IEET > GlobalDemocracySecurity > Vision > Interns > HealthLongevity > Ben Scarlato > Futurism > SciTech
Day 2 Afternoon Liveblogging H+ Summit: Ray Kurzweil
Ben Scarlato   Jun 13, 2010   Ethical Technology  

Ray Kurzweil didn’t cover much new this afternoon. As one wag said on Twitter, “Shouldn’t Ray’s five year-old stump speech be 10,000 times more interesting and only five minutes long?” But Ben does his best to summarize. By the way, check out Kurzweil, J. Hughes, and a cast of thousands in this New York Times story on the Singularity University - J.

Ray Kurzweil is the author of The Singularity is Near among other books and accomplishments.

First, Alex Lightman introduces Kurzweil and lists several of his many honors and accomplishments.

Kurzweil say he’s frustrated he couldn’t be there for more of the conference, and that he wants to get through his presentation quickly so there’ll be time for questions.

He recently gave a talk for SETI. If you take the standard SETI assumption that are hundreds or thousands of civilizations advanced enough to have things like radio assumption, that would mean there would also be many civilizations doing galaxy wide engineering.

Since we don’t see that, Kurzweil says we’re actually in the lead.

The negative findings of SETI is significant because it gives us an even greater responsibility than stewardship of our planet: we’re responsible for the way intelligence emerges in our universe.

The Law of Accelerating Returns means the rate at which we adapt technologies gets faster and faster. The paradigm shift rate is getting faster and faster.

The linear expectation is what our intuition is, which is why half-way through the genome project it was called a failure because it’d only completed 1% of the genome. But it doubled every year, and finished a year ahead of schedule.

Amazingly, it wasn’t an approximate doubling. It was a precise doubling. These things are incredibly predictable.

There were two key innovations in homo sapiens, one was the large neocortex which enabled tool and language. Another was human directed technological cultural innovation.

Again, rates of adoption are increasing. Think back to 3 years ago, most people didn’t use social networks, let alone Tweets. In 2004 Facebook was invented. Both Facebook and Google emerged from Google dorms.

One criticism of Kurzweil’s ideas is that exponential curves eventually stop increasing. However, he says that’s only true for a specific paradigm. Computers have already switched paradigms 5 times. Eventually, we’ll go to the 6th paradigm to a kind of 3D computing.

Kurzweil emphasizes how predictable his graphs are, there’s no disruption from any of the world wars or anything.

Technological acceleration is very predictable at a high level, just as the properties of gases are predictable at a high level due to things like the law of large numbers. But you couldn’t predict, say, which company would succeed in a new technology area.

Some economists have expressed concern that as more of the economy becomes information based, there’ll be a contraction of the universe. However, we more than double than our consumption of information technology each year, because new applications become available.

We’re now doing amazing things like simulating protein folding. This is the bridge 2 he talks about in his health books, reprogramming our biology.

Kurzweil’s phone is probably updating right now, but his body’s software hasn’t been updated in thousands of years.

Now onto his new book, Reverse Engineering the Brain: the Ultimate Source of the Templates of Intelligence.

There’s a lot of new theories about consciousness, such as Penrose’s quantum theories of consciousness. Kurzweil says consciousness is a bit mysterious.

He says there was an assumption that because both quantum mechanics and consciousness were mysterious, they must be linked.

There really is no scientific experiment we could imagine that would definitively define conscious. The Turing Test only determines whether someone seems conscious.

A lot of things are growing exponentially, including non-invasive brain imaging and the amount of information we’re gathering about the brain.

There are some pretty slides of cerebral cortex simulation images.

One of his critics, John Horgan, wrote a paper saying we could never simulate the human brain because it’d take a quadrillion lines of code.

Horgan used a picture to illustrate this. Kurzweil did some research, and it turns out it wasn’t a real picture; it was actually a simulation. He says the significance of this is lost on some audiences.

Kurzweil asks where are these quadrillion lines of code? Everything in the human brain is encoded by the human genome, which is a lot smaller.

He says it’d be at most be half a million lines of code. But how is that possible with trillions of neuronal connections? Because there’s a lot of redundancy.

The brain is a probabilistic recursive fractal. That doesn’t mean it’s simple, but it’s a level of complexity that we can deal with.

Exponential growth doesn’t mean instantaneous, in fact it can be sub-linear. So Wall Street discounted exponential growth after the Dot Com boom.

We have 10 thousand times more energy than we need coming from the sun; it’s just in the wrong form. But we’re applying nanotech to change this, and it’s following an exponential trend.

It’s only 8 doublings at 2 years each from meeting 100% of the world’s energy needs. The reason we don’t see this is that exponential growth is always ignored when it’s low, but the whole point is it starts to grow much faster as it gets bigger.

He quickly goes through some slides of predictions up to 2029.

We’re going to become hybrids of machines and nano-biotechnology.

He’s finished, and the crowd gives a standing ovation. Now for a Q&A with Alex Lightman.

Ben Scarlato, a former IEET intern, studied Computer Science at Rochester Institute of Technology and works as a software engineer focused on security.


2010: “Supercomputers will have the same raw power as human brains (although not yet the equivalently flexible software).”
ill have the same raw power as human brains (although not yet the equivalently flexible software).”

This assertion raises some difficulties. It is hard to say just what “raw power” and “equivalently flexible software” are.

My interpretation of the oracle is: “Supercomputers will be as smart as humans, excepting that they probably will not be.”

Will videos of the H+ summit be available at some point ?


Yes, Humanity+ said they will be releasing the videos over time.

Google is trying to build Strong AI, aka ‘Skynet’. There is no debate or refutation to this fact:

YOUR COMMENT Login or Register to post a comment.

Next entry: Day 2 Afternoon Liveblogging H Summit

Previous entry: Martine Rothblatt’s Lifenaut Project Featured in New Scientist