Oh my!!
I am having my very own personal singularity event NOW:
September 1st, 2017 the day the freak-o-meter went straight up.
I mean intellectually at least how can there be a tomorrow?
I'll try and post before I lose consciousness or perhaps I am transported to a different universe or dimension.
1. Exhibit A for the prosecution.
http://www.biorxiv.o...175406.full.pdf
Go to page 20 of the pdf above or look at the figures attached below. Notice in Figure 3 of the pdf the two panes on the left which have continuous traits and their genotype sample tops out at 1 million. The three lines of interest are those in red in both of these figures. These lines are for intelligence, years of educational attainment and cognitive performance. The two left panes show that at a million samples one would expect 1000-3000 SNPs and these SNPs would explain 40-60% of the variance for these 3 traits.
Now go to page 19. This figure shows the distribution of expected effect sizes (look for the red lines).
So basically we have a nice big normal distribution with perhaps a slight bias to higher IQ, yet this is not clearly shown in the figure.
On first impression this does not really seem to be anything worth writing home about.
After further contemplation one should feel a certain nagging apprehension that something might be amiss.
If this is all just a random process in a normal distribution with nano-scale effect sizes both positive and negative, then this should mean that
the actual samples that are drawn one by one for the thousands of SNPs for IQ will be nowhere close to an upper limit. How likely would it be
to flip 10,000 coins and wind up with 10,000 heads?
How could we all have been so dumb?
The implications of the genetic architecture of intelligence has been missed by almost everyone.
Few if anyone questioned the depiction of Brave New World in which the alphas were not really that special.
From what is clear now the alphas will be very very special indeed.
They will not be somewhat smarter than others, they will be an entirely different species intellectually and physically.
There should be no great concern about interbreeding between them and us.
This would essentially be impossible.
The Bell Curve argument was a fight among fools.
It was about fighting about less than 10 points in most communities.
That is what the basis of class is based on?
Are they kidding?
It is laughable.
The Other Bell Curve (page 19 above) is the real Bell Curve.
It looks exactly the same as the Bell Curve we know and love.
To arrive at someone's IQ we sample from this other Bell Curve.
While the mean might still be 100 and the SD still 15, the broader implication can be more clearly seen.
The top of the range of this Bell Curve is no longer hitting 5100 positives versus 5000 negatives for a
net of 100. The optimized result is hitting 10,000 favorable coin tosses. In actuality the Bell Curve that we
have been arguing would form only a sliver at the center of the Other Bell Curve. Adding up all the pluses and
minuses gives you a slice of reality, though it is like looking through a pin hole. Randomly summing up all these
pluses and minuses basically gives you 0. While the optimal result would be 5000.
We have been drifting on the surface of an ocean of variation without realizing that there is a tidal wave underneath.
Um, oh yeah, so basically at 1 million sample size we already have 2000-3000 of the SNPs identified. That is basically where we already
are with the already announced 1 million Educational Attainment GWAS. So basically what that should mean is that if
we were to CRISPR a few thousand SNPs that are known we should be able to create super ridiculously intelligent people.
Possibly not the optimal, let's call it one third of optimal, so give or take 1000 IQ. Today, Tomorrow, we can engineer
people with 1000 IQ back of envelope calculating.
2. Exhibit B.
Almost the entire conception of the future had been based on an idea that there would be a gene for intelligence, a gene for that.
Even into the 1970s, futurists were talking about a 200 IQ genetically engineered world.
Even when the polygenic nature of IQ was first proposed it still was not recognized what this implied.
Here we are now and we have been bonked on the head and it is still a little foggy.
Basically the highly polygenic nature of IQ and many other traits will mean that intelligence levels will not be increased a slight amount,
they will be able to be increased a MASSIVE amount.
Why didn't all these people writing books ever visit a farm or talk to animal breeders? It has been known for centuries that some traits
(i.e., polygenic ones) can be manipulated to an extreme extent. The example given below is that egg laying frequency can be changed
by a factor of 30 through genetic selection.
This is not a guess. This is completely understood and accepted within the farm breeding community.
The article below notes that von Neumann was perhaps 6 SD above average IQ. However, it is thought using the calculations in the paper that
genetic engineering could produce people with intelligence 100 SD above normal. 1500 IQ roughly? It is not clear to me whether this
includes optimizing the negative effects which I guess would add another 100 SD? 200 SD above normal ? 3000 IQ?
A description of von Neumann's cognitive ability already gives me chills.
And he would in a genetically engineered world be over an order of magnitude below everyone else?
This level of intelligence is scary. It is very very scary.
Anyone think offering kids with a 3000 IQ a hit of LSD would be a good idea now?
https://arxiv.org/pdf/1408.3421.pdf
This observation applies to all the rest of highly polygenic traits. Essentially we have argued endlessly about minor differences in traits for over 100 years.
No one bothered to mention that this has been a largely futile debate about very small differences. The social impact of this observation could be considerable.
Obviously the argument should never have been about the 30 point IQ difference that separates us from us (we from we), but the 3000 IQ point that will separate
us from them (we from them).
It will take no great perception to recognize the eugenica. With an IQ of 3000, a height of 25 feet, a perfect eideic memory, ... , they should be reasonably conspicuous.
3. Exhibit C.
In the figures above, the sample sizes topped out at 1 million.
At 1 million, they were already moving to 50% explanation of IQ variation.
1 million seemed like a very big number.
Sample size of 1 million for Educational Attainment is already in the publication pipe.
http://programme.exo...esentation/214/
Interestingly, Ancestry.com has a DNA database of 5 million growing at 1 million per quarter,
23andme has a 2 million person database, FT DNA has a 500,000 person DNA database ....
Apparently these databases are small potatoes.
The US government's NDIS DNA database has 15.5 million criminal DNA samples (5% of the US population).
(Is that even constitutional?)
The UK tops that it, it has a DNA database of 7% of its population (5 million people).
{I would not feel comfortable living somewhere in which DNA testing was that widespread and so coercive.
There are fundamental questions of human rights that are being pushed up against.}
http://www.nanalyze....t-dna-database/
It should not be unexpected that the criminal justice system has very sophisticated psychometric profiling within their prisons
so that highly accurate phenotypes should exist for most of these prisoners. Furthermore, using this resource for the benefit
of the taxpayers who financially support such institutions does not seem unreasonable. Identifying information would not be
necessary. Research has found that the high and low ends of the IQ distribution mostly draw on the same genetic variants.
Thus, a prison population would still provide very important information about the genetics of the entire range of IQs in the
community. Leveraging this database for the good of the community would be very difficult to argue against. The level of
political support for such an initiative would likely be overwhelming.
Edited by mag1, 01 September 2017 - 10:58 PM.