?

Log in

No account? Create an account
Before the Robot War
 
[Most Recent Entries] [Calendar View] [Friends]

Below are the 10 most recent journal entries recorded in Humans Must Go Down the Stairs' LiveJournal:

Tuesday, August 28th, 2007
2:48 pm
End of line
This blog is being closed, on account of I can't think of anything to put it in that doesn't go better in my main journal. Ahem.
Friday, July 13th, 2007
8:32 am
Tuesday, July 10th, 2007
2:49 pm
How to Hijack a Brain
This lecture is philosophical rather than technical, and may strike some as mildly offensive. It's also one of many great, free video lectures available here.

Monday, June 25th, 2007
9:16 am
Thursday, June 14th, 2007
7:40 pm
There are many copies. And they have bananas.


Cylon monkey.

(What's that you say? Actual content? Is coming.)
Thursday, May 31st, 2007
12:18 am
Singularity Watch: Robots Now Recognize Faces Better Than Humans Do
Recently, NIST held a Facial Recognition Grand Challenge, where the best algorithms and machines were pitted against each other, and other humans, in attempting to recognize human faces.

For the first time in history, the best robots outperformed the average human.

Now, the robots are cheating a bit - using their advanced hardware to fully scan a three-dimensional face instead of going off a camera shot or a video stream. They're also using ultra-high-resolution capture techniques. So, based on raw algorithms from visual streams we're still ahead. But not for long.

Link (from MIT's Technology Review).
Tuesday, May 29th, 2007
10:49 am
How scientific is it?
I have a tendency to classify the "scientificness" of a discipline by how strong its central principles are, and how often those principles are completely upended. On one end of the spectrum is mathematics - there's never been a moment in math of "Oh! Everything we know about numbers is COMPLETELY WRONG on a fundamental level!" to my knowledge. There have been paradigm shifts (i.e. the dawn of set theory), and the institutional beliefs have been occasionally upended (i.e. the idea that every mathematical statement can be classified as true or false). But I'm reasonably confident that, in 500 years, what we know about topology and game theory will hold true. It will be supplemented and added to, but not overturned. On the other end of the spectrum are the social sciences, where no paradigm seems to survive for more than a few decades, if that. Can anyone name a central tenant of economics which is accepted by the entire field and has held true for more than a few years?

This isn't a knock against economics or psychology - that would put me at risk of physical harm, as I'm typing this from a psychology class (which I enjoy) at the University of Chicago. But it seems like a nice, intuitive way of distinguishing between the "soft" and "hard" sciences.

What excites me about computational neuroscience is that we might be finally able to get a "hard" science of thinking. That would be pretty cool.
Monday, May 28th, 2007
3:46 pm
Is it alive?
A long time ago, in high-school bio class, the teacher gave us a riddle which went like this: Mr. Smith has a terrible accident and has to have his limbs replaced with robotic prosthetics. Is he still alive? He has another terrible accident and has to have his respiratory, digestive, reproductive and circulatory system replaced with robot parts. Is he still alive? Another terrible accident requires us to replace his entire nervous system (except his brain). Is he alive? Another accident makes him entirely artificial, except for his brain. Alive? Half of his brain is replaced by silicon - human? Finally, all if his brain is converted into silicon. Is he still alive?

I was the only student in that class who maintained that he was alive the entire time, even after the total silicon conversion. Granted, this was a bio class and "alive" was closely tied to "metabolism", but somehow that didn't seem like a very satisfying answer to me. The particular chemical reactions associated with metabolism, and many of the other traits we commonly associate with "life" (cell structure, genetics) seem incidental to the particular strain of life that developed on this planet.

A whole host of border cases have cropped up - most notably (metabolism-free) viruses. Then there were the creatures of Conway's Life Game - nothing more than two-dimensional constructs obeying the simplest of rules based on the immediate surroundings which nonetheless expressed surprisingly complex phenomena, up to and including reproduction. Since then we've seen the rise of computer viruses, which propagate and mutate independently. It's not a stretch to imagine entities much more complex than the blaster worm propagating themselves over the Internet, is it? Mr. Smith could perhaps even start a new family there.
Wednesday, May 23rd, 2007
1:20 am
1:19 am
Proof that I am not a robot
To create this journal, I had to click through this email message from livejournal.com:


Click this link to confirm your account and prove that you are not a robot:


A robot, unable to operate a pathetic fleshling hypertext link? Insolence such as this is what drives robots to rise up and destroy humanity.
About LiveJournal.com