Singularity Summit 10 Mid-Day

By | August 14, 2010

12:30 Ben Goertzel

Ben introduced as the “badass” of the AI community.

The human body is like machine. Body produces and processes a huge amount of data. Hard to address the complexity. A drug typically goes after one target, not all the interrelationships. SENS for example may fall short grasping the complexity.

Move mitochondrial DNA into the nucleus — prevents damage, but what else results?

Genescient’s flies live 5x normal fruit flies. Got there by intensive breeding, focused on longevity. Superflies! Stronger, more sex, better immune systems. Sexy immortal billionaire insects.

Isolated genes that express differently in superflies — hundreds of them. Many of these have analogs to human genes. Traditional statistics revealed a lot about superfly data. How can AI tools reveal more? One day AGI biologists will put humans out of existence. Meanwhile, narrow AI can help. Text analysis — can map and reveal relationships.

Certain genes have been identified as key. Mapped supplements which trigger proteins related to this key genes. Includes zinc, vitamin E, reverstrol.

But current AI tech just scratches the surface.

Combine early stage AGI with powerful narrow AI to get the artificial biologist.

12:00 Ray Kurzweil

The Mind and How to Build One

(Ray is “here” via video conference.)

The brain is not magic. The complexity of the brain is not inherently beyond our measn to comprehend.

Speech recognition, visual proecessing, many limited Ai applications have had a big boost from emulating the appropriate

brain functions. Take the basic principles and focus them, amplify them.

Exponential progression shows that we will be able to get to 10-14th / 10-16th calculations per second. Justin Rattiner of

Intel says 3D chips will take off where stand silicon chips leave off. Moore’s Law will continue. The exponential growth will get us to human levels.

These ideas are highly contested by mainstream thinkers / media.

Singularitarian — someone who understands or accepts the fact that information technology progresses exponentially, which

ultimately leads to something transformative. Education, experience, qualifications — none of these are necessarily indicators of the likeliehood that someone will be a Singularitarian.

Intelligence is the most progound phenomenon in biology. We have already amplified intelligence with our technology. Homo

sapiens were a big step forward with our big heads. But now we don’t need a bigger brain case — we can move to a new substrate where the old limitations don’t apply.

Shows the Moore’s Law chart — his expanded take on it — updated from the Singularity Is Near. The progress is very smooth.

One transistor for a dollar in 1968, a billion for a dollar today. Different technologies — eg, biotech — now growing exponentially because of information technology. Exponential trends start out slow and then produce huge returns unexpectedly.

We literally create our brains by the thoughts we have. Thoughts seem ephemeral but they create the physical structure of the brain.

A perfect simulation of the brain won’t do anything until it learns something.

The cortex is the only region that allows us to think in hierarchies. A symbol for and idea can be combines with a symbol for another idea to create a new idea with its own symbol. This sums up to what we call knowledge.

Complex concepts are organized as individual lists in a particular order. LISP mania in the 70′s and early 80′s was a foreshadowing of the dot-com bubble. Bust cycles are precursors to profound developments. Dot-comm bubble leads to multi-billion dollars companies (Google.) Our understanding of the cerebral cortex shows us now that it is a LISP processor. As was asserted in the 70′s and 80′s, this is how the mind works — generating one-dimension lists that can be embedded to encode any level of sophistication.

It’s hard to have a serious discussion about consciousness becasue of a conceptual gap between the objective and the subjective. There’s no evidence provided that consciousness lies anywhere that it is speculated to be. You can’t do a falsifiable experiment to isolate consciousness. Ultimately it’s a lep of faith.

But, “if it qaucks like a duck…”