There’s a new effective altruism forum up! It has a bunch of awesome posts, including one by me on choosing a career/job. (Even though it’s on the EA forum, it should be of fairly general interest.)
Follow-up from a previous link: the amazing Elm debugger has expanded into Elm Reactor and moved into the core Elm distribution. I almost wish I was coding for the browser again just so I could use this stuff, it’s that cool.
This one’s not just a follow-up but a repost, but it merits it: now that I’m not a slave to the Harvard meal plan, power smoothies have totally taken over my house. They’re like Soylent except they’re actually made of food and they also taste good. My initially-skeptical roommates are now also hooked. If I start talking about $500 blenders, somebody please stage an intervention.
Think you’re good at generating random numbers? This webpage uses a battery of advanced statistical tests (like the Exact Fishy Test) to falsify your hypothesis. Try to see how high of a p-value you can get! (My personal best is 0.11.)
GiveWell has some cool new stuff going on! They rebranded GiveWell Labs as the Open Philanthropy Project and have been picking up steam, making a bunch of grants in macroeconomic policy and labor mobility.
They’re also planning a grant for start-up funds to an organization focused on increasing organ donation rates via providing benefits to organ donors, which would save lives and save the healthcare system money. As far as I know, this is the first organization that’s actually been started with a GiveWell grant—hopefully there will be more in the future!
At the Effective Altruism Summit (okay, yeah, this one’s a bit old), data scientist and journalist Jonathan Stray gave an interesting presentation on qualitative research methods, with an associated reading list. The basic idea was that while a lot of the attendees were very familiar with standard quantitative academic methods, not many (if any!) knew much about research in more qualitative fields (editorial: probably mostly due to hard-science chauvinism). I’m working my way through it as part of my project to get a bit outside my incredibly small filter bubble—going slowly so far but it looks like an interesting perspective that’s orthogonal to most of the other stuff I’m reading, so helpful in that sense.
Piero Scaruffi is a guy who has written a truly astounding number of reviews of jazz, classical music, rock, art, literature, nonfiction, film, travel destinations, hikes, charities, and probably most other objects that exist. Also histories of those things. Also original books and essays. Seriously, it’s insane. (via Holden Karnofsky)
Comments
Hah, clever joke. But it makes me wonder:
How items in a sequence of numbers from 1-100 do you need to look at in order to decide with 99% confidence whether the numbers were generated by a human trying to be random, or by a true random generator? How would you train such a classifier? (You’d want to teach it the sorts of patterns that humans generate when they’re trying to be random…)