Genescient has NOT paralleled human mating habits that would predictably
shorten life. They have only started from a point well beyond anything
achievable in the human population, and gone on from there. Hence, while
their approach may find some interesting things, it is unlikely to find the
things that are now killing our elderly population.


On Tue, Aug 10, 2010 at 11:59 AM, Ben Goertzel <> wrote:

>> I should dredge up and forward past threads with them. There are some
>> flaws in their chain of reasoning, so that it won't be all that simple to
>> sort the few relevant from the many irrelevant mutations. There is both a
>> huge amount of noise, and irrelevant adaptations to their environment and
>> their treatment.
> They have evolved many different populations in parallel, using the same
> fitness criterion.  This provides powerful noise filtering

Multiple measurements improve the S/N ratio by the square root of the number
of measurements. Hence, if they were to develop 100 parallel populations,
they could expect to improve their S/N ratio by 10:1. They haven't done 100
parallel populations, and they need much better than 10:1 improvement to the
S/N ratio.

Of course, this is all aside from the fact that their signal is wrong
because of the different mating habits.

> Even when the relevant mutations are eventually identified, it isn't clear
>> how that will map to usable therapies for the existing population.
> yes, that's a complex matter
>> Further, most of the things that kill us operate WAY too slowly to affect
>> fruit flies, though there are some interesting dual-affecting problems.
> Fruit flies get all the  major ailments that kill people frequently, except
> cancer.  heart disease, neurodegenerative disease, respiratory problems,
> immune problems, etc.

Curiously, the list of conditions that they DO exhibit appears to be the
SAME list as people with reduced body temperatures exhibit. This suggests
simply correcting elderly people's body temperatures as they crash. Then,
where do we go from there?

Note that as you get older, your risk of contracting cancer rises
dramatically - SO dramatically that the odds of you eventually contracting
it are ~100%. Meanwhile, the risks of the other diseases DECREASE as you get
older past a certain age, so if you haven't contracted them by ~80, then you
probably never will contract them.

Scientific American had an article a while back about people in Israel who
are >100 years old. At ~100, your risk of dieing during each following year
DECREASES with further advancing age!!! This strongly suggests some
early-killers, that if you somehow escape them, you can live for quite a
while. Our breeding practices would certainly invite early-killers. Of
course, only a very tiny segment of the population lives to be >100.

> As I have posted in the past, what we have here in the present human
>> population is about the equivalent of a fruit fly population that was bred
>> for the shortest possible lifespan.
> Certainly not.

??? Not what?

> We have those fruit fly populations also, and analysis of their genetics
> refutes your claim ;p ...

Where? References? The last I looked, all they had in addition to their
long-lived groups were uncontrolled control groups, and no groups bred only
from young flies.

In any case, since the sociology of humans is SO much different than that of
fruit flies, and breeding practices interact so much with sociology, e.g.
the bright colorings of birds, beards (that I have commented on before),
etc. In short, I would expect LOTS of mutations from young-bread groups, but
entirely different mutations in people than in fruit flies.

I suspect that there is LOTS more information in the DNA of healthy people
>100 than there is in any population of fruit flies. Perhaps, data from
fruit flies could then be used to reduce the noise from the limited human
population who lives to be >100? Anyway, if someone has thought this whole
thing out, I sure haven't seen it. Sure there is probably lots to be learned
from genetic approaches, but Genescient's approach seems flawed by its

The challenge here is as always. The value of such research to us is VERY
high, yet there is no meaningful funding. If/when an early AI becomes
available to help in such efforts, there simply won't be any money available
to divert it away from "defense" (read that: offense) work.


RSS Feed:
Modify Your Subscription:
Powered by Listbox:

Reply via email to