The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • EA - Selfish Reasons to Move to DC by Anonymous EA
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Selfish Reasons to Move to DC, published by Anonymous EA on September 5, 2022 on The Effective Altruism Forum.
    You’ve probably heard the arguments for working in DC policy. Basically, the US government has lots of power and money, and it’s possible to get into influential roles surprisingly quickly. But many in the EA community have a negative view of the quality of life of a DC policymaker: the hours are long; the policy wins are big but rare; the pay is terrible; the bureaucracy is even worse; and you can’t wear jeans to work. Doing policy work in DC is sometimes framed as a noble sacrifice that EAs make in order to have a higher impact.
    In the run-up to EAG DC, I want to offer some reasons that DC is not only a great place from an impact perspective, but also a really nice place to live (for a certain sort of person). I think these selfish factors matter quite a bit for whether a career in the federal government is going to be sustainable for you over the long term. Note: I’m deliberately including some items on this list that will sound terrible to some readers, as long as they’re perks from my perspective.
    The EA community in DC rocks. This honestly bears like 90% of the weight for me (and probably most EAs in DC). I’ve found the DC EA community to be by far the most warm and welcoming of any EA community I’ve encountered. One possible reason is that EAs who decide to come to DC are self-selected for being more extroverted than the median EA. There’s also a strong culture of networking and making introductions in DC, which makes it easier to get integrated quickly.
    DC is beautiful. DC has tons of parks (the most of any major US city?), lots of beautiful architecture and lots of interesting trails to explore. It’s a particularly great place for runners and bikers – you can feel like you’re out in nature pretty quickly on a regular basis without needing to actually leave the city.
    The dating market is good. I haven’t been on the market while living in DC, but friends tell me there’s an abundance of young, fit, single, highly-educated, socially skilled, do-gooding people in the area.
    Non-EAs living in DC tend to be pretty impact-oriented and ambitious. It’s not uncommon for my non-EA peers in DC to be really excited about the work they’re doing and enjoy talking about their ambitions for impacting the world.
    It’s not huge. Most places and people I want to visit in DC are at most 30 minutes away. In fact, much of the EA community in DC is very concentrated in just a couple neighborhoods. If you live in one of those places, you’ll live within walking distance of lots of cool people. And because DC is beautiful (see point 2), the walks are pleasant.
    The food scene is great, especially for vegns. I honestly don’t care much about food, but I hear that DC’s restaurants are actually a huge draw for my foodie friends, particularly vegans and vegetarians. I’ve definitely appreciated the abundance of really high-quality and relatively affordable “fast casual” vegan-friendly spots around DC.
    Free museums! The Smithsonian museums are awesome, and it’s pretty great to be able to walk in any day of the week without paying anything. Lots of free activities make it easier to live on the modest salaries of early policy jobs.
    DC is cool. Alright, maybe only according to me. But I think between the grand architecture, the security clearances, proximity to the “halls of power,” and wearing suits all the time, my life in DC is just a lot cooler than it was in other cities. Movies like “All the President’s Men” and “The Report” convey some of the coolness of DC.
    You don’t have to work in government. This isn’t exactly a benefit — more of a PSA: it’s not crazy to move to DC, even if you don’t see yourself having a long career in government. It could still make sense to come to DC and wor...
    5 min
  • EA - An entire category of risks are undervalued by EA [Summary of previous forum post] by Richard Ren
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: An entire category of risks are undervalued by EA [Summary of previous forum post], published by Richard Ren on September 5, 2022 on The Effective Altruism Forum.
    Word count above is inaccurate; 5 minute read (1353 words) excluding footnotes.
    This post is a brief summary of a longer forum post I wrote on systemic cascading risks & their relevance to the long-term future (an EA criticism competition submission).
    I make three strong claims:
    (1) The cascading sociopolitical & economic effects of climate change, pandemics, and conflicts are undervalued in the mainstream longtermist community. These systemic cascading risks can be extremely important to the long-term future through shaping the development of powerful technologies in the next 10-30 years.
    (2) Institutional resilience is the generalization of the solution to systemic cascading risks. A resilient food, water, energy, and infrastructure nexus are key to ensuring system stability around the necessities to live during a crisis, helping to tractably hedge against all systemic cascading risks at once.
    (3) The systemic cascading lens fills the missing gap between current events & longtermism, solving a key question in EA epistemics.
    0. What is a systemic cascading risk?
    Envisioning our societal structures as a graph with nodes and links, a given systemic cascading risk shocks a subset of societal "nodes" and causes n-th order effects that cascade across systems & magnify in volatile, harmful ways – exploiting underlying systemic flaws and interdependencies.
    I subsequently discuss COVID, the Russia-Ukraine war, and climate change as examples of pandemic, conflict, and environmental systemic cascading risks respectively.
    Pandemic and Conflict Risk: COVID & Russia-Ukraine War
    COVID-19, alongside the Russian invasion of Ukraine, served as a catalyst for food inflation and undermined perceived institutional legitimacy:
    COVID's economic stresses, for example, catalyzed violent unrest and riots internationally. Fiscal stimulus-based recovery drove global debt levels to 320% of global GDP, putting many countries at risk of default.
    The Russia-Ukraine War disrupted a third of global wheat & barley exports, 75% of sunflower oil exports, and drove up gas and fertilizer prices, resulting in global maize and wheat price spikes of 42% and 60% (from January 2021 to June 2022). International export bans and a violent Sri Lankan debt default have worsened the crisis.
    High food inflation levels are expected to last until 2024, further testing fragile states reliant on food imports and triggering social distress. Historically, the absence of adequate food, water, and energy drives political instability – e.g. 1977 & 1984 Egyptian and Moroccan bread riots, 1989 Jordanese protests, 2011 Arab Spring.
    Environmental Risk: Climate Change
    In the next 30 years, anthropogenic climate change is projected to cause ~216 million internal climate migrants due to heat stress/desertification/land loss, ~150 million displaced by sea level rise, ~5 billion people living in moderately water-stressed areas, risks for multi-breadbasket failure and disruption to food supply chains, and related inflationary and poverty effects.
    These stresses on our societal systems are likely to contribute to
    a) violent civil conflict & social tensions – an estimated 46 countries containing ~2.7 billion people will face a high risk of violent conflict due to climate change
    b) political extremism & violation of international norms – e.g. possible far-right and extreme governments being elected in the first world due to climate-based economic and social pressures
    1. How are they relevant to the long-term future?
    Much of EA focuses on tail-end risks – like whether a conflict could cause nuclear war or climate change as a direct existential risk, rather than whether an event...
    13 min
  • EA - Do AI companies make their safety researchers sign a non-disparagement clause? by ofer
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Do AI companies make their safety researchers sign a non-disparagement clause?, published by ofer on September 5, 2022 on The Effective Altruism Forum.
    Among AI companies that employ AI alignment/policy researchers (e.g. DeepMind, OpenAI, Anthropic, Conjecture), which companies make such researchers sign a non-disparagement clause?
    Also, what are the details of such non-disparagement clauses? (Do they aim to restrict such researchers indefinitely, even after they leave the company?)
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    1 min
  • LW - Bay Solstice 2022 Call For Volunteers by Scott Alexander
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Bay Solstice 2022 Call For Volunteers, published by Scott Alexander on September 4, 2022 on LessWrong.
    It's September, which means going to beaches, holding barbecues, and . . . preparing for this year's Bay Area Winter Solstice. This year's team will include me, Shauna, Nova, and Mingyuan. In the meantime, we're looking for volunteers.
    First, we have logistics volunteers. Some of these, especially AV and lighting, will require a commitment to be kind of agenty in figuring out what you need to do, attending dress rehearsals to help test things, etc. The simple tasks will just require you to show up day of. If you want to help with any of the following logistics tasks, please email [email protected]:
    Someone to monitor the Twitch stream/simulcast and give updates to AV on that.
    Someone to work closely with Nova on the video side of AV; filming the event and running the live stream (possibly that's worth a separate role, they'd be responsible for the audio submix to the stream)
    Someone to help setup and run lighting
    5-10 people who can help with simple tasks on the day of, like ushering, taking tickets, and moving large objects
    Second, we have creative tasks. We seem pretty good for people willing to do creative tasks this year, so you don't need to volunteer unless you're really enthusiastic. If you do volunteer, expect to have to audition. If you're interested in any of the following creative tasks, please email [email protected]:
    Giving a speech (either writing your own, or giving a reading selected by the team)
    Singing a song (either your own, or performing one selected by the team)
    We'll give volunteers more information by email, and we'll try to give everyone more information as we get closer to the event.
    Thanks!
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • EA - The Base Rate of Longtermism Is Bad by ColdButtonIssues
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: The Base Rate of Longtermism Is Bad, published by ColdButtonIssues on September 5, 2022 on The Effective Altruism Forum.
    Cross-posted from Cold Button Issues.
    Sometimes philosophers make bold, sweeping claims for other philosophers and modest, palatable claims for the general public. Consider Peter Singer’s writing on philosophy which includes endorsing situational infanticide versus his more popular writing where he makes hard to dispute claims like “[l]iving a minimally acceptable ethical life involves using a substantial part of our spare resources to make the world a better place.”
    Will MacAskill and Hillary Greaves wrote a paper arguing for strong longtermism the “the view that impact on the far future is the most important feature of our actions today.” Then Will MacAskill wrote a New York Times best-selling book that argued that caring about the future is somewhat morally important.
    MacAskill didn’t need to water down his claims to convince me. In theory, I’m fully on board with longtermism. There’s probably tons of future people who matter just as much as we do so let’s prioritize them, hurray! Despite being willing to endorse the philosophy of longtermism, I think building a movement around longtermism or taking actions for the sake of the longterm future are likely to backfire.
    Some friends of mine in the effective altruism movement have said they would be excited about the shift to longtermism if there were successful past examples of longtermist movements.
    But I think past examples of longtermism are easy to find- it’s just hard to find successful examples.
    When GiveWell was relatively young and not as influential as it is today, it commissioned work on the history of philanthropy, to answer questions like when did ambitious philanthropists succeed, when did they fail, and what effective altruists could learn from the past. I think repeating such a process for longtermism, by taking even a quick look at past efforts to prioritize the longterm future- casts doubt on longtermist efforts.
    Benjamin Franklin, Failed Longtermist
    If George Washington was the Captain America of the Founding Fathers, Benjamin Franklin was Iron Man. The fun one, the cool one, the guy who invented the lightning rod. He’s probably the closest thing America has to a Leonardo Da Vinci. He signed the Declaration of Independence, ran the post office, was the ambassador to France, and kept on inventing things.
    He also tried to be a longtermist. When he died, he left a bequest to the cities of Boston and Philadelphia that was to accrue interest for the next 200 years before the cities could access the whole principal. As Will MacAskill recounts, the amount grew to $5 million and $2 million respectively. The money mostly went to fund a private college.
    Because of this he’s sometimes favorably cited as a successful example of how people can intentionally try to help the longterm future and succeed. The problem is, of all the things that Franklin did that shaped the future, his intentional future-oriented bequest was basically a rounding error. No disrespect to the Benjamin Franklin Institute of Technology which benefited from his generosity, but that’s not what made Benjamin Franklin important to the world.
    What else could he have spent this money on? Taking better care of his health so he lived longer, supporting relatives to start a family yielding hundred of additional Franklin descendants over the years, running a few more experiments.. He could have thrown another party! This might sound like I’m joking but a big part of an ambassador’s jobs is to be a charming bon vivant and a great party host. An even stronger friendship between the United States and France would surely have been more consequential than founding this private college.
    There’s no contradiction between spending money now...
    12 min
  • LW - The ethics of reclining airplane seats by braces
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: The ethics of reclining airplane seats, published by braces on September 4, 2022 on LessWrong.
    I enjoyed reading the replies to this tweet, since it's a lower stakes issue that has all the contours of broader ethical debates. Granted, what triggered the tweet was not low stakes:
    There are passionately held beliefs on both sides (see replies to the original tweet for more). As with any argument, different principles lead to different conclusions. One principle with many adherents is that the rules follow directly from the design of the airplane:
    But this could still screw over the person behind you. So maybe reclining is bad, based on the fact that it harms more than it helps?:
    These views, by the way, are similar to what a travel industry analyst says to the NY Times: "Airplane etiquette is you only recline when necessary, and if you must recline, just put the seat back a little bit to get the comfort you need without encroaching too much on the person behind you."
    Another principle: the person in back of you could have "property rights" over the area directly behind your unreclined seat:
    But reclining may also be justified based on the consequences:
    Many other variables. Long haul vs. short haul:
    Dimmed lights:
    Meals:
    Height:
    If the replies are at all representative, this issue is in a bad state where a significant share of people have opposing beliefs about what's right and when. So we should expect to see more conflicts between passionate passengers.
    One potential solution is that the airlines try to coordinate everyone. An announcement could say "Our policy is that passengers should feel free to recline. Just check to make sure you do not spill the drink of the person behind you." This should douse the passions and lead to less conflict. A grumpy person being reclined on should feel less empowered; the loudspeaker announcement is common knowledge.
    Another thing airlines could do is sell reclining and non-reclining tickets. Then everyone knows what they're getting---another way of making the policy more explicit.
    This is not to say that either policy would be the best. Maybe an "asking equilibrium" is optimal, so that people can forge personalized agreements. "Is it alright if I recline?" "Sure, let me move my drink" or "Yes could I have a few minutes to finish eating?" or "I'm sorry but actually no, I'm extremely tall." I worry that this system would disadvantage nice people though.
    This seems like an issue where kind and reasonable people could disagree. With no (clear) connection to other ideological commitments, perhaps it's a useful exercise for understanding the other side.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • LW - Private alignment research sharing and coordination by porby
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Private alignment research sharing and coordination, published by porby on September 4, 2022 on LessWrong.
    Coordinating alignment efforts seems difficult when the work involves information that would ideally remain private. I'm not aware of any existing infrastructure to help with this, nor are the people I've asked so far. My understanding is that there are informal social networks which serve this purpose to some degree (e.g. passing google docs around), but it's difficult to safely scale these up.
    I've seen some attempts to help manage potentially risky information- for example, Conjecture's recent policy discussions: Conjecture: Internal Infohazard Policy - LessWrong
    These policies are primarily interested in containing information. That's important, but it sure would be nice to have a system that simultaneously helps with both privacy management and coordination.
    What kind of use cases could be targeted?
    Motivating hypothetical example 1:
    Hypothetical me is an independent researcher focusing on characterizing out of distribution behavior. I stumble across One Weird Trick for rapid grokking and generalization. I've run a variety of benchmarks and, aimed properly, it does remarkably well on HHH-style benchmarks. Unfortunately, the technique is trivially applicable to capability research; I observe extremely rapid training convergence and perplexity beating text-davinci-002 in a 2B model I trained on my workstation.
    It seems like I should tell someone about the fact that this exists, on the off chance it interacts positively with some other safety work, but telling the wrong person could shorten timelines by approximately 100%. I'm not deeply embedded in the field and don't have high-trust working relationships with anyone. Now what? Send an e-mail to Eliezer composed mostly of AAAAAA?
    Motivating hypothetical example 2:
    I'm another independent researcher looking to start a new alignment research project, but it's a topic which lightly touches on capability. The capability parts are already out in the public literature, but they're in an obscure subfield and I suspect mostly unknown by mainstream capabilities research. I worry that drawing attention to it could be a net negative even if my work is not capability focused.
    I think the research has potentially significant alignment value. I can't find any other related work in alignment. I'm not sure if this is because other people aren't publishing about it, if I'm not looking in the right spots, or if this actually an unexplored topic.
    I could ask in one of the public alignment-relevant communities, but I'd like to avoid leaking capability-relevant information if possible. Without already having lots of high-trust connections in the field, how can I find out if I'm doing redundant work, how can I get feedback about my research plans, and how could I safely tell people about the work if I do it?
    Motivating hypothetical example 3:
    I'm working in one of Googetasoft's seven AI departments. Some internal research incidentally found something that has some very interesting implications for alignment, but I strongly doubt the company is going to pursue the line of research because it isn't relevant for increasing capability. And they're also not interested in publishing it, because it isn't relevant for increasing capability.
    I've poked around the department, trying to piece together enough interest to get the relevant bits released. I've gotten some positive responses, but also some pushback- despite not being very interested in it, because it isn't relevant for increasing capability, the information is treated as risky to release because it could reveal future product efforts.
    Given the difficulty of separating the information from the implications it has for product development, and given the lack of secure channels for ...
    9 min
  • LW - Let's Terraform West Texas by blackstampede
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Let's Terraform West Texas, published by blackstampede on September 4, 2022 on LessWrong.
    [I'm half serious about this. I'm somewhat doubtful that it would work, but I like the idea. At the very least I think it would make a fantastic plot for a solar punk novel. If you want to rip it off for a story, go right ahead.]
    I went to an ACX meetup yesterday and mentioned an idea that I’ve had off and on for a few years- terraforming the West Texas desert. They suggested I write something about it, so here is that something.
    Land
    West Texas has a huge amount of cheap land. I just went looking again and the first page of results were all between about $800 and $1700 per acre. Sorting by cheapest price per acre, I found several lots of 50 acres or so for 27-30k, which works out to about $550 an acre- and there are cheaper prices for larger chunks of land- there are multiple listings for 1000+ acre lots. The reason I mention this is to make it clear that the land out there is cheap and plentiful. The reason why it’s cheap and plentiful is because it’s damn near worthless- it’s desert. The water underneath the land is often brackish. The heat is unrelenting. The sun never stops shining.
    Sunlight
    West Texas is hot, dry and bright almost year around. Go to this website and select El Paso from the drop down- that region gets 5.87-7.42 peak sun hours every day, where “peak sun hours” is sunlight at maximum, unobstructed intensity- ~1000 watts of power per square meter. This is prime solar-panel land, but you can’t get that power to anywhere that people actually want to live. At a minimum, it would require hundreds of miles of power lines and infrastructure. If you wanted to use your electrical power locally, however, you have a near-endless supply. The only problem is storing it.
    Storage
    Conveniently, one of the other things that West Texas has in large quantities is brackish water, which coincides neatly with the cheapest land prices. You could store your electricity by pumping this water up into large reservoirs and retrieve it by allowing the water to flow back down to the aquifer through turbines. This is called pumped-storage hydroelectricity and it’s about 70-87% efficient- you lose 20-30% of your power, but once the water is pumped to the reservoir, there’s no loss of power over time- it can sit in a covered reservoir forever. Cool- but now what? You have endless electrical power and a method of storing it until you need it, but what’s the point?
    Water
    The point is this mid-sized industrial desalination plant. With endless electricity and gravity-fed brackish water, you could divert some of that brackish water from generating electricity to a desalination plant. I haven’t been able to find any detailed discussion of the power requirements for one of these machines. As for capability- the manual states that it can handle up to 2000 ppm total dissolved solids (TDS), which maps to the “slightly saline” regions in this report (page 5). There may be other, larger desalination plants for sale that could handle more-brackish water (3k-30k ppm TDS) which would give you more options, land-wise.
    These desalination systems can produce 10-50k gallons of water per day. Which is a lot, but I'm not sure how much you would need- I imagine you could just have more than one desalination plant hooked up in parallel if one wasn't enough.
    Waste
    The waste from the desalination plant is all the crap that was in a large volume of brackish water, now concentrated into a much smaller amount of water. It seems like there are two possible ways to handle this waste. You could either dump it back into the aquifer that you pumped it out of or extract the minerals from the wastewater and sell them. Saline water contains salt, magnesium, calcium, potassium and a lot of other useful minerals. If it were dump...
    8 min
  • AF - The shard theory of human values by Quintin Pope
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: The shard theory of human values, published by Quintin Pope on September 4, 2022 on The AI Alignment Forum.
    TL;DR: We propose a theory of human value formation. According to this theory, the reward system shapes human values in a relatively straightforward manner. Human values are not e.g. an incredibly complicated, genetically hard-coded set of drives, but rather sets of contextually activated heuristics which were shaped by and bootstrapped from crude, genetically hard-coded reward circuitry.
    We think that human value formation is extremely important for AI alignment. We have empirically observed exactly one process which reliably produces agents which intrinsically care about certain objects in the real world, which reflect upon their values and change them over time, and which—at least some of the time, with non-negligible probability—care about each other. That process occurs millions of times each day, despite genetic variation, cultural differences, and disparity in life experiences. That process produced you and your values.
    Human values look so strange and inexplicable. How could those values be the product of anything except hack after evolutionary hack? We think this is not what happened. This post describes the shard theory account of human value formation, split into three sections:
    Details our working assumptions about the learning dynamics within the brain,
    Conjectures that reinforcement learning grows situational heuristics of increasing complexity, and
    Uses shard theory to explain several confusing / “irrational” quirks of human decision-making.
    Terminological note: We use “value” to mean a contextual influence on decision-making. Examples:
    Wanting to hang out with a friend.
    Feeling an internal urge to give money to a homeless person.
    Feeling an internal urge to text someone you have a crush on.
    That tug you feel when you are hungry and pass by a donut.
    To us, this definition seems importantly type-correct and appropriate—see Appendix A.2. The main downside is that the definition is relatively broad—most people wouldn’t list “donuts” among their “values.” To avoid this counterintuitiveness, we would refer to a “donut shard” instead of a “donut value.” (“Shard” and associated terminology are defined in section II.)
    I. Neuroscientific assumptions
    The shard theory of human values makes three main assumptions. We think each assumption is pretty mainstream and reasonable. (For pointers to relevant literature supporting these assumptions, see Appendix A.3.)
    Assumption 1: The cortex is basically (locally) randomly initialized. According to this assumption, most of the circuits in the brain are learned from scratch, in the sense of being mostly randomly initialized and not mostly genetically hard-coded. While the high-level topology of the brain may be genetically determined, we think that the local connectivity is not primarily genetically determined. For more clarification, see [Intro to brain-like-AGI safety] 2. “Learning from scratch” in the brain.
    Thus, we infer that human values & biases are inaccessible to the genome:
    It seems hard to scan a trained neural network and locate the AI’s learned “tree” abstraction. For very similar reasons, it seems intractable for the genome to scan a human brain and back out the “death” abstraction, which probably will not form at a predictable neural address. Therefore, we infer that the genome can’t directly make us afraid of death by e.g. specifying circuitry which detects when we think about death and then makes us afraid. In turn, this implies that there are a lot of values and biases which the genome cannot hardcode.[This leaves us with] a huge puzzle. If we can’t say “the hardwired circuitry down the street did it”, where do biases come from? How can the genome hook the human’s preferences into the human’s world ...
    45 min
  • LW - Why was progress so slow in the past? by jasoncrawford
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Why was progress so slow in the past?, published by jasoncrawford on September 1, 2022 on LessWrong.
    What explains the hockey-stick shape of world GDP over time, with seemingly no progress for thousands of years, followed by soaring growth?
    The first question you might ask is: is this just an exponential curve? If so, then the explanation is simple: we see a constant growth rate every year, and the steep upward slope is just what exponential curves look like. There’s no more mystery than there would be about the shape of a population curve.
    To check this, we can plot the same numbers on a logarithmic y-axis. In such a chart, exponential curves become straight lines. But when we do this, we still see an upwards slope. That is, the growth rate has increased over time:
    Nor is it just a consequence of population growth, because we see the same pattern in GDP per capita:
    I’ve previously said that a core reason for this is that progress compounds, creating a flywheel effect. Here’s another way of looking at the same idea. Why wasn’t the threshing machine invented in, say, the 1300s? Consider all the barriers to such a thing:
    Human capital
    First, who would have invented it?
    Would it have been a farmer? (Well over half the workforce were farmers.) When would he have found the time to tinker? Labor productivity and incomes were low; there wasn’t much spare time or material for inventing.
    Who else could have done it? There was no established professional class of inventors, engineers, or entrepreneurs. The closest were skilled craftsmen who made machines, such as clockmakers or millwrights. Overall there were many fewer inventors per capita than today (ok, I don’t have data on this right now, but I’m pretty confident in this assertion).
    If it were, say, a millwright, he would have to learn enough about machines to go beyond the kinds that he had been taught to make through apprenticeship, and invent something entirely new. Where would this knowledge have come from? There was no printed material, and no mechanics’ institutes.
    And if our inventor did have mechanical skill, why why would he decide to apply it to a practical invention for farmers? It was more prestigious and lucrative to make clockwork novelties for the aristocracy. Even if the inventor did have a practical bent, there were social taboos against labor-saving devices.
    Manufacturing
    Suppose that despite all of this, some inventive person is determined to make a threshing machine and acquires the resources and time to experiment. Maybe he avoids the trap of making a machine that mimicks human motions, and hits on the idea of a rotating drum with teeth.
    He will find that making the machine work reliably is very difficult. It requires a high degree of skill in the mechanic that crafts it by hand. There are no machine tools to create precision parts. Wood is too soft for precision work; metal parts are required. Machines that are shoddily constructed break easily or bruise the grain instead of threshing it. (See my full post on the threshing machine for elaboration.)
    Distribution
    Suppose our inventor overcomes all these obstacles, and manages to create a practical threshing machine. What next?
    He could use it on his own farm, if he is a farmer. If that’s all he does, then his invention has had no significant impact on the economy and no impact at all on history. This is not what we are seeking to explain.
    To matter for progress, the invention needs to be distributed. And here our inventor faces more obstacles.
    Who is the market for his invention? Will other farmers be receptive to it? They need to change their methods and take a risk on something new, something they are not used to doing.
    Even if they are willing to take that risk, do they have the capital? Buying a piece of agricultural equipment is an investment that ...
    11 min