Intrinsitivity

Behind the Fall of Imus, A Digital Brush Fire – WSJ.com

On the morning of the original broadcast, there was little response to Mr. Imus’s slur. Media Matters posted the video and transcript on its Web site and sent an email blast to several hundred reporters, as it does nearly every day. The post received dozens of comments, many heated, some more than 300 words long. The next day, top news outlets didn’t mention the incident.

According to the article quoted above, the firing of veteran radio host Don Imus was a direct result not of Imus’s racist slur, but of a video and transcript of the slur posted on mediamatters.org. American radio hosts are infamous for impropriety.

Continue Reading »

If your screen allows, I recommend watching this video full size.

The article below discusses recent research on computer-based image recognition proving that computers can now identify images as fast and accurately as humans. As an aside, I note that the article ends by noting that “research was supported by grants from the NIH, DARPA, ONR, and NSF.” This makes me wish I could be involved with applications of this research. Surely, some of those applications will be publicized.

ScienceDaily: First Impressions: Computer Model Behaves Like Humans On Visual Categorization Task

Now, in a new MIT study, a computer model designed to mimic the way the brain itself processes visual information performs as well as humans do on rapid categorization tasks. The model even tends to make similar errors as humans, possibly because it so closely follows the organization of the brain’s visual system.

Last but not least, below is a video of a Google TechTalk given by Luis von Ahn, who in 2005 received his Ph.D. in Computer Science from Carnegie Mellon University. Luis explains the problem of computer-based image recognition and the emerging trend of using human computation (i.e., technocratic slave labor; just kidding) to solve problems that computers currently cannot, at least, as of his talk.