Continued from here.
**The unusual warning.
Life was busy for me* in these days. I was being theoretician for the 5th Generation Warfare efforts, wrote many initial proposals for individual Psyop campaigns under my Generalissimo hat. At the same time, I was working on the design elements for the next-generation civilization, and also coordinating the research on the Tessel technologies for my new company, Tessellated Genetics.
The animal work on the technology of Tessels proceeded rapidly.
Of course it did! I have always been astonished at people who exclaimed astonishment. Nobody in the Status Quo ever wants to learn anything from the world’s many skunkworks, but they produce cheap innovations. Get the best team, tell them to do airplanes or spaceships or decode the human genome, give them resources and stand out of the way. That can’t be scaled : as soon as you start budgeting and planning, forget it, your system is of a size and complexity that is also easily gamed. Organizations have bugs just like software.
Part of the problem of scaling is the payoffs. One of 2016’s more profound practical philosophers was Nicholas Taleb, who discussed risk/return for financial instruments and careers. Dentists do well every year, stock traders many years, futures traders less often, perhaps never. And, as “Big Short” shows, once in a while, futures trading can do very well indeed.
Big Short shows 2 sides of a very large and intricate game, also an evolutionary system. It shows periphery vs center, rates of evolution. Remember the bit about where the big bank, can’t remember which, needed to sell puts on the CDOs? And the only buyers were little guys (maybe a bank in Germany also) who were the seriously systematic and independent thinkers? That is different rates of evolution. The losses in the banking industry, and for taxpayers, were the center losing one of the games. So far, both banks and taxpayers have learned the wrong lesson, I thought. It isn’t only species that go extinct, imperial presidencies are on the wane.
This same line of evolutionary thinking, I realized, is why nothing has worked so well as laissez-faire-with-skin-in-the-whole-game. In world history so far, all other forms of organizing systems have resulted in economic failure and war, as this one shows obvious signs of doing.
Another aspect of scaling has been that ‘large’ == ‘gameable’. Systems themselves are always political at some level, that is tolerable so long as reputations inside and outside were relatively in synch, and your actions are local enough for effects to affect your reputation. Where you have skin in the game in several of the dimensions of your life, and where your reputation follows you through those dimensions and affects your future. ‘Open’ systems, transparent systems, local-enough systems where many eyes scrutinize every fact, and indications of wrong-doing propagate to levels able to appropriately influence the persons doing them.
For people to take more than $ and career into account, they have to be players in and on a local stage, skin in the whole game, lives in the community where their reputation is important for future happiness, where you can’t substitute a power relationship for genuine good will.
That is the part of civilization that hasn’t scaled. The best institutions do not run on power relationships, but more and more do. Thus, the declines and failures. And thus, the need to kill off power relationships along with the excess male sex drive that produced them. I didn’t think there would be no more power relationships, of course, just of the ones where people have no skin in the games they can play, where they have non-local effects beyond retribution.
Our local project was an example of how an efficient, voluntary system can work. Different payoff sizes and probabilities attract different kinds of personalities and require different kinds of resources. Ours was a huge payoff, personally in scientific reputation and satisfaction in being on the leading edge of something new, and possibly commercially. The total resources required were an animal lab and ability to extract fertilized eggs before they implant, raise the eggs to blastula-stage in tissue culture, inject blastula-stage animals with iPSCs from a clone-line tissue culture converted to iPSC, implant the resulting mosaic in the uterus of the foster mother, raise mice to 60 days and rats to 90, behavioral testing of many kinds, followed by many other tests and analyses, biochem to electron microscopy used for 3D reconstruction of 1 cubic millimeter of cortex.
Turns out, those aren’t bleeding edge of modern lab science, so not a high bar, especially as the people who liked that kind of high-risk, high-impact project tended to know each other in a research community and the openness of the project made it easy to collaborate. There were a lot of collaborations.
The Jackson Laboratory had taught everyone about raising mice in the lab, and mice breed at 6-8 weeks, it takes perhaps a month to get behavioral data and that can go on as long as the animal is alive, and all this is cheap. Almost every lab has some excess capacity in some part of their operations. Collaborations made those available for this research.
We were doing 2 things to accelerate this program, both achieved by open online lab books. First, we wanted to scale this research across as many labs as possible. We did that by selling research materials and training. We started that effort as soon as we were confident of our own results, all from the initial half dozen labs I had convinced to collaborate. We made $ out of that, and told everyone we were. Also, that all our profits went into this research, which they did.
Second, we were going for fast turn around in the experiments. That meant an N of 8 animals in each group, the size necessary to show an effect of a size that would interest us, and multiples of that size were what most of those labs could slip into their work using free resources. That would allow every lab to test many variations of parents, but have a reasonable check on measures for each combination.
We had learned from the problems of replications in science exposed by Ioannidis, so each lab replicated someone else’s test to get started, and if the results were not the same, both labs, and often another, would do it again to see what was wrong. Rats were a longer schedule, so we tended to skip some of the combinations and make up for that with more behavioral testing of the more intelligent and longer-lived brains.
Further, every animal got attention from specialists, because juveniles were created in one lab, tested in Skinner Boxes or mazes in another, and passed to biochemistry, physiology, or neurophysiology labs, then on to anatomy labs. For the behavioral testing, often the entire lab was automated so a mouse or rat could be shepherded from home cage through walkways to a testing apparatus, be put through the tests, and then shepherded back home. All the animals were chipped, so following them lab to lab and collating the data in our open database was easy, mostly right from the lab to our database via the local control system. Everyone kept local data so we had multiple levels of backups and checks for the inevitable bugs. Much of this checking was automated independently, so our checks were checked.
The behavioral work showed positive effects, large positive effects, and the other labs started gearing up for the analysis. That part was slower, and the first 2 stages overran the rest for a while.
All of the other labs needed technicians galore, training them was a major part of the scaling effort. Videos and virtual classes handled some of this, but many of the labs did training of some kind, reciprocal favors most often, tho some got back their investments in developing new techniques via training fees.
Nevertheless, it often took several months to train a technician in how to extract a brain and prepare it for anatomy : you have to know how to handle the animal, give it an intra-peritoneal injection of nembutal, cut open the chest cavity, clip the right atrium to let it bleed, and insert the large needle into the left ventricle, open first the saline flow to flush the animal’s blood, then the preservative, often formaldehyde, nasty stuff, so this should be done under a hood. Then cut the animals head off with the handy guillotine, and spend 30 minutes carefully picking the brain out of the skull, putting it into a jar, labeled correctly with the animals’ number, verified by the chip under their skin that has followed them since birth. If details of cranial nerves, spinal nerves, autonomic nervous system ganglia were also needed, that could take a day of work under a low-power microscope.
Biology, physiology, anatomy labs are full of jobs like that. For anatomy, the brain is put into a block of some plastic or paraffin or ice, cut very- to exceedingly-thin, depending on whether light or transmission electron microscope is to be used, stained, mounted on glass slides for light microscopy and imaged. Electron microscopy needing ultra-microtomes producing 100 micron sections and requiring heavy metal staining and counterstaining is a lot more work.
Having slides or images, somebody may sit and count particular types of cells or dendrites or axons or …, all little specs it takes a long time to learn to recognize reliably, in a systematic survey of brain tissue. Handling the tissue has largely been automated for routine work, but much of this research was not routine. Depending on the study, it could take many weeks per brain, although the automated analyses could identify most cell types in light microscopy and synapse types in electron.
For for recording the electrical activity of a single neuron, glass pipettes pulled very thin and filled with saline were needed for intra-cellular recordings. Tungsten electrodes were sufficient for extracellular, are sputter sharpened by dipping them in and out of mercury baths under oil with the minimum sparking current through them, then coated with lacquer. Recording neurons requires very good amplifiers and a very low-noise environment, 50 micro- to small milli-volt signals through 10 megOhm glass or 1-2 megOhm tungsten. Lower resistance than that, the tips weren’t sharp enough to penetrate cells or isolate single cells from the outside.
Recording EEGs is much easier, as it uses stainless steel wire, beads on the recording end formed by melting via an electrical current, leaving the insulation behind the bead. These could be placed individually on an animal’s exposed brain, or implanted under the dura mater that covers the brain. Often, for recording from a set of such electrodes, we would use a small piece of soft, thin plastic to hold them, put the entire set on the brains surface in one operation.
For electrophysiology record of single cells or deep in the brain, the animal is anesthetized and electrodes guided to particular places in the brain using a stereotaxic instrument and atlas of the mouse, rat or cat brain. As with the EEGs, those electrodes can be implanted and recordings made with the animal free.
All that and the very many such processes involve techs and grad students and post docs and junior professors. By mid-level, most scientists are writing proposals and papers and supervising research. Here we had another major advantage over standard research projects, because our labs skipped the funding proposals. This was easy, as some labs had excess capacity for generating the Tessels, others for behavior, etc. So the proposal stage was “I can do this part of an experiment, what do people have to offer?”, pick collaborators and go.
Our senior scientists were spending their time looking at each other’s data hot from the lab. These data points all got much attention, and every experiment generated a dozen more. The ideas outran the lab’s combined capacity, and memes were in competition, ideal for evolutionary progress of all kinds. Labs did many experiments in parallel, the game of predicting other experimental outcomes as the results of each came in was a major source of ideas.
We soon found that the brains of our multi-parent hybrids are different, and the most intelligent are quite different.
In psyops, the message is the op.
*Generalissimo Grand Strategy, Intelligence Analysis and Psyops, First Volunteer Panzer Psyops Corp. Cleverly Gently Martial In Spirit
**The unusual warning is not written yet, as it requires a spark I have not yet encountered.
The usual warning***, however, is that propaganda is bad for mental agility, grasping the reality behind the layers of facade.
And that this site wants to seize your mind, just as do all others. In fact, this site is far more dangerous to your current viewpoints than other sites because it uses very advanced techniques to convince you, as much honest, full-context discussion from as many points of view as possible, using impeccable logic and pointing out all uncertainties, as well as the fact that the world is full of unknown unknown unknowns. Possibly raised to the Nth power unknown unknowns.
And, that warnings are useless, they disarm you. And that propaganda about propaganda is still propaganda, and you shouldn’t let anything into your mind without careful inspection from many points of view.
I have written much better warnings, do read several to get a better measure of the insidious insidiousness of ideas I am trying to put into reader’s heads.
***The difference between my usual warnings and unusual warnings is that the propaganda elements embedded in unusual warnings are explained, usually in ways that are true and also elegant propaganda devices themselves. Which then themselves are exposed and explained, … Recursion and meta are my thing.
At this point, I had an idea. The difference between usual and unusual is whether I add one more statement.
Which type do you think this is?
I thought that was cool as hell.
You aren’t being careful enough.