A bunch of links

Okay, I’m hopping on the “posts full of links” bandwagon.

• A piece of John Ioannidis’ research I haven’t heard before: in medical research, observational studies seem to agree surprisingly well with randomized trials. He rightly notes that there are a bunch of caveats: in six out of 25 comparisons the two methodologies didn’t agree (and in half of those cases the effects were in opposite directions); the sample was small (and didn’t include several well-known failures of observational trials, like beta-carotene and alpha-tocopherol) and almost all interventions are not studied with both observational and randomized studies, introducing sampling bias. Still, it’s a pretty interesting result. Does anyone know of similar studies in global health/epidemiology?

• Effective altruists considering entrepreneurship take note: the NBER paper Schumpeterian Profits in the American Economy suggests that entrepreneurs and innovators capture about seven percent of the social value that they create. This suggests that the positive externalities from a company you start might be on the same order of magnitude as the good you do by donating the money you earn. It also suggests that entrepreneurship in the developing world might be competitive with that in the US: even though the markets are smaller (so you’ll make less from it), the good that you do through externalities is greater (because it goes to people who are worse off).

• For some reason all my friends are into low-attention-cost food replacements now. The best-known is of course Soylent, although production has been held up for months due to manufacturing delays and some people have expressed skepticism about the “throw in isolates of all known micronutrients” (whole-foods based options are more robust if we discover some micronutrient that we didn’t previously know about). Holden Karnofsky (of GiveWell fame) thinks you should drink Power Smoothies, which are drinkable like Soylent but have bananas and spinach and stuff (and the ingredients are way easier to source). (When I worked at GiveWell they went by the shorter if more off-putting name of MIAC, or Meal-In-A-Can.) Romeo Stevens and John Maxwell think you should eat their MealSquares, at least once they start producing them—the recipe is designed to be nutritionally complete and the one I had (from a previous iteration) tasted delicious, like dense banana bread.

• Open Data Stack Exchange has a lot of interesting info on freely-available datasets. Want to find some open data to play around with? They have one thread that serves as a database of databases of open databases. I just did a project for a psychology class analyzing a ton of Twitter data and it was pretty fun—if you have similar assignments, or if you’re just curious, this could provide some interesting fodder for analysis.

• Ryan Carey has a list of status updates from effective altruist organizations and individuals. Great work compiling all this stuff, Ryan! Makes me wish there were some way this could happen automatically (i.e., without Ryan combing the Internet every six months or something).

• Less Wrong turned five years old a while ago! It’s a common observation among LW readers that much of the material seems revelatory when you read it for the first time, and totally obvious when you re-read. Scott Alexander does some intellectual history and figures out what we’ve all forgotten we didn’t know.

• Albert Wenger of Union Square Ventures is doing some interesting blogging about where he thinks the future is going:

  Here are some of the pieces that I will be grappling with in the coming days and months

  1. The massive cracks that are appearing in the existing system. In particular, the industrial system is built on people selling their labor and using the proceeds to buy goods (and services). But the price of labor is under tremendous pressure from machines and from globalization. Other cracks include the increasing importance of non-rival information goods and the missing prices for the environment.

  2. The early signs of what a different system might look like. As I mentioned above, these include many of the USV portfolio companies. But they also include projects such as Polymath and not for profits such as the Freelancers Union and the growing cooperative movement.

  3. The nature of mankind, which admittedly sounds awfully pretentious. But our views about our own nature are at the heart of what systems we believe can work. For instance, if you conceive of people as fundamentally selfish and greedy and prone to violence then your outlook will be very different than if you believe in curiosity, altruism

  4. History and systems. These may seem like an odd pairing but I have a good reason for it. Systems sustain themselves through complementarities between their components. We tend to be too obsessed with trying to pin change on a single force whether that’s technology or policy or individuals.

  5. Ideas about what we can do as individuals and organizations to help shape the transition.

  6. The relationship to existing philosophies and schools of thoughts. I have come to realize that both capitalism and Marxism are essentially rooted in industrial society. They are like the proverbial generals fighting the last war. We need to get past both of them.

  Along the way I am looking forward to lots of poking at my thinking from readers and a gradual process of collaboration and contributions. I have not yet figured out what the best way for that is, so for now, please comment!

  I’ve found it quite thought-provoking so far. Plus, he actually responds to the comments!