Next Event

Friday, October 24, 2008, 07:00 PM: Life Extension with Jerry Emanuelson

Jerry will be talking about his 25+ year experience with life extension treatments, including hormone injections, his longevity doctor, what treatments to ask a doctor for, how to find and guide a doctor, an interesting example of the 'medical priesthood vs. empowered patient' conflict as more healthcare treatments are about prevention/enhancement, getting his DNA scanned with deCODEme and opensourcing his genome on the SNPedia.com, and more.

More...

Back to Event List

Artificial Intelligence

This is a PAST event. See "Meeting Notes" section for audio, video, documents and other information.

Original event date/time: Friday May 25th, 2007, 7:00 pm to 9:00 pm

We decided this topic at the end of our last meeting on Nanotechnology. The two topics seem to go hand in hand, so we're going to do them back-to-back. Let everyone who you think might be interested in AI know about this meeting!

Abstract:

Videos to watch (collected from earlier boulderfuture posts):

Eliezer Yudkowsky's video from the Bay Area (Silicon Valley) Future Salon

Novamente (Ben Goertzel) videos

Hugo de Garis's "Bulding Gods"

Sebastian Thrun's "Winning the DARPA Grand Challenge"

Jeff Hawkins "Hierarchical Temporal Memory"

And stuff to read...

Marshall Brain's websites:

Robotic Nation

Mana

And of course, who can forget that mother-lode of AI websites, KurzweilAI?

KurzweilAI


Video of Ray Kurzweil talking about exponential change in technology and how to predict the future...

Killer App Expo - Ray Kurzweil

This is a past event.

Meeting Notes:

We have audio for this event.

Participants: Wayne (me), Doug, Rich, Sarah, Christopher, Robert, Ann D., Ann O., Annie.

A random meandering conversation touching on many subjects such as

  • Playing Go and playing Chess

  • The Turing Test

  • The Grand Challenge clever tricks that humans can't do

  • your brain is smart at figuring out what your eyes see

  • brains and evolution, brains evolved only 700,000 years ago

  • The Loebner prize, and how humans were judged to be machines

  • how people confess to Eliza, how Eliza should be hooked to Google

  • neural nets worse than insects, and can't introspect

  • possibility of making AI's as intelligent as a philosopher

  • programming from the conscious mind vs subconscious mind

  • Monica Anderson's robot that walks like a starfish, versatility and adaptivity of walking gaits

  • is it a cockroach, or is it a tape recording?

  • genetic algorithms to learn to walk -- by somersaulting

  • "I want a program for my car to display the last speed limit sign seen"

  • stitching together photos to make a panaramic view, motion processing of the eye

  • our ability to remember faces, how much processing power do you need for image processing?

  • language is very high level and can't be broken down into a block diagram

  • how to screw up the Google Translater when translating from Chinese

  • Stuff between the lines is between the lines, double meanings, languages are inherently ambiguous, "trouble may wander far and wide, but it is always near"

  • Shannon's Limit on communication bandwidth

  • how can a program take multiple branches on an "if" probabalistically?

  • Computers can do integrals (the Risch Algorithm) like the world's best mathematicians, but can't do anything a 4-year-old can do

  • The human genome is 3.31GBP of DNA, or about 800MB

  • 800MB is not enough? much more happening in a cell

  • the genome has learning algorithms, learning how to see

  • 10^11 cells in the human brain, other cells in the brain, besides nerve cells

  • emergent complexity during growth, if you cut tendons in birds, wings don't grow

  • what is the difference between instinct and knowledge?

  • algorthims for figuring out how to figure out how to wire the neurons

  • monkey neurons speak a "set" language?

  • The axioms of mathematics is about 30, but the axioms we need is not a finite number

  • neurons are computationally complete, the fact that humans can do arithmetic proves humans have all the capabilities of a computer, humans natural inclination is not for doing integrals

  • the emotions of crossing the street

  • what's the algorithm for open-ended learning?

  • do horses dream of walking? do babies dream of the equations of inverted pendulums?

  • snowboarding and big wave surfing

  • humans are intelligent because we are inbred?

  • the brain is a good use of energy (for some people), brain uses a lot of calories, it has to pay off, does for humans and dolphins

  • what makes humans different from other primates? is it just brain size or is it cellular?

  • the information in the fertilized cell -- how much? vs the DNA? There is information in the nucleus not in the DNA

  • Von Neuman machines is just a set of operations and a stinkin load of memory -- what makes a blank slate that thinks?

  • are the materials in the factory (the lipids and proteins) the same as the blueprints for the factory?

  • Conway's Game Of Life

  • with DNA you can't "run" it without everything that goes with it

The link is: http://www.archive.org/details/BoulderFutureSalon-ArtificialIntelligence


(C) 2007 Boulder Future Salon and the Acceleration Studies Foundation.