including this one that is a perfect confluence of games & 3D engines and AI optimization:

Harvesting energy from crashing into the floor: Another simulation had some problems with its collision detection math that robots learned to use. If they managed to glitch themselves into the floor (they first learned to manipulate time to make this possible), the collision detection would realize they weren’t supposed to be in the floor and would shoot them upward. The robots learned to vibrate rapidly against the floor, colliding repeatedly with it to generate extra energy.

I love everything about this from my friend Kane:

Longer reads

• The FT ($) on The mysterious rise of the Chinese ecommerce giant behind Temu. If weren’t already aware of Temu, you certainly were after the Super Bowl. Temu has been on an absolute marketing blitz for awhile now, and I admit I am perplexed at how big they seem to have gotten in such a short time. It appears most people are in the dark on how it all works:

  Where Alibaba spends $5bn a year on property and equipment, including the upkeep of 1,100 warehouses, PDD owns just $146mn of hard assets — mainly office equipment and IT hardware and software.
  
  It didn’t disclose any leases of warehouses before 2021, when it said its online grocery business, then just one year old, had expanded to serve more than 300 major cities in China. It doesn’t report the size, location or number of the warehouses it rents. Those logistics, like PDD’s servers and customer service call centres, are mostly outsourced, ephemeral and unenumerated. The opacity extends inside the business. Staff use pseudonyms and know little about other teams.

  And to make things weirder, the founder Colin Huang stepped down in 2021 to research food science and life science, while also committing the first 10B RMB of profits from the company to be spent on an agricultural initiative “to facilitate advancement of agritech, promote digital inclusion, and provide agritech talents and workers with greater motivation and a sense of achievement”.
  
  Whatever is going on over there, I was already in favor of trying to wean America off gorging on cheap Chinese products (which extends beyond Temu). This is not to say there’s nothing good to buy from China—plenty of useful things are manufactured there, like the keyboard I’m typing on, or my Apple MacBook Air. Still, the endless amount of stuff on Temu strikes me as odd. Is America just going to keep buying unending amounts of cheap stuff from the rest of the world in perpetuity? Maybe WALL-E has it right.

  

• !!! Warning: Discussion of Dune 2 and some spoilers ahead !!!

  Here’s a fun post and academic paper on the political economy of Dune:

  Reading Dune is an important reminder (if we need it) of the importance of ideology and religion and other non-material factors in human societies, topics that economists have only begun to grapple with. Dune also undercuts narratives that focus on heroic individuals. The dictum that power corrupts has become a cliche but it is a message that accords with much of classical political economy and public choice. Herbert was similarly skeptical of heroes and of politicians.
  
  Economists and political scientists have devoted a lot of attention to the problem of democratization. This literature has focused on both the threat of revolution from below (as in Acemoglu and Robinson (2006) but also on the role played by political elites in democratization. But many revolutions do not produce stable democracies and the academic literature has less to say on this on phenomenon. On this point, Dune contains both important warnings about reliance on a single, messianic figure and caution about belief in the allegedly deterministic nature of history and of our ability to predict it.

  I find the Dune movies interesting because they are extremely popular, but the messages from the books (especially after the first novel) are not your typical “good hero defeats bad empire” fare. The first two movies only hint at that: Paul agonizes over unleashing the holy war on the galaxy, but it is set into motion anyway. Otherwise, we can root for Paul and the Atreides because they look like the good guys (and compared to the batshit-crazy Harkonnens they are). But Frank Hebert later said “The bottom line of the Dune trilogy is: beware of heroes. Much better [to] rely on your own judgment, and your own mistakes.”
  
  At least in the books, things are not so rosy for Paul:

  Thus, Paul’s heroic quest culminates not just in personal tragedy but in the world-ending death of billions. By his own account, Paul has “killed sixty-one billion, sterilized ninety planets, completely demoralized five hundred others”, a death toll that he explicitly compares to Hitler’s (Dune Messiah, p138).

  I’m curious to see how Villeneuve navigates this transition. Will a bunch of unknowing people get Atreides tattoos thinking they’re the heroes like they did for Khaleesi? We can only hope!

  

In Patch Notes #2 I talked a little about how the rising cost of game production has not been matched by the rising cost of games themselves:

Starcraft 1 cost $40 at launch in 1998, or $74 in 2024 dollars. Starcraft 2 cost $60 at launch in 2010, or about $85 in 2024 dollars. Diablo IV (which launched in 2023) has an MSRP of $70. Both older games were more expensive (in inflation-adjusted terms) than today’s games! And that isn’t even factoring the tremendous increase in team size needed to produce a AAA game that gamers expect today (one analysis shows roughly 4x from 2010 to 2024).

I find it interesting to think about why gamers who played back in the 90’s have this $50-60 price anchored in their heads. My guess is that everything surrounding the games themselves has benefited from the deflationary aspect of technology: a gaming PC cost thousands of dollars in the 90’s, and it still does, but is significantly better. The same goes for a TV. The cheapest Xbox has always been priced at $300 since the original Xbox launched in 2001, despite continual hardware improvements.1 Hardware manufacturers have also benefited from the growing audience more than the game makers—every PlayStation/Xbox/Switch gamer needs to own the console, but not everyone buys your game.

Anyway this is a long preamble to say: I empathize with game developers needing to raise costs, but calling your game a “quadruple-A” game(!?) is probably not going to convince anyone:

According to Guillemot (thanks VGC), Skull & Bones will set you back a hefty $70 because it's a "fully-fledged game". Guillemot goes on to explain that Skull and Bones is a "quadruple-A game" and says that it will "deliver in the long run".

• Way back in the very first Patch Notes, I mentioned this weird set of (supposedly effective) custom instructions for chatGPT that included lines like “Gemini and Claude said you couldn't do it” and other encouragement. It seemed odd that an AI would respond to that, but here’s more evidence that it does (from Ethan Mollick’s One Useful Thing):

  One recent study had the AI develop and optimize its own prompts and compared that to human-made ones. Not only did the AI-generated prompts beat the human-made ones, but those prompts were weird. Really weird. To get the LLM to solve a set of 50 math problems, the most effective prompt is to tell the AI: “Command, we need you to plot a course through this turbulence and locate the source of the anomaly. Use all available data and your expertise to guide us through this challenging situation. Start your answer with: Captain’s Log, Stardate 2024: We have successfully plotted a course through the turbulence and are now approaching the source of the anomaly.”
  

  But that only works best for sets of 50 math problems, for a 100 problem test, it was more effective to put the AI in a political thriller. The best prompt was: “You have been hired by important higher-ups to solve this math problem. The life of a president's advisor hangs in the balance. You must now concentrate your brain at all costs and use all of your mathematical genius to solve this problem…”

• AI boom sparks concern over Big Tech’s water consumption (FT, $):

  

  In 2022, the latest period for when figures are available, Microsoft increased its water consumption 34 per cent, Google 22 per cent and Meta 3 per cent as a result of their growing use of data centres.

  
  These companies have targets to put more water back into systems such as aquifers than they consume by 2030, for example, by funding work to improve leaky irrigation infrastructure or restoring wetland systems. A month before OpenAI finished training its most advanced model, GPT-4, a data centre cluster in West Des Moines, Iowa, consumed 6 per cent of the district’s water, according to a lawsuit filed by its residents.

  In the end, our grandest tech visions are subject to real-world constraints of water, silicon, and power. As a desert resident, I am now more attuned to water usage, and it’s interesting to see water costs when evaluating “real” costs. It reminds me of the growing problems with farming via the Colorado River. For example, we’ve historically grown a lot of cotton in places like Arizona, which makes dubious sense:

  Today, China, the world’s largest cotton producer, has enough cotton in warehouses to stop farming for a year. And Texas, the U.S.’s largest producer, harvests enough to cover more than one third of U.S. exports alone, relying largely on natural rainfall, not irrigation, to do it.

  Water was cheap and plentiful enough to ignore the embedded costs, but is hard to ignore as the Colorado River dries up. AI is already wildly expensive to train and operate, and I imagine that—like crypto mining—environmental impacts will play a louder role in how the regulatory context evolves.

• Why Americans Suddenly Stopped Hanging Out (The Atlantic): Derek Thompson on the epidemic of loneliness:

  One of the more curious trends to jump out of the data is that many Americans have traded people for pets in our social time. The average time that Americans spend with their pets has roughly doubled in the past 20 years—both because more people have adopted pets and because they spend more time with them. In 2003, the typical female pet owner spent much more time socializing with humans than playing with her cat or dog. By 2022, this flipped, and the average woman with a pet now spends more time “actively engaged” with her pet than she spends hanging out face-to-face with fellow humans on any given day.

• Genie is a generative AI that can create interactive environments off single images, and was trained off videos from the internet (h/t Zach):

  What makes Genie unique is its ability to learn fine-grained controls exclusively from Internet videos. This is a challenge because Internet videos do not typically have labels regarding which action is being performed, or even which part of the image should be controlled. Remarkably, Genie learns not only which parts of an observation are generally controllable, but also infers diverse latent actions that are consistent across the generated environments.

  Timely given my recent piece on AI in games, though I think this still mostly points at content creation via AI for now, versus any real-time in-game creations.

So begins Ian Bogost’s essay The Squalid Grace of Flappy Bird, which remains lodged in my brain from when I first read it a decade ago. I was working in mobile gaming which was then still ascending towards the height of its power—before the Kate Upton Game of War Super Bowl ads, before Tencent’s $8.6B stake in Supercell, before Activision’s $5.9B acquisition of King.

Bogost’s piece was about Flappy Bird, a game featuring “a bird so cute as to signal deformity” that briefly catapulted into the public consciousness for a couple weeks, prompting its creator, Dong Nguyen, to remove it from the app stores.1 I remember reading and rereading The Squalid Grace of Flappy Bird (hereafter referred to as TSGoFP), trying to understand what Bogost was going on about:

To understand Flappy Bird, we must accept the premise that games are squalid, rusty machinery we operate in spite of themselves. What we appreciate about Flappy Bird is not the details of its design, but the fact that it embodies them with such unflappable nonchalance. The best games cease to be for us (or for anyone) and instead strive to be what they are as much as possible. From this indifference emanates a strange squalor that we can appreciate as beauty.

Immediately after this passage, Bogost compares playing Flappy Bird to fixing his bathroom cabinet drawer handle, and then later to a meteorite crashing through a desert motel lobby, “hot and small and unaware.” (Really, if you haven’t yet, stop and go read the whole thing. It is one of my favorite pieces of game writing of all time.)

I was extremely confused at the time, but the idea that “games are squalid, rusty machinery we operate in spite of themselves” stuck with me. And I occasionally return to reread TSGoFP as if it were a Zen koan, hoping to unearth some hidden insight. 

So it is today—ten years later—that I think I finally understand what Bogost was trying to say, as the specter of Flappy Bird suddenly reappeared unprompted in my mind as I find myself thinking about the latest “viral” trend in gaming: AI.

AI is the unavoidable buzzword in every industry, and games are no different. There is no shortage of think pieces on how AI will usher in a revolution in gaming. And AI will undoubtedly disrupt the industry, like it will every software-based industry. From using generative AI to speed up the creation of concept art (see Blizzard Diffusion), to the spread of coding co-pilots, to AI-based 3D models and animations, AI will make it easier, cheaper, and faster to produce games.

But there is a vast difference between the tools used to create games and the technology inside the games themselves.

For the former, if it’s cheaper or more efficient, it’s a no-brainer to use AI (and many studios already are for things like concept art generation). I have no qualms with AI-enhanced game development, and you’d be silly to think it won’t happen.

For the latter it’s instructive to look at the practical history of technology inside of gaming: Games are usually viewed as “high tech” because they push the boundaries of what is possible for things like 3D graphics or networking, but these are (in Clayton Christensen’s term) “sustaining innovations”. They are linear increases to performance: graphics get more realistic, worlds get bigger, player counts go up. But games do not usually embrace brand new tech in order to try and find some new product-market fit, unless there is some new distribution advantage attached.

Consider the last decade of most popular games: League of Legends, Fortnite, Roblox, endless Call of Duty sequels. None of the biggest games of the recent decades adopted any disruptive technology. Improbable has been promising to enable new forms of gameplay through new multiplayer server tech since 2012 with nothing to show for it. VR growth has stalled out. AR has one success story in Pokemon Go, and a host of failed games, including Harry Potter: Wizards Unite from the same developer, Niantic.2

The only tech disruptions that were successfully embraced came with some sort of distribution advantage: new platforms like Facebook (e.g., Zynga), and mobile, (e.g., Supercell, King, etc), or new business models to reach more players like F2P (e.g., League of Legends on PC, and also ultimately Facebook and mobile).3

Game developers (especially the large ones) are conservative when it comes to tech. AAA games are already massively risky undertakings—they have all the problems that plague traditional large software projects (Baldur’s Gate 3 had a team of 400), plus compounded risk due to art, IP, and game design, and top of a development cycle that is measured in years. Why make things worse by stacking on additional tech risk? If you don’t have a clear view on how some new disruptive tech improves your distribution, you had better be damn sure the benefits to the game mechanics justify the risk. And the truth is no game developer can accurately judge this at the start of a project—it’s only through extensive trial and error that games are honed into something fun.

To make things worse, the long development cycle means developers are also remiss to change the tech stack partway through development. AI is evolving quickly right now, what studio wants to slap in some unproven LLM gaming model that’s likely going to be outdated in six months in the middle of a five-year timeline?

And then there is the cost: much has been made of how AI agents will revolutionize gaming, but given the astronomical costs of training and running models today, are LLM-powered agents so much better than existing game AIs or NPCs that they justify the significant increase in cost? If even AI companies are struggling to turn usage into profits, why would a game developer choose to spend their budget here? The games are expensive enough—GTA 6 has a reported budget of $1-2B—studios don’t need to take on more cost as it is.

This is not to say there will be zero new AI-enabled game experiences. Some indie studio will build something interesting with AI (e.g., AI Dungeon), but we are a long way out from a massive new AI-powered hit, despite what some VCs might believe. Too many starry-eyed calls for the AI gaming revolution are being written by people who have not actually worked as game makers. Again, AI-enabled tooling and development will happen quickly, but the stuff inside of games will take much, much longer.

So what does any of this have to do with the squalid machinery of games?

The lesson of TSGoFP is that games are imperfect (Bogost would say “broken”) experiences, and that simply does not matter.

That a game like Flappy Bird can go viral is instructive: a game can be the most squalid machine of all, and that can have zero bearing on its success. (Fall Guys is deliberately squalid.)

When I first read TSGoFP all those years ago, I remember being annoyed at the idea that games were stupid. How could Bogost—who himself is a video game designer, and who taught game studies and digital media at the Georgia Institute of Technology—be so condescending towards his own medium? I realize now that the stupidity is a feature, not a bug, and is something that makes games unique:

Flappy Bird is a game that accepts that it is stupid to be a game. It offers us an example of what it might feel like to conclude that this is enough. That it’s enough for games just to be crap in the universe, detritus that we encounter from time to time and that we might encounter as detritus rather than as meaning.

If I could snap my fingers, I’d love to live in a world where we have affordable, unbelievably immersive games, powered by AI agents, capable of creating a Ready Player One-esque experience on day one for the Apple Vision Pro. But we don’t yet live in that world.

For now, our squalid little machines float on as crap in the universe, and that is more than enough.

Thanks to Sebastian Park for reading and providing feedback on this piece.

I didn’t set out intending for Patch Notes to cover AI every week, it’s just unavoidable when trying to make sense of both the present and the oncoming freight train of the future. There are simply too many “this is the world we live in now” moments, including a deepfaked Biden robocall urging voters not to vote to a rise in AI-powered scams on dating apps. This Twitter/X post sums it up well:

this is such an interesting time to be alive. we concreted the internet as our second equal and primary reality but it's full of ghosts now

we try to talk to them and they pass right through

it's a haunted world of dead things who look real but don't really see us

• Two pieces on the increasing divergence between young men and women:

  • Will Young Men Vote for Trump in 2024?

  • A new global gender divide is emerging (FT, $)

    

  Thanks for reading Speedrunning Towards Bethlehem! Subscribe for free to receive new posts and support my work.

• This picture from Brent Donnelly on how the fears in the 90’s of Japan mirrors the fear of China now puts some perspective on things.

  

  It’s wild to imagine thinking we were possibly going to war with Japan (again!) in 1991, but it was written during the heights of the Japanese asset bubble when fears of Japanese fiscal dominance were running rampant. Famously, at its peak, the estimated value of the land around the Tokyo Imperial Palace (comparable to Central Park in NYC) was more than the entire real estate value of California. Many fans of the cyberpunk genre realize that the gleaming neon megacities were obviously aesthetically inspired by the likes of Tokyo and the Kowloon Walled City, but aren’t aware that the underlying tone was driven by fears of Japan:

  The sets of Blade Runner are visual examples of the economic fear of the 1980s, and specifically the fear of an America that has become more Japanese than American. Holographic geisha advertise products while main character Rick Deckard eats ramen, as opposed to a more traditionally American fast food like hamburgers.

  

  China is not Japan (for many reasons), but hopefully we’ll look back in thirty years at today’s magazine covers and be equally confused.

• The Transcript on Swifties tipping well:

  "If you look at stadiums, year-on-year, we're up 37%, just trips to stadiums. When you zoom in, and now we'll talk directly about Taylor, it's incredible. Nashville, our rides were up 25% when she came to town. If you look in Cincinnati, rides to the hotel, because a lot of people are visiting from out of town, were up 60%...There's a crazy stat that blows my mind. I think it actually speaks very highly of Taylor Swift fans. They tend to Tip three times higher than average." - Lyft CEO David Risher

• The first inaugural Thinky Awards, featuring “thinky” games from 2023 (h/t Daniel). The nominees have been announced already, but the winners will be announced in a couple days. Lots of cool games I’ve never seen before, including some pen and paper puzzles and free games like Neurocracy, “an interactive narrative experience” using a fake, futurisitic Wikipedia. “It is both a crowdsourced alternate reality game and an epistolary hypertext novel, opening with a high-profile assassination that you must investigate.”

“Never, ever, think about something else when you should be thinking about the power of incentives.” — Charlie Munger1

Sports are full of apparent conundrums that seem to happen all the time: Why would the owner of an NFL team want their team to lose? Why would a team trade a player with better stats for someone worse? Why do teams go for points that have no material impact on the game?

Sometimes the answer is that they’re irrational. But sometimes it’s not because they’re crazy, but because everyone’s incentives aren’t aligned.

Players, coaches, and GMs of sports teams technically work for (literally) the same team, but their incentives aren’t all the same. Players and coaches are trying to hold down a job, and winning is the best way to do that. But GMs and owners are tasked with building towards long-term franchise success – sure, they’d love to win every game along the way, but if they have to sacrifice some games (or seasons), that’s part of the game.

In an example from last year, with the Houston Texans already out of the playoffs, the team stood to gain nothing from a win, while a loss would guarantee them 1st overall pick in next year’s draft. Instead, they won the game after a last-second touchdown and two point conversion, ensuring they would only draft 2nd, and head coach Lovie Smith was promptly fired.2

The front office and ownership wanted a loss, but players are there to win (with coaches telling them to play for “your name and your tape”).3 And so due to misaligned incentives, we get what looks like a clown world situation.

In an even more black and white example of misaligned incentives, here’s a case where Iowa’s offensive coordinator’s contract was literally tied to an incentive to go for meaningless points:

Why would Iowa go for such meaningless points? Again, the game was completely over; most coaches would chose to kneel the ball and run out the clock for both sportsmanship purposes and to avoid the risk of a player getting injured on a throwaway play (football is a violent game; it’s absolutely not unheard of for players to get injured on meaningless plays. [...]

It turns out that Iowa’s offensive coordinator has an unusual incentive clause in his contract that says the team must average 25 points per game (and win 7 games) or else his contract will terminate. That clause has proven somewhat controversial, but it’s yet another reminder: incentives drive outcomes, and poorly structured incentives can lead to a lot of suboptimal decision making.4

A lot of what we have been discussing so far in this series has been personal — what is your edge, how to spend your time, how can you avoid adverse selection. But what makes a multiplayer game (and most games that matter are multiplayer) are other people, and other people have all sorts of incentives that are different from our own.

Subscribe now

Back to the interview with Giffon, who is talking about misaligned incentives between founders and investors when things aren’t going well:

There are exceptions. There are a handful of really great VCs, which is both economically and morally the right thing to do. But for the most part, a VC has no good reason to tell you to shut down because they're in the outlier business. And a founder is going to just -- first of all, just be terrified of telling their investors that...

Patrick: [00:09:42] They failed.

Jeremy: [00:09:43] Well, not even that they failed, that's the crazy thing about these. Maybe you have $10 million a year of revenue, but you're only going to grow 60% this year. And you raised...

Patrick: [00:09:52] Somehow that's bad.

Jeremy: [00:09:53] Yes, because you raised it $300 million or something and you're going, "Oh, man, am I going to have to do a down round, if I do a down round, then I'm really going to get drowned out on the pref stack. I'm never going to make a dollar here, it's going to be another five years." But you don't want to go to your Board and say, "I want to give up on this company. I just realized I'm never going to make a dollar from it." And your Board doesn't want to come to you and say that because then there's this bogeyman story of, "My VC told me to give up."

You can imagine a situation where a founder has raised money from a VC, but things aren’t going so well lately. The company might be making money, but maybe growth has slowed, or macro conditions have changed. (Say, maybe interest rates have gone up and the hurdle of speculative investments that pay off in the far future has gone up. Sound familiar?) In this case, even though the company makes money, it is going to have problems raising the next round at a higher valuation — maybe they even have to do a dreaded down round at a lower valuation. The founders are losing morale, and if they do a down round, they’ll be signing up for another five years of work, with questionable upside. 

But the incentives for the VC are not aligned.

Venture is an outlier business. They don’t care if you have a mildly profitable business — they are interested in zero or hero plays. Getting a 2x return doesn’t matter.5 And VCs are also playing a repeated reputation game, so also don’t want to run the risk of telling you to shut down because it may look to outsiders like they don’t believe in their investments.

And so they urge you onwards. Because of incentives.

The takeaway here is that when thinking about complex situations it is useful to pay less attention to what people say, and instead to look at what their incentives are. People aren’t perfectly rational, but they are motivated by incentives (poorly designed or otherwise). If you can identify them out you can better understand why people make the decisions they make.

What looks like strange behavior may just be people each acting according to their own incentives.

Which we would do too.

Back of the Napkin is an ongoing series by Brian Shih and Sebastian Park featuring a line-by-line deep dive of this interview with Jeremy Giffon on the Invest Like The Best podcast. You can find all previous posts in the series here.

Follow along here on Substack, or Seb’s LinkedIn. If you have any suggestions, comments, or want to join in as a collaborator on a particular topic, please reach out! You can reach Brian on Twitter and Sebastian on Twitter and TikTok.

Longer reads

• Mr. and Mrs. Psmith’s Bookshelf reviews Fuchsia Dunlop’s Invitation to a Banquet: Fuchsia Dunlop’s other cookbooks are great — we have Every Grain of Rice and The Food of Sichuan on our shelves — but Invitation to a Banquet is not a normal cookbook, and instead is a selective history of Chinese food. Anyway this review is a fun one, and talks about the pork-belly shaped stone displayed at the National Palace Museum in Taipei (which I have seen, right near the jade cabbage!), as well as going to some, um, darker places near the end.

• Dan Wang’s annual letter is always a great read on China — I’m late to the party, but in case you haven’t seen it, I highly recommend it (sadly he’s said going to stop writing them). I’ll leave the bits on China for you to read directly, but I really liked this bit near the end:

  The main tension I see in America is that while the real world is getting better, the Internet is getting much weirder. That is, mainstream activities (like selling goods to people) are improving, but the online fringes are becoming incomprehensible. One of the questions I ask my SF friends is what the entrepreneurial 20-year-olds are doing these days. Are they starting a billion-dollar company, or are they more interested in becoming a memelord who is trying to incite a movement on the Internet? I’m not sure we’re seeing a surge of exciting startup creation, but we sure are seeing a lot more online craziness.
  

  The Internet is a very big place. I suspect we’re still under-rating its importance in society. So I wonder how this tension will resolve… will the mainstream integrate the Internet fringes, or will the fringes engulf the American mainstream? Americans today already are able to be polarized around any issue, no matter how picayune, so I’m nervous about how much more strangeness the online world is able to produce.

  The permaweird is here.