talk sense into the public; but Americans don't listen, probably because they they can't bear the hint that someone might be smarter or kinder; they treated the Clintons like dirt and dumped their economic policies at the first opportunity in favor of a Keynesian hogfest.

And there are admirable people all over the place, doing good works and living responsibly, who get ignored in order that shallow Hollywood rubes get all the press. Please.

First, the "software" ain't trivial. I'll grant we'll be *closer* to that goal. The devil's in the details. It's real easy to wave your hands and say "dna algorithms".

Second, the hardware will *not* be up to par by then. Moore's law is dead; long live diminishing returns. We're going parallel (google CUDA) as a way of dealing with that, but one of the fundamental problems with the parallelism currently implemented in hardware (google "grid computing") is that getting the vast quantities of data to our super-de-duper fast processors is the bottleneck. The human brain, with its huge number of crappy processors, naturally handles the wide-band input provided by our senses. A semiconductor brain-in-the-box is going to have to bruteforce a lot of the work, which will suck away efficiency.

I'd say simulation/replication of the human brain is going to have to wait on nanotech assemblers who can also function as computation units. At that point we'll be able to just recreate the exact topology and function of the brain (including fake hormones, serotonin, blood sugar effects, etc), and it ought to work. 50 years minimum. Probably more like 200.

What does this guy get paid to do? What the hell is a futurist? Are there any objective criteria to guage the accuracy and efficacy of a "futurist?" Must be nice to just wax poetic about technology and the future - and get paid to do it!

No, we will not have computers able to recreate the subtlety and suppleness of the human brain by 2020, whatever the hell that means (does the phrase "It's so bad, it's not even wrong." come to mind?).

What is going to happen is stem-cell research is being developed to grow human brains. The problem lies not in the progression of Moore's Law, or the genetic algorithms implemented on silicon, or the dialectric k constant. No, the problem lies in the interface. How do we interface with a human brain? The logical, sequential, deterministic algorithms of computers coupled to the intuition and massive parallelism of organic tissue. THAT is the bottleneck.

I hope Salon keeps this video in archive. Show it to engineering and comp. sci. classes in the year 2020, so they can write thesis papers on the failure, yet again, of the promises of AI.

And just one more thing - the down side, the negative consequences, of this technology and computational power are not going to be talked about in the mainstream media. Because it will be part of a zero-sum game equation: the corporations gain, the average citizen loses. All this technology will be used for profit and power, and the average citizen, human being, is going to become even more of a commodity, even more of a data mining signature, even more of an actuarial projection. Don't even think it will benefit your quality of life, unless you consider being fat, lazy, and stupid, and having every sensual desire met, while your heart and your spirit is starving for something that is real, is considered qaulity.

This guy did not say anything that was not said in 1955.