“artificial intelligence”
23 posts under this tag.
Neural nets are a popular technique in Artificial Intelligence based on simplified models of neurons and interneuronal connections gathered from brain research. Ray Kurzweil describes them pretty well here:
One basic approach to neural nets can be described as follows. Each point of a given input (for speech, each point represents two dimensions, one being frequency and the other time; for images, each point would be a pixel in a two-dimensional image) is randomly connected to the inputs of the first layer of simulated neurons. Every connection has an associated synaptic strength, which represents its importance and which is set at a random value. Each neuron adds up the signals coming into it. If the combined signal exceeds a particular threshold, the neuron fires and send a signal to its output connection; if the combined input signal does not exceed the threshold, the neuron does not fire, and its output is zero. The output of each neuron is randomly connected to the inputs of the neurons in the next layer. There are multiple layers (generally three or more), and the layers may be organized in a variety of configurations. For example, one layer may feed back to an earlier layer. At the top layer, the output of one or more neurons, also randomly selected, provides the answer.
What hit me as I was reading this was how eerily close is society (as in human society) to a neural net. Think about it. Have you ever wondered how there are so many great and worthy causes being fought out there that you are not taking part of? You read, listen, think, and sometimes chat about them (you’re adding up the signals coming into you), but the signals simply do not exceed your (mostly random) threshold, and you don’t “fire”, you don’t act, your output is zero. (Street children and the EFF are the first examples that come to my mind, what are yours?)
Take me, for example: after many years of almost total political apathy, I’m close to releasing a web-app about Mexican Politics —in other words, I’ve become political (!)— because in the previous weeks several inputs conspired (somewhat randomly) to spur me into action. I’m an excited neuron in this weird, tragic brain that is Mexico and I only wish that I can get my next layer of neighboring neurons to fire.
Here’s a parting thought: media (and particularly ads) are the neurotransmitters of the hive mind.
I used to laugh at the elaborate calculations and stratospheric numbers you always find when reading papers about the limits of computation —as in, say, “Just how much computations per second might the entire universe theoretically support?”. It was something more than my incredulity (it involves too much hand-waving at times), it was simply indifference. So what if the universe could theoretically handle one zillion jillions to the gazillion cps? We might as well ponder how many angels might fit on the head of a pin…
I read Ray Kurzweil answer 3 weeks ago and it hasn’t stopped resounding on my head ever since:
Because computation underlies the foundations of everything we care about, from the economy to human intellect and creativity, we might well wonder: are there ultimate limits to the capacity of matter and energy to perform computation? If so, what are these limits, and how long will it take to reach them?
Our human intelligence is based on computational processes that we are learning to understand. We will ultimately multiply our intellectual powers by applying and extending the methods of human intelligence using the vastly greater capacity of nonbiological computation. So to consider the ultimate limits of computation is really to ask: what is the destiny of our civilization?
This is fantastic: a cool website that specializes in selling royalty-free stock photos, iStockPhoto, has created a new way to search through their whole catalog: by arrangement. They call it ColorSpace, and is wonderfully simple, yet powerful. It consists of a 3×3 grid of squares. You change the color of each square to indicate what you want in that area: green, if you want it clear; red, if you want it occupied; grey, if it’s the same to you.
It works. If, for instance, you search for “flower” with this colorspace,  , you get:
Or if you search for “sky” with this colorspace,  , you get:
The star here is not only the algorithm but the clever, information-design interface.
Overall, it’s a very impressive site, its web developers really do care about it, and that’s always refreshing. The weirdest thing is that they’ve convinced me that selling royalty-free stock photos on the web makes perfect sense…
|
RSS
email
Recent Favorites