I program by day, and program some more by night. I just finished Go Scoring Camera, an Android app that does computer vision to interpret Go boards, and I'm starting on Keyboard Builder, a tool for customizing the Android Phones' on-screen keyboard. I'll keep writing phone apps until they're sufficient to provide a passive income.

Making bayesian statistics easier and more accessible by coding advanced sampling algorithms for PyMC

Some background: I took statistics in high school because it seemed vaguely useful. Unfortunately, the material seemed very dry and involved mostly memorization and few general principles. It was boring and limited. College statistics was the same thing. I did some internships and statistics seemed very useful for figuring things out, but I didn't know how to do very much.

Later I started reading Overcoming Bias, and Yudkowsky kept mentioning this thing called "Bayes theorem" and how it was really powerful. I read a stats book on Bayesian Statistics and my mind was blown. The statistics that I had been taught was a collection of formulas that gave answers but not much insight, but Bayes theorem encapsulated not just all of the statistics I had learned but the very notion of "learning from data." I was hooked.

Later I figured out that the curse of dimensionality makes complex problems difficult (even though the simple statistics taught in stats classes now are still easy).

My project: Bayes theorem provides a simple coherent framework for learning from data. It mas... (read more)

I'm writing the core of a system similar to Google Wave, but much simpler, more flexible, and mathematically elegant. It works peer-to-peer, too, which is nice. It's got some clever algorithms in it, and even cleverer data structures, and I've been through several iterations of the design and code. It's probably the most complex thing I've ever written, and it could be pretty darn handy in the future. This is a surprisingly difficult problem to solve well.

(I also did a quick detour one-off project: pwstore, for anybody wanting to store passwords in Haskell. I feel pretty good about this one; it's useful.)

I'm also writing a NaNoWriMo book, which I'd left half-finished back in November, because it actually was a lot of fun to write. I've got one wonderfully snarky character; if I give her first-person narrative, the story becomes extra fun. I don't even know how many words I'm up to now. It's a sprawling fantasy adventure thing.

I'm writing a blog post every day in order to cultivate a writing habit, create a useful body of introductory material for the subfields of philosophy I find most worthwhile, and perhaps make a good impression on future employers (I'm still a college student and don't have the competence-demonstrating portfolio I'd like).

Edit: my blog is located at http://www.wrongbot.com

This is a great idea for a thread! We should have these at regular intervals.

I'm working on turning my PhD work into a product. The idea is to take logic-based specifications, expressed as structured english, and produce working information systems (such as web apps) from them. I have a prototype which got me the PhD but I need more work to get to a viable product. An intermediate step I am focusing on at the moment is using this technology to write validation checks for existing large datasets to pick out errors/outliers. So someone would apply the rule "Each patient with diabetes type I must be prescribed insulin" on a relevant dataset and the system would pick out cases where this does not hold. Normally this would need a complex SQL query that must be written by a developer and cannot be verified by the domain expert.

The other thing I'm working on is nurturing my addiction to the 17x17 challenge. This is my first forray into 'serious' math and I'm finding it extremely addictive. I usually dive into this when my motivation is low, and lo and behold, I'm motivated again! Having spent 7 of the last 14 months on this, still no solution, but I do think that if the filters ... (read more)

I'm working on tightening the bounds on the Jordan-Schur theorem. I've improved the best known bounds but not by much. If this project does succeed it might end up turning into my PhD thesis.

1) Expanding my programming capabilities. This is because it's vital if I'm ever to develop one of my other big ideas. I can come up with algorithms and models pretty easily, but am not at the stage where I can put them into a user-friendly, non-command-line format, or build off of others' code (which rarely comes with easy compiling instructions).

I've tried various frameworks, but they're all hard to get started on -- more so than the inherent difficulty of programming would suggest. I follow the instructions just to get it set up, but they always leave out something vital so that I have to get an expert to look at why it doesn't set up right. I've tried Django+Python, the Android SDK, Matlab (incl. with Simulink), and .NET, becoming most proficient with the last two.

It's not all gloomy, though. Some successes: completing a useful internal development project at work that involved setting up a GUI to allow for easy database lookup and clean presentation of data. In technical languages, I've set up something similar but for signal processing and automated generation of relevant plots. I was able to modify a Firefox extension so that I could deftly browse websites from the ke... (read more)

I am thinking about the design of quantum money, quantum copy-protected programs, and quantum program obfuscation. Basically, understanding whether and how quantum computers can implement the strongest possible cryptographic operations (all of which are known to be classically impossible and to have a wide range of applications).

I am working on the development of collaborative filtering / recommendation protocols for large communities. This includes, for example, a system for aggregating product reviews which is robust to the presence of large numbers of planted reviews, or a system for spam filtering based on trust which remains robust in the presence of a supermajority of sophisticated spammers. More generally, this work seems like an important first step in the development of automated and robust systems of trust.

I started working on these projects because they were interesting problems I was well equipped to work on which seemed likely to yield publications in time for graduate school applications (optimistically, these positions have been born out in both cases). I don't recommend this decision making process.

I think automating trust and designing better recommendation systems are more important than the overwhelming majority of theoretical problems currently studied, but I have realized more recently that there are more important issues to deal with. I am continuing to work on these problems now because of inertia, the fallacy of sunk costs, and a desire for status.

I'm in the early stages of my PhD research in metabolomics, which specifically means that I'm constructing what will end up being a library of tagged proteins to use in probing protein-small metabolite interactions. While the genomes and proteomes of quite a few model organisms are well-understood, cross-pathway regulatory interactions are another extremely important factor in metabolism, and these have had only minimal effort put to characterizing them. My work is therefore aimed at finding and characterizing these interactions on a broad scale, and inc... (read more)

Investigating programmer productivity. Learning about the history of the programming profession in general, of the "software engineering" meme in particular, and detailed history of the "agile" meme. Learning some R and stats on the side.

Why - because there is a lot of leverage in looking at the question of why programming skills are what they are, and how they could generally be better, compared to just improving my own. There seem to be some low-hanging fruit, such as encouraging programmers to be more aware of cognitive biases affecting their work, or of the nature of probability and tools for working with it.

I am trying to understand personal interaction in a semi-technical fashion in order for me to have mental models which are intuitive enough to me which can guide me in casual (and, later, business) interaction. I've read through some of the OB posts on status and I'm currently reading the book Impro (which Robin Hanson recommends here). I've also partially made it through Dale Carnegie's How to Make Friends and Influence People. Book suggestions are very welcome.

I am working on a 2D adventure game that features some topics from lesswrong: rationality and recursive self-improvement.

I love making games and this seems like a good way to take the basic concepts we take for granted on LW (like AI going FOOM or cryonics) and present it in a different way, from a different angle, in a simpler fashion, in a different medium, to new audience.

Goal: make enough money that I can do this again. And again. And again. Continue making games about various lesswrong topics.

I'm working on making music that doesn't suck. Bleepy electropop, to be precise. Using LMMS, a cheap'n'cheerful open source clone of FruityLoops. I've been an avid record nerd for nearly thirty years now, it's time I saw what I could do.

Current state: First demo. Something almost done. A minor diversion. I'm getting to know the capabilities and limitations of LMMS and Audacity (which is basically the four-track we all wanted twenty years ago as a computer program). I'm accumulating a huge pile of fragments, which need putting together into actual pieces. I... (read more)

I'm working on an iPhone game I started when learning Objective-C. It's unknown (about 3000 free downloads/week) but loved (mostly rave App Store reviews), so I'm trying to make it go viral.

The version awaiting Apple review has incentives for introducing new players, and usage analytics to get a more accurate idea of what to fix and simplify. (Fan feedback is great, but they're already converted & over the learning curve). I think I'll still need to lower the learning curve/barrier to entry & boost the purple cow factor, but I'll bet I see somethin... (read more)

I've had a little success with two books collecting some of my blog posts so now I'm trying actively to turn my blog writing into longer-form work suitable for book publication. I'm currently working on three series of blog posts simultaneously, all of which I hope to turn into books.

The first, How We Know What We Know is on a lot of the things people talk about on here, trying to explain Bayes' theorem, Kolmogrov Complexity and the scientific method to a lay audience. I'm doing this because a lot of my blog readers, especially those who enjoyed my book &q... (read more)

I am working on a new approach to epistemology -- essentially an extension of Frege's context principle -- that explains the nature of meaning and existence. I am finding that it provides traction on many problems in epistemology, linguistics, rationality and philosophy in general. I also believe that it provides traction on problems in more practical fields like physics and computer science.

I think that my approach to epistemology is pretty cool, that it is essentially a "Theory of Everything"; however epistemology is not my field of training or... (read more)