The prosecution responsible for the recent nation wide injunction against embryonic stem cell research has written this article, in their defense. A fellow neuroscience graduate student described it "one of the least factual arguments I have ever read". Many of the comments ( explaining why Mr. Bowmen is bad at science ), are insightful. I don't need to duplicate why Matt Bowman is oh so very wrong here, but it reduces one word : "epigenetics". Adult derived stem cells aren't going to replace embryonic lines any time soon.
--- [[ meta ]] ---
Beck has noted that we seem to have degenerated into commenting on events that happen in the news. This is fairly accurate. The truth is that few blogs sustain themselves with daily updates of original content. The ones that do are called news, webcomics, and the weather report. Blogs with large readership scrape the internet, post, link, maybe sometimes even analyze or comment. Who wants to spend all day looking for funny videos of rats staring at each other just to keep readers fed with a constant stream of novelty? We do not need to be such a blog. We have few readers, and therefore can post whatever the hell we want. I think we can even swear in posts without upsetting people.
Beck has noted that he feels bad about posting raw research on WeAlone. I say this is nonsense, all of us are in graduate school now, and most of what we have to say is going to take the form of raw research for the foreseeable future.
Here, I'll start : what do you think the most natural generalization of principle component analysis (PCA) to finite fields ( like bit vectors ) is ? In an offline discussion, Beck and I decided that the notion of "vector of maximum entropy" makes more sense than "vector of maximum variance", when your vectors have elements from a finite field. So, we think its possible to define a PCA like operation for distributions of vectors with elements from a finite field using entropy instead of variance. However, we have no indication that this is even remotely useful.