Printed: 2019-06-16

Institute for Ethics and Emerging Technologies





IEET Link: https://ieet.org/index.php/IEET/more/Brin20170909

How did humans “overshoot” in intelligence? Plus a Tech Roundup!

David Brin


Contrary Brin


http://davidbrin.blogspot.ca/2017/08/how-did-humans-overshoot-in.html

September 09, 2017

Phew. The news cycle has slowed enough to get our heartbeats down just a tad.  So how about some stimulus on the positive-hopeful side? You are still a member of a spectacular, scientific civilization. The War on Science (and all other fact-using professions) will not succeed if we keep our spirits up. So let’s roll up our sleeves and dive into some amazing stuff.

But our main feature this time? A riff on why humans may have shot way beyond "threshold" levels of intelligence.

== How did we get so smart? And what does it imply? ===
 
How did we evolve intelligence? In a classic paper, Nicholas Bostrom (author of Superintelligence: Paths, Dangers, Strategies) and Carl Shulman appraised the Hard Intelligence problem with some panache. Yet they left out a crucial observation — (I believe I am one of the few to state it explicitly) — called the Glass Ceiling Effect.  That the rise to sapience is blatantly non-linear.
 
We see dozens of species that cluster near quite similar levels of threshold semantic ability and basic tool use.  A level that (in my novels) I have long called “pre-sapience.”  This clustering, which includes several diverse mammals (apes, elephants, sea lions and possibly prairie dogs), birds (corvids, parrots), and perhaps even cephalopods, implies two things:
 
1- that rising to this level is relatively easy and
2- moving beyond it is - for some reason - very hard.
 
#1 and #2 seem to be logically derived from observation. There are added implications, though. The stunning degree to which we crashed through and moved beyond this ceiling — leaping many orders of magnitude in semantics, tool use and other realms — suggests either:
 
3- moving through the ceiling, while rare, opens up whole realms of mentation in a nonlinear way.
 
 Or else
 
4-  the opposite. Reaping any benefit from sapience requires a species to continue evolving very rapidly, moving to a distant and very difficult plateau, or its fitness profile will collapse back to pre-sapience. Alas, #4 fits the facts better than #3, by far.
 
Either way, the implications are that human sapience is likely to be rare.
 
But there’s more. Consider dogs and goats.  As soon as humans developed a 100 word vocabulary and fire and stone spears, we were the top predators - especially after we got a partnership with dogs - and we could then defend goat herds, which proliferated to denude vast stretches of land, causing deserts to spread long before agriculture. Hence, we were already wrecking the planet at that level of borderline effective sapience. (Of course, irrigated agriculture then spread deserts even more dramatically.)
 
We might now save the Earth!  (See my novel Earth.) But only because we leaped ahead to be capable of ecological science just 10,000 years after we began herding goats. That’s an evolutionary eye-blink, so rapid that much of the planet is still in decent shape!
 
Picture a species that crosses this gulf more conventionally, or more slowly. Then by the time they get smart enough to understand ecology, it’s already too late. Their world is too impoverished to support a major, industrial civilization, capable of spaceflight.
 
In other words, our non-linear leap from threshold pre-sapience to interplanetary tech and ecological management might affect the Fermi Paradox in two ways. First, it happens only rarely, and second, it must be very non-linear, almost exponential in order to leave the species with adequate resources to expand.
 
What might be a mechanism for this non-linear leap? Roger Penrose’s hypotheses merit some scrutiny here. Is it possible that this exponential nonlinearity of mental growth happens because we  reached a threshold, where new modes became possible, suddenly?  Perhaps Penrose’s quantum effects in the brain. Or else dramatic leaps in available software. (The latter is my own theory about this… successful software reprogramming revolutions, 100,000, then 40,000, then 15,000, 6000, 2000, and 250 years ago.  I describe this in Existence.)
 
Furthermore, Nick Bostrom’s speculations about numbers of neurons winds up being quaint and irrelevant if you ponder recent discoveries about intracellular and inter-cellular computation, which suggest levels of computability many, many orders of magnitude greater than mere synapses.
 
This suggests that sapience may not be as common as we assumed. Simon Conway-Morris of Cambridge - author of Life’s Solution: Inevitable Humans in a Lonely Universe - is the accepted maven of Convergent evolution. But he has lately retracted his earlier stance that full sapience will naturally and convergently evolve.
 
Finally, there is Nick’s attempt to draw conclusions about the difficulty of artificial intelligence… These  I shrugged aside as tendentious leaps without much justification. In my talk on the future of A.I. to a packed house at IBM's World of Watson congress in Las Vegas, October 2016, I presented a tour of big perspectives on Intelligence, as well as both artificial and human augmentation.

And yes, this will be nonlinear, as well.
 
== Tools and more tools ==

Stanford’s new, four-layer 3D-chip design replaces silicon with carbon nanotubes (sheets of 2-D graphene formed into nanocylinders). The top layer has sensors, then resistive random-access memory (RRAM) cells. Then two logic layers. Three-dimensional integration is the most promising approach to continue the technology-scaling path set forth by Moore’s law, allowing an increasing number of devices to be integrated per unit volume…
 
…though in fact, Moore’s law is collapsing in what I call the Big Flip, as the last 50 years of advancement in computational hardware slows down to its long-awaited S-Curve… but progress in software (which had been glacial) seems to have taken off spectacularly - especially in Learning Systems - in just the last couple of years.
 
A team of researchers at Michigan State University has developed a new type of solar concentrator that when placed over a window creates solar energy while allowing people to actually see through the window.
 
Pharmacy on Demand: DARPA has a really neat project in battlefield medicine for a small (dorm fridge size) synthesizer that can produce pretty much any basic pharmacy drug (chemical) to GMP levels using base ingredients.
 
Their next project which is well underway is a biologicals machine (enzymes, mRNAs, etc.) In developing a flexible, miniaturized synthesis and manufacturing platform, Battlefield Medicine will lead to distributed, on-demand small-batch pharmaceutical production in austere environments. 
 
It correlates with other advances such as the recent Qualcomm "Tricorder XPrize" which advanced the capability of hand carried disease diagnostic systems.  http://tricorder.xprize.org/  Many pieces are coming together at the same time.
 
A “living” programmable “ribocomputing” device based on networks of precisely designed, self-assembling synthetic RNAs (ribonucleic acid). The RNAs can sense multiple biosignals and make logical decisions to control protein production with high precision. The research was performed with E. coli bacteria, which regulate the expression of a fluorescent (glowing) reporter protein when the bacteria encounter a specific complex set of intra-cellular stimuli. But the researchers believe ribocomputing devices can work with other host organisms or in extracellular settings. What could go wrong?
 
Will we see the return of storing bulk data on… magnetic tape? Oh, but at the recent Science Foo Camp (on the Google Campus) George Church told us about the near feasibility of storing all the world's books in a cup of DNA... (Also ask him about resurrecting mammoths!)
 
All around the world, scientists are building repositories of everything from seeds to ice to mammal milk — racing to preserve a natural order that is fast disappearing.  Both disturbing and reassuring in some ways… though I admit some pique that the description of “life arks” in EARTH (1989) gets no mention.
 
Warnings: Finding Cassandras to Stop Catastrophes, a new book by Richard A. Clarke & R.P. Eddy, offers insights into how we can weigh predictions, especially when it comes to national security, threatening technologies, the U.S. economy, and possibly the fate of civilization. In Greek mythology Cassandra foresaw calamities, but was cursed by the gods to be ignored. Modern-day Cassandras predicted the disasters of Katrina, Fukushima, the Great Recession, the rise of ISIS, and many more. Like the mythological Cassandra, they were ignored. There are others right now warning of impending disasters, but as Ray Kurzweil asks: “how do we know which warnings are likely to be right? … Clarke’s and Eddy’s penetrating insights are essential for any person, any business, or any government that doesn’t want to be a blind victim of tomorrow’s catastrophe.”  
 
Alas, short-shrift is given to the truest font of such alarums… hard, high-level science fiction.
 
== And some setbacks ==
 
Rep. Lamar Smith (R-TX) -- who is bafflingly chair of the House Science Committee -- penned an op-ed praising the “benefits” of climate change. He goes so far as to celebrate carbon dioxide emissions for melting Arctic ice to allow for “new commercial shipping lanes”. They've spent years telling us climate change wasn’t happening, then human-generated CO2 wasn't responsible... and now it’s suddenly a good thing?' Sorry 314 guys. This has been going on for years. 

These towering hypocrites shift the goal posts with stunning agility. Vast farms in Canada will replace those lost to desert in Mexico and Texas! (Oops, there's no topsoil up there, and even if things warm enough for crops, there'll just be one, short growing season, to replace two long ones, down south.) And thawing tundra will pour gigatons of methane into the atmosphere. But this is the sort of raving monster the GOP puts in charge of the Science Committee. And even if your crazed uncle is beyond reach, maybe his wife isn't. Go have coffee with your aunt. 

He may relish the end of the world; she'll frown and worry about her grandchildren.
 
Finally....
 
Here's a good article on METI - the rash cult wanting to send "messages" to aliens - and the response of a dozen SETI thought-leaders, including myself, asking for discussion.
 
Not that I’m unhappy with how things turned out… but where was this “sapiosexuality” movement, back when I was a frustrated student at Caltech? hm? Well, it's a new and better and wiser generation.

David Brin Ph.D. is a scientist and best-selling author whose future-oriented novels include Earth, The Postman, and Hugo Award winners Startide Rising and The Uplift War. David's newest novel - Existence - is now available, published by Tor Books."

Newsletter: http://ieet.org/mailman/listinfo/ieet-announce

Contact: Executive Director, Dr. James J. Hughes,
IEET, 35 Harbor Point Blvd, #404, Boston, MA 02125-3242 USA
Email: director@ieet.org
phone: 860-428-1837