This year, I’m going to try to organize this Part 2 reflection by a few free-form chapters of what I hope to be interesting lessons and ideas from 2017. Hope you enjoy!
A measured life
I’ve spent quite a significant amount of time on personal accounting this year. I’d be doing financial accounting anyway, as I’m sure most of us do, so it wasn’t too much of a stretch to add a few more items in my spreadsheet. Why log so much? For me, it’s mainly about goal-setting and accountability. I can have subjective goals each year like spending less, or sleeping more, or reducing my carbon footprint, but I can only hold myself accountable and track my progress if I measure my progress (i.e. you can’t improve what you can’t measure). Seeing the true results also reveals surprises that help calibrate my own biases, and that I hope are informative for readers.
First, my financial accounting is quite complicated as I need to separate out personal expenses and income from those from my pass-through entity and new nonprofit. My Stanford work complicates things further as I am regularly making huge expenses for SUS and getting reimbursed a few weeks later, which screws up any weekly tracking I’d want to do. But since I categorize all my expenses into a few high-level categories, I can look at my breakdown over the whole year as such:
A few caveats: Gifts include charity and gifts to friends and family. Expenses for eating out and entertainment are over-weighted in the sense that I’m counting plenty of times I’m paying for others and not proportioning out the expenses over to the Gifts category. Transportation looks particularly good because I don’t own a car, and I get a GoPass from Stanford for free Caltrain rides (so essentially this is Bart and my monthly Muni pass). Supplies is a little bit of a miscellaneous category, including things like electronics, laundry, and Google Drive. Health insurance will show up as a new category next year when I turn 26. And lastly, of course this doesn’t account for the gifts I receive.
So what does this all mean? I definitely am always hoping to improve the balance between eating out and cooking, though that was difficult throughout the year because of my busy hours down in South Bay, which mean I’m regularly not home until 8 or 9 or 10pm. I mentioned in Part 1 my big change in habit around purchasing music, which shows up in the Entertainment category, and hovering around 10% of my overall expenses, my entertainment spending seems pretty reasonable. I’d like to separate out personal gifts and charity next year, and see both of those categories get to 5%.
I’ve tried my best to accurately count my sleep hours and work hours to monitor my productivity and division of labor. Sleep was pretty easy because I’ve worn a Fitbit all year, and I can just quickly sync to my phone and look back at the daily log. Work hours are a little bit less rigorous, but I’ve tried to apportion out rough time spent working on Stanford business vs. Cloud Arch Studio business vs. City Systems business, usually in 1/2 hour increments, and subtract out non-productive times eating or commuting. I also have hours from Nueva where I was still teaching in Spring of 2017, as well as time reading and blogging here. It’s also worth noting up front that the numbers below are an absolute average over all days and weeks of the year, so they include the weight of weekends and holidays. That’s just a disclaimer if it looks like I’m barely doing a typical 40-hour workweek.
First off, I’m personally quite disappointed by the true balance between Stanford work and my personal ventures, which I ultimately would like to see get closer to 50-50 within five years, as I grow the SUS initiative to be more sustainable without my management (or like I like to call it, firefighting). Reading and blogging can also clearly increase, though it became crystal clear by the midpoint of the year just how outlandish my New Year’s resolution was to embark on a personal writing project; it just became impossible to keep up with the weekly and daily onslaught of crazy news and ideas. I would like to more explicitly track both media reading (short form and long form) and book reading, as well as podcasts, to get a better view of the Reading category as intended. With that change in accounting, I’d like to see Reading get to 1 hour per day, and Writing 0.5 per day, in 2018. Lastly, sleep doesn’t look too bad; I’ll probably keep my goal at 7 hours per night.
I’ll see if I can get my overall work productivity from 42.9 to 45 hours per week in 2018 as well, or more. More fundamentally, one of my greatest frustrations with my current work life, and something these numbers don’t really capture, is that I have so few opportunities to just work on a task by myself, for two or more hours, without having to manage something else. When I think about a genie’s wish, I immediately think about having an 8th day of the week, where the world is held in suspension, and I can just devote a few 12 hours to personal projects. Imagine if we could schedule our lives like that, with weekly or at least monthly sabbaticals.
By the way, since I won’t really mention it anywhere else, if you want to know what I’ve been doing professionally this year, it’s pretty much all here. And the best way to follow the work I’m doing is by subscribing to this, and following this blog for some big personal nonprofit updates in 2018.
This next section will focus on a subset of my larger interests in carbon reduction, so I’ll discuss that first. I’ve been on a general journey to reduce my carbon footprint, which, according to carbotax.org, is about 10 metric tons of carbon equivalent (MTCO2e) per American. Here’s my breakdown after answering the questions on that survey:
|Annual MTCO2e||Avg American||Derek|
|Waste & Water||0.4||0.1|
|Food & Beverages||2.4||1.2|
|Travel (Air, Hotel, Boat)||0.7||15.5|
If you’ve looked at this and other carbon calculators, you’ll know that some of the outputs are influenced by the carbon content of local utilities that you don’t have direct control over, while others are more directly tied to personal actions. There are also plenty of model-based estimates taking place in here which I could spend more time refining with my own calculations, but I just haven’t found the time to undertake yet.
Interpreting the numbers as they are, obviously, the killer for me is the air travel I do for business and for family/leisure (almost 65k miles this year). Assuming that in the broad utilitarian sense, I’m mostly flying to work on global sustainability issues, and that my work will reap many multiples’ worth of benefit, and at the very least I’m donating to try to offset my carbon, then where I personally focus on is reduction in daily and weekly material consumption. My transportation is pretty good on account of mostly using public transit, and so this year I was particularly interested in taking as much meat as I could out of my diet.
Food is particularly interesting because, for me, there’s also a significant mitigation-of-suffering goal in effect (more on ethics later).
So how did I measure my diet? I split each meal into the categories of vegetarian, seafood, chicken, pork, and beef/lamb (in order of carbon content, and in my opinion, ethical standing). If a meal has a mix of two types of meats, I accounted for the more carbon-intensive one. I effectively only logged the meat meals each day, so the remaining, including most breakfasts, were vegetarian (totaling 21 meals per week). Unfortunately I don’t have a good baseline to compare to since I haven’t tracked this before, but I can pretty safely say that prior to trying to be vegetarian, I was raised and lived with the understanding that literally every meal should have meat, so my daily meat intake was at least 2 meals. Here’s how I fared in 2017:
I didn’t quite make it to full vegetarian, but what I achieved is probably the biggest lifestyle change I’ve ever had. (What it effectively amounted to was a lot of salads and tofu, which I do in fact really love.) However, I couldn’t quite shake the temptation of cravings for Chik-Fil-A and In-N-Out, so for most weeks I was effectively a weekday vegetarian.
So if my daily meat intake dropped from about 2 meals to 0.9, what was the outcome? If I hold onto the same proportion of meat types, this would roughly be equivalent to the following in combination (assuming 1/4lbs as a meat serving):
- 180 servings of chicken avoided, which, assuming 2 lbs yield of meat from one factory chicken, would be the saving of about 20 chickens.
- 90 servings of pork avoided, which based on this would be around 15% of a pig’s yield, or the saving of 3 out of 20 pigs.
- 90 servings of beef avoided, which based on this would be around 5% of a cow’s yield, or the saving of 1 out of 20 cows.
- 45 servings of fish avoided, which based on this would very roughly be on the order of saving 10 sole-sized fish.
And in carbon, based on this, that’s about 1 MTCO2e offset. That seems about right compared to the average American statistics. (In the next section I’ll dive down the rabbit hole of the full ethical implications of eating animals.)
For next year, I certainly would like to see my vegetarian meal count approach 3 per day, but assuming I’ll still eat some amount of meat, I’d like to see seafood be the greatest proportion. About halfway through the year I made a measly attempt to go vegan; next year I’d like to begin tracking eggs and dairy from the start so I can see my progress towards veganism as well.
I’d love to hear what you track on a daily basis, what your goals are, and whether you have any recommendations for apps that make it easier to track goals!
The ethics of eating animals
I’ve attempted to articulate my ethical journey through a few blog posts this year, but unfortunately I haven’t been able to do my thoughts justice. In advance of hopefully much deeper investigation next year, as my ideas continue to develop, I’ll just use the example of vegetarianism to illustrate exactly how my mind tries to grapple with the ethical calculus, which may be a better representation of my ethics than me trying to extrapolate some higher-level themes.
How do you compare environmental and mitigation-of-animal-suffering goals in your diet, if you care about both? If you have to eat meat, what’s the right balance of the four key types of meat (chicken, pork, beef, fish) to maximize both ethical values? (I can’t remember in exactly which podcast, but Ezra Klein makes his own case for eating beef over chicken, which I intuitively disagreed with and which got me really thinking about this.) There are a lot of variables at play. First I have to estimate the typical weight and yield of a factory animal, which lets me estimate the number of servings the animal yields, as well as its life cycle carbon footprint. To summarize using the same few references I used above:
|Full weight/animal, lb||Edible weight/animal, lb||Servings/animal||kgCO2e/kg animal||kgCO2e/serving||Animals/serving|
If you accept the assumptions above (without worrying about sensitivity analysis for now), now we need to figure out an equivalency between kgCO2e/serving and animals/serving. Basically, we need a common denominator, say, our valuation of a human life. Of course, this is where things get super subjective, and it’s probably best to calibrate our heuristics on orders of magnitude (I’ll also note here that I’m not considering dairy/eggs for now).
To figure out the environmental side of the ethical equation, the basic question is: how many kgCO2e does it take to cause one human death (or whatever unit of human suffering we want to consider)? Well, with a very brief amount of Googling, I found two key numbers I’m willing to work with for now. This estimates our global carbon budget at about 1 trillion MTCO2 if we want to keep temperature rise below 2 degrees C. This (specifically Table 20.16 on page 64) attributes about a hundred thousand deaths to climate change (I couldn’t figure out exactly which climate scenario this used, and didn’t bother to examine the details of the methdology, but I’m just taking the order of magnitude here as a reasonable estimate of DALYs-worth-of-deaths exclusively attributable to a 2 degree C temperature increase). If you accept those two vast assumptions, then we’re talking something on the order of tens of billions of kgCO2e equating to a human life. If we normalize the kgCO2e/serving of each animal by its contribution to a human death, and then add a bunch of 0’s (1e11) to get a good-looking integer and call that a util, then we get the following utils for the environmental impact of 1 serving of each animal:
Next, to figure out the animal suffering side of the equation, the basic question is: if I had to choose between the death of 1 human or the death of X of each of these animals, what is X? First off, some people would say that no number of any non-human animal life is equivalent to a human life, but let’s just assume for argument’s sake that you accept Peter Singer’s claim that animal suffering is at least measurable in the same currency as human suffering. Then let’s just say for argument’s sake that we think 1 human life is worth 1 million cow lives (i.e. the trolley problem, but the first track has 1 million cows on it). Then using our util, we end up with 1 serving of beef equating to 38 utils. Notice that this is on the same order as the 31 utils of impact from 1 serving of beef due to its carbon emissions. I ended up picking 1 million to reach equivalency between the two sides of the equation, to basically demonstrate that, taking all other assumptions for granted, if I think that cows are worth less than a millionth of a human life, then I should forgo beef mostly because I care about the environment. But if I think that cows are worth more than that (let’s say a thousandth of a human life), then animal suffering becomes by far the greatest weight to my ethical calculus.
Now in terms of the difference between the animals, if we believe the following are reasonable claims:
- 1 human life is worth 1,000,000 cow lives
- 1 cow life is worth 3 pig lives
- 1 pig life is worth 65 chicken lives
- 1 chicken life is worth 2 fish lives
Then we get this final comparison of subtotal and total utils for a serving of each animal:
|kgCO2e/serving||Animals/serving||Environmental utils/serving||Suffering utils/serving||Total utils/serving|
Again, to demonstrate the balance point, I’ve picked equivalencies between each of the animals to roughly balance out the impact of eating any of the animals (65-70). And, since these final assumptions were fundamentally subjective variables, here’s a spreadsheet you can download to play with the numbers. But basically, here’s what I take this all to mean:
- The impact of eating animals is small but ethically meaningful, both because of the direct impacts of animal suffering as well as the indirect impacts of greenhouse gas emissions on climate change and suffering (for humans, and I guess animals too).
- I personally don’t know how many cows need to be on the first track for me to pull the switch to kill one human stranger instead, but it’s definitely a number with a lot of 0’s behind it. Maybe it’s 10,000, maybe it’s 1,000, maybe it’s a googleplex, I really don’t know without understanding human and animal suffering more. But I’m willing to believe that the ethical weight of climate change and animal suffering are within a few orders of magnitude of each other, and so I care deeply about both.
- Between the animals, I would agree with cows and pigs being within the same order of magnitude, and birds being lower, with brain size being my only meaningful indicator of suffering. In fact it would take a lot more chickens, hundreds, for me to pull the lever to kill one mammal instead, and hundreds of fish to kill one chicken, which ultimately means that poultry and seafood are orders of magnitude better than eating mammals, by all my ethical accounts.
OK, let’s extrapolate some higher-level themes
The simplest insight I’ve had on ethics this year is that it can be measured, just like I’ve measured so many things in my life. This becomes crystal clear if you think about values as valuations. The hard part, of course, is what you’re valuing and how you’re valuing it, and the harder part is, how do our real-world actions affect outcomes in value terms, and the hardest part is, how do we agree on our metrics of valuation.
The most important tool we have in ethics is reason. Reason is quite literally having good reasons for why you do what you do. And if values rely on rational accounting of some kind of currency, then if we agree on our underlying assumptions, reason enables us to agree on our ethical choices.
The most important spectrum of underlying assumptions is between principle-based, or deontological, and outcome-based, or utilitarian, methods of valuation. Without getting too much into detail, utilitarianism is the true domain of empiricism, or measurements in the real world. In other words, utilitarianism is the true home of reason.
What I value is the suffering of conscious beings. Even though our observable measurements of suffering are imperfect, I can meaningfully compare outcomes based agreed-upon assumptions about suffering (like I demonstrated in the previous section). Most directly this line of reasoning leads to the classic utilitarian goal of maximizing universal well-being in our personal or policy-scale decisions.
There’s an important meta-value that I’ve grown to appreciate in 2017 with another simple thought experiment called the veil of ignorance. Basically, since we were born into this world without a choice of exactly who we were, we should make ethical decisions as if we had no information about who we were. It’s like designing a perfect world under the condition that after we designed that perfect world, we were then placed into it at random. What this Rawlsian thought experiment gets at is the importance of fairness, or equity, alongside maximization of well-being. In other words, the difference between a Darwinian self-interest and a post-evolutionary or progressive selflessness, or Singer’s ‘expanding circle’.
I’ll emphasize what I meant by that last statement. I believe that the most fundamental type of political difference tracks the development of human ethics from selfishness to selflessness, and that conservative ideals (capitalism, libertarianism, western religion, homeownership) are our most natural of human values because they are the product of our Darwinian genes, while the future of humanity lies in the discovery of post-evolutionary truths about the meaning of suffering, and that those who adopt liberal ideals (socialism, equality, globalism, and reason) are quite literally martyrs of the future.
The link between my personal ethics and the ethical city (urban systems being my professional area of focus) is as simple as evidence-based decision-making. I believe if we can come up with fairly simple ways to encode the parallel goals of maximization and equity in well-being, and we use the scientific method to continually find better ways maximize well-being, minimize suffering, and reduce inequalities, we can guide both individuals and societies towards that progressive future.
Here’s a practical way to think about the ethical city. If our city is a room full of people with a floor and a ceiling that represent the worst and best of outcomes, then as designers, engineers, and policymakers, I think we have two basic jobs:
- Raise the floor.
- Build as many ladders as we can.
And now, for some lighthearted gaming
If you’re still with me, I’ll let you in on the best game I discovered in 2017. Ready for it?
Set. On Google Play. Seriously. Download it here, and add me as a friend.
I played Set maybe once or twice in high school, but thanks to Paul, I’ve rediscovered it in digital form and it is intellectual paradise. I literally feel like I could write a book about how my mind works on and is worked by this game. Some brief observations on the three hundred times I’ve played this game:
- I’m not convinced there’s a stable equilibrium of strategy for this game. If I try to develop a systematic process of elimination to find the set, then my brain begins to over-rely on probabilistic rankings and wastes time on rare combinations. Then if I switch over to a broader, more intuitive view of the board, I’m a little bit slower on average per set. Maybe the perfect strategy is out there, but I’ve experienced the game more as a rotating set of gym workouts that has exercised multiple parts of my brain.
- That being said, the intuitive muscle in my brain has really surprised me at times when playing this game. I’m beginning to suspect that my brain knows, with an immediate subconscious register of colors and shapes, what the pattern combination is likely to be, and then it’s up to my frontal cortex to stagger towards the correct identification of specific cards. It’s an incredible feeling.
- Also incredible is a kind of meditative experience I have when I’m really in a state of flow in this game, and I can tap into a meta-level of thinking and basically observe myself in thought. This literally is the closest I’ve felt to the meditative experiences that Sam Harris talks about on his podcast.
- By the way, one huge perk: I can play this game while listening to podcasts.
- Also by the way, this makes for a very fun 2-player game, and even 3-player game, but technically that is distorting your Google Play statistics…
- At this point, what I’m essentially trying to do is change the shape of my distribution of games from normal to lognormal. It was a month-long endeavor to get my three-minute bar to meet my four-minute bar, and now they are neck and neck. I wonder if someday my two-minute bar will become the mode…
OK, enough geeking out about Set. This year, thanks to Wayland, I also got really into some party games which I can also geek out about like I did above, but will spare you the embarrassment:
- Codenames: probably the ultimate party game for both old and new friends. Turns out it even works well for English vs. non-native English speakers (as I discovered in Monterrey, MX).
- No Thanks: close to Set in its intellectual wonder, but more from an econometrics angle. Playing with the minimum three people is a perfect never-ending oscillation of game theory. Also, I tried playing this in Thailand in a cafe and the staff told us to stop because it looked like we were gambling, and turns out, gambling is illegal in Thailand.
- One Night Ultimate Werewolf: a marked improvement on Mafia and quite fun if your group is willing to commit at least an hour to it, to play a satisfying number of rounds.
- Skull: maybe the essential game of bullshit and chance. But as a result, it has a fairly short half-life on account of how mentally stressful it is.
And finally, as a confession, in the midst of an incredibly busy Fall, I did manage to make the time to buy a Nintendo Switch and beat both Breath of the Wild and Odyssey. My, has video gaming improved since the days of Pokemon Yellow and N64 Smash.
Hope you get to try some of these games, and reflect on the measurable and immeasurable in life, this holiday season!