The Pace of Innovation Never Falters

Notebooks

The Pace of Innovation Never Falters

(Page 2 of 2)

March/April 2010
By Steve Jurvetson

Onwards and upwards: In this abstraction of Moore’s Law, Ray Kurzweil plots computational power on a logarithmic scale, and finds a double exponential curve that holds over 100 years (a straight line would represent a geometrically compounding curve of progress).
Ray Kurzweil

Transforming Energy

816 Days Ago
03/01/2010

A bit skeptical of the "fastest" claims

I'm not so sure your computer will outperform conventional computers at any and all tasks.

It is my understanding that a quantum computer is essentially a classical machine with a quantum core. While the laws of superposition might be great for performing computational tasks in the realms of cryptography, integral transforms (signal or video processing), or even direct quantum mechanical modeling, I highly doubt it would be useful for more "mundane" tasks such as weather modeling, logistical optmization, or even database management/search.

In other words, you may well soon own the best RSA code cracker or prime factorizer, but it will likely be a one- or two-trick pony.

Oh, and FYI, Moore's law predicts a minimum cost of integrated circuit density that doubles in every 24 months, not in 12.

813 Days Ago
03/04/2010

The original Moore's Law paper (1965) predicted an annual doubling of transistor counts in the most cost effective chip, and he revised it in 1975 to every 24 months. With a little hand waving, most reports attribute 18 months to Moore’s Law, but there is quite a bit of variability. The popular perception of Moore’s Law is that computer chips are compounding in their complexity at near constant per unit cost. This is one of the many abstractions of Moore’s Law, and it relates to the compounding of transistor density in two dimensions. Others relate to speed (the signals have less distance to travel) and computational power (speed x density). Kurzweil would argue that the original prediction was correct over the very long term. More info: http://www.flickr.com/photos/jurvetson/3656849977/

As for quantum computing, the original canonical application was factoring integers, but D-Wave has a different adiabatic architecture that is application-specific to logistical modeling and optimization problems in general (traveling salesman route optimization, discrete variable machine learning, etc.). It would be a co-processor to traditional machines, and by no means would I claim that it would useful for all computation. Think of it as expanding the bounds of computation where we cannot compute today. Here's an example from Google:

http://www.flickr.com/photos/jurvetson/4171280876/

Wish I had more room to include details like David Deutsch's NYT observation: “Quantum computers have the potential to solve problems that would take a classical computer longer than the age of the universe.” from http://www.flickr.com/photos/jurvetson/376098497/

813 Days Ago
03/04/2010

quantum leap in logic

Yes, someday something other than silicon will take over Moore's Law and there will likely be a step function non-linearity when that happens. This is pretty straight-forward, will it be 2010, maybe, but I think it will take longer than that (it almost always takes longer than we expect).

The connection to synthetic life being coded in the lab in the same time frame is a classic monumental leap of logic. It reminds me of an article a few years ago about the Swiss supercomputer centre building a big new computer to model the brain, and the journalists talking about the machine waking up, implying that such a machine could achieve some level of consciousness. Maybe they were kidding, but I doubt it and many non-technical folks took it as serious.

The fact that we are beginning to understand how DNA is coded and have learned how to splice characteristics that we like, it's an astronomically long way from being able to program from scratch new DNA. The author should watch again the beginning of "I am Legend". The coding of life is far more complex than any program designed my humankind, and these programs are full of errors, so much so that bug management is treated like the management of a random processes. I agree that life is a product of intelligent design, but that intelligence is greater than the aggregate intelligence of all humankind.

813 Days Ago
03/04/2010

Oh, those are unrelated predictions - there is no connection to the quantum computer prediction.

The first synthetic life form will be from Craig Venter's company Synthetic Genomics. They have created viruses from scratch, and bacterial scale creatures are due any day now, after a bit of a detour form some epigenetic learning. Last year, Venter demonstrated that you can swap 100% of the DNA of two microbial organisms (as genetically different as mice and human) and change them into each other.

And yes, the fundamental limiter on the pace of progress is our understanding of complex systems development. I suspect we will design for evolvability in the near term. George Church has already created 4 billion genetic variants per day. And here are some earlier Tech Review musings on the different approaches: http://jurvetson.blogspot.com/2006/07/dichotomy-of-design-and-evolution.html

Notebooks

The Pace of Innovation Never Falters

Transforming Energy

MAGAZINE

People Power 2.0

The Library of Utopia

A Faster Fourier Transform

People Power 2.0

Turing's Enduring Importance

A New Net

"Tectonic Shifts" in Employment

Turing's Enduring Importance

Kenya's Startup Boom

Foundation Medicine: Personalizing Cancer Drugs

Videos

Consumer-Driven Disruptions

Newsletters

Follow Us