TODO: insert description here
I picked up Alien Clay because I like Hugo nominees and I like Adrian Tchaikovsky. Adrian’s great love is hideous nonhumanoid aliens, either buglike or tentacle-based or both. He relishes the biological details, from the properties of humanoid skin, to the lifecycle of terrestrial fish parasites. How could an ecosystem develop so that for example grasping limbs were actually symbiotic creatures?…
David Foster Wallace was one of my earliest “real literature” dives. From the days when, sometimes after a long day of work, I just didn’t want to do more work on the Google shuttle. I got back into reading after not reading so much in a personally-directed way in college. Recently, flipping through Brief Interviews with Hideous Men, I found a business card from Google that I had been using as a…
Sometimes you know things because of theorems whose form is, when you know one true thing, then you know another true thing. P implies Q . If x is even, then x^6 is divisible by 64 . And 10 is even, therefore 10^6 must be divisible by 64 . Figuring out when this applies is the “library search” problem I wrote about last month . Rewrite search is a bit different. A rewrite happens when you know…
Last month I wrote about building a math AI and mentioned that I think we need to optimize the “library search” problem. In this post I’d like to explain what library search is, why it’s important, and what we can do about it. Motivation A few years ago I got excited about the IMO Grand Challenge. I thought it would be interesting to look for a team of smart people working on it and see if I could…
How can we make an AI that is really good at mathematics? Some AIs are okay at math right now. GPT-4 can do okay on the math SATs. But personally, intuitively, I feel like math is like chess. A math program should be able to completely dominate humans at math, the way that chess programs dominate humans at chess. Is scaling all we need? Maybe Nvidia launches the H200, H300, H400, H500, all the…
I feel a bit detached from my resolutions for last year. Like I have to drag myself into considering them. A quick review. 1. Health I’ll give myself a “does not meet expectations”. It sounds bad when I put it like that, huh. Well, I was below my 189 weight target for 4/12 months of the year. Yearly average 189.3 so it wasn’t a complete disaster. So, maybe given your typical performance review…
The core idea behind “AI alignment” is that superintelligent AI will be an agent maximizing some utility function, either explicitly or implicitly. Since it’s superintelligent, it will be really good at maximizing its utility function. So we, as humans, need to be sure that this utility function is “aligned” with something that humanity finds acceptable. In particular, we don’t want the AI to turn…
My resolutions from last year went pretty well so let’s do that again. 1. Health I plan to keep working out four times a week, but this is enough of a habit that I feel like I don’t need to “resolve” to do it. Diet-wise, my target weight is below 189. Counting calories is the most straightforward way I’ve found to achieve this, but I want to be clear about the metric that I most care about here.…
Last year I made some New Year’s resolutions . Let’s take a look at how they did. 1. Calorie Counting I got weak at actually counting calories toward the end and skipped like half of December. But I did a lot better than I expected at losing weight. Somehow I am reluctant to make an explicit weight goal on here, but the bottom line is to get healthier and for getting healthier for an overweight…
Previously I wrote about using radio telescope arrays to look for aliens and today I have even more to say on the topic. What is a Radio Telescope Array Again? It’s a bunch of radio telescopes that are all pointed in the same direction. A radio telescope takes measurements in the form of complex numbers. These are voltages, but I prefer to just think of them of big 2d arrays of complex numbers.…