Eliezer S. Yudkowsky wrote:
> Billy Brown wrote:
> > OTOH, I doubt that such studies will have much relevance to intelligence
> > enhancement. It doesn't look like biotech is moving fast enough for
> Nothing but the FDA and lack of funding prevents me from using existing
> neural stimulation* technology and the Algernic bank shot** to start
> enhancing intelligence tomorrow.
If everything works exactly the way you expect, maybe - but how often does that happen? The problem with making Algernons is that it takes several years to figure out if your first experiment worked, and longer to find out what went wrong. If you go through a few mistake cycles before getting it right, then wait 10 - 15 years for the first kids to get old enough to be helpful, you're looking at the 2030 - 2040 time frame. Plus, the funding and legal problems are likely to delay the whole project substantially.
Meanwhile, research on neuron/microchip interfaces already has solid funding and legal acceptance. Primitive examples exist in the lab, and the cycle time for building improved models is only a couple of years. It appears that prosthetic limbs and sensory organs using such techniques will be in production in a few years. If we get a typical progression curve, I'd expect cybernetic enhancements to exist in the 2020 - 2030 time frame.
But even if the cybernetic approach takes longer, it is still likely to be more important. Algernic enhancement would produce sharply limited enhancement of a few few-defined types. Artificial enhancement is an open-ended process. It probably isn't as fast as pure AI enhancement, but it would advance much faster than a purely biological approach.
Billy Brown, MCSE+I