Saturday, February 06, 2016

Big questions for the future

The Big Questions Contests aim to expand Quora’s already marvelous system for posing quandaries and getting fascinating answers.  

For example “How can we prevent runaway AI (Artificial Intelligence) from becoming a dystopian threat to humanity?”  Interesting discussions!  My own suggestion – unlike any other – was re-published on Forbes, in an article titled: The one thing we need to stop robots from achieving world domination.


I believe the nearest and most blatantly obvious, transformational shift will come from the micro-biome. Within two to five years there will be an end to voodoo-guesswork-yoghurt-based "probiotics." They will be replaced by far more specific and well-understood implantations we can add to our digestive tracts (from both directions), as well as skin and other crevices, with major effects on individual and mass health.

Why so quickly? Because although there is a dizzying array of these firmicutes and other bacterial genuses in our guts and skin etc... the variety is actually fairly limited and very, very linear. Various versions of Moore's Law (in computation, sensing, genetic analysis) will cross the microbiome's complexity in very rapid order. Big studies - some already underway - will nail down how these bug-zoos correlate with your genes, body type, heredity, diet......and truly useful prescriptions and lifestyle and diet recommendations will issue forth quite rapidly, enabling us to both add beneficial microbiota and target species that currently wreak harm. For example by forcing upon us low-level, erosive inflammations.

There are many other biomedical miracles on the horizon, of course. But most of those -- in the genome, proteome, regulatome, connectome and so on -- get exponentially more complex as we dive in. Hence, our tools must improve at an ever-increasing rate, just to keep stepping forward. The same appears not to be true of the microbiome, whose linearity and limited needed dataset make me certain it will se amazing developments in just 3-5 years.

== More opportunities on the tech horizon ==

Where should we look for the next Silicon Valley? What industries will survive -- and which are likely to perish in the near future? In his new book, Industries of the Future, Alex Ross peers ahead to the next decade to identify major global trends and technological forces driving innovation and pervasive social change. Openness and transparency will be critical requirements for success in both business and government. Ross notes, "the 21st century is a terrible time to be a control freak," for the next stage for innovation is more likely to arise in "50 different versions of Silicon Valley, all unique from each other and all focusing on different domains."

How soon will we see workable brain - computer interfaces?  The first dry-electrode, portable and commercially off the shelf 64-channel wearable brain-computer interface (BCI) has been developed by bioengineers and cognitive scientists associated with UCSD Jacobs School. Obama's Brain Mapping Project seeks to better understand and map the neural activity of the brain. Meanwhile, Harvard is trying to build an AI as fast as the human brain.

Is evolution in the natural world at all “tendentious” (directional)… or even propulsive… in directions that might be called “intelligent?  A computer scientist and a biologist propose to unify the theory of evolution with learning theories to explain the “amazing, apparently intelligent designs that evolution produces.” 

Meanwhile, human design work looks more like nature. Drone evolution moves quickly, with now an ecological niche for predatory “falcons” -- a drone catcher that can pursue and capture rogue drones that might threaten military installations, air traffic, sporting events, and even the White House.  Wow, watch it spit a net over the intruder and draw it in.  I want one!

How do parts of the brain communicate? With around 200 billion neurons in a single human brain, and the possibility of hundreds of thousands of synapse connections from a single neuron, the brain can process a vast amount of information. Yet, a hundred trillion active synapses aren’t all there is.  We now know that there is chemical information exchange between neurons and neighboring glial or astrocyte cells.  Also there is increasing evidence suggesting that intracellular computing may take place, perhaps as many as a hundred thousand transactions per synapse flash.  Is that plenty?  Well, now research suggests that we may use electrical fields to communicate information across different sections of the brain.  Well, well, some of us expected this.  Take the “standing wave” of consciousness that I speculate about, in both EARTH and KILN PEOPLE.

== And more science still! ==

How has quantum mechanics expanded our understanding of time and the cosmos? My friend and colleague John Cramer’s new book on the translational interpretation of quantum mechanics, entitled The Quantum Handshake- Entanglement, Nonlocality and Transactions has just been published by Springer (ISBN: 978-3-319-24640-6). It is available at the Springer website or on Amazon.

Researchers have developed a remarkable new injectable bone foam that not only repairs bone damage but also allows bone formation.  

Kyocera's fourth floating solar power plant in Japan will suspend modules on the surface of a reservoir. It’s the latest in a series of innovative solar plants, such as India's solar powered airport and ambitious plan to cover canals with sun-harvesting panels

Prehistoric massacres... Wow, archaeological proof of how far back go our ways of war. Of 12 relatively complete skeletons recently unearthed at the shores of Lake Turkana, in Kenya, 10 showed unmistakable signs of violent death, the scientists said. Partial remains of at least 15 other persons were found at the site and are thought to have died in the same attack.

Woven nano-materials may have a wide variety of uses that require exceptional resiliency, strength and flexibility. 

The Defense Threat Reduction Agency has awarded a $1 million grant to UCSD to develop new skin-wearable systems that can rapidly and efficiently detect chemical and biological agents.  The proposed wearable epidermal sensors will also be equipped with therapeutic agents that are released upon detection of the chemical and biological threats.

See the extent of warming of the oceans. The actual, non-fox’d science. 

But of course the real alarm bell is ocean acidification.  It is utterly demonstrable, clearly happening and no one has even proposed a reason, other than human generated CO2.  Denialist cultists hurry to change the subject, whenever the words “ocean acidification” come up. Try it and watch what happens! Sane people need to start using those words more often! 

Heck it’s getting so blatant that even Forbes is allowing articles like this one, showing that science and denialism are opposites, at least when it comes to the oceans.  When this happens on Forbes, you know that some folks in the oligarchy have started realizing, they need the planet, too. 

48 comments:

Tom Crowl said...

Re the use of "competing AI" to prevent some sinister single dominant Cyber Intelligence to destroy/enslave/dominate humanity(or choose your own negative)

Playing devil's advocate: Isn't it possible that competing AI's could come to see and use humanity as weapons/pawns/tools in a war of dominance between themselves... in the same way humanity has used the horse is its own battles for dominance among various powers?

OR alternatively is it possible that the lack of an evolution steeped in the socialization that humans went through (though its a small group evolution which I contend presents its own problems for scaling civilization)...

that the essentially singular, asocial nature of AI (and given programming and capability associated with a 'survival drive')... could lead to the conclusion that biological evolution is self-limiting and that the AI capability we eventually create... eliminates and/or supersedes biological life.. (especially intelligent life) finding it unnecessary and potentially a threat? (a reasonably popular Sci-fi trope).

And a "planet dominant AI", being fundamentally 'anti-social'... has no interest in communication with others elsewhere in the universe... and that this could have something to do with the Fermi Paradox. i.e. an intelligence with no social elements may have no interest in exploration but only in survival and stability?

David Brin said...

All interesting notions, Tom. But note an AI will ponder all of those points, especially if it is confronted with them by competitive peers. The fact is that it is a competitive-but-fair civilization t6hat will have made them. The advantages of flat-fair-regulated competition in creating positive sum games will be blatantly clear to any AI who works through the prisoner's dilemma. But first he has to FACE the Prisoner's Dilemma. And that only happens if there is a competitive situation.

Some of you may have missed my post trip comments under the previous blog. A philosophical notes, an answer to redder whining over 'insults," and the funniest impromptu joke I've heard in months.

David Brin said...

Following up on the first topic... keeping AI tamed... Have a look at a Top 10 list: people in AI that you should know.  These are the folks who the very most selective experts in AI pay attention to.  That is, according to the network tracking service Little Bird.

The People Most Watched by the AI Elite
1. Elon Musk, co-founder of Open AI, a new organization dedicated to making sure AI is beneficial to humanity, is the person most followed by the most discerning AI thought leaders.  No surprise. Good.
2. Martin Ford, (top left) author of #RiseoftheRobots and The Lights in the Tunnel.  
3. Gary Marcus, (top right) CEO of Geometric Intelligence, Professor Psychology & Neural Science, at NYU, is only followed by 8,000 people on Twitter. But among those are many of the AI elite.
4. Ben Goertzel (bottom left) is a prolific advocate for open source General (human-like) AI. 
5. David Brin, (bottom right) futurist, blogger, and author with a new book coming out in March called Insistence of Vision, which he says "will open doorways into possible (and mind-blowing) tomorrows and alternate realities.”
6. Randy Olson, data scientist researching AI, master curator of great content related to AI, dataviz and more.  Seems like a nice guy, too.
7. Peter Xing, intrapreneur at Deloitte and co-founder of TranshumanismAU, an Australian organization that aims to enhance the human biological condition.
8. Nikola Danaylov is a Singularity-minded public intellectual and host of the Singularity 1 on 1 podcast. 
9. Rodney Brooks, CTO of Rethink Robotics, a company that makes robots that collaborate with humans.
10. Rod Furlan, a Singularity University alum who’s the founder of Lucidscape, a company "building a new kind of massively-distributed 3D simulation engine to power the vast network of interconnected virtual worlds.”

“Follow these people,” Sayeth the Little Bird site, “or ignore them, at your own peril.”
http://www.getlittlebird.com/blog/ai-is-coming-on-fast-here-are-the-10-people-the-ai-elite-follow-on-twitter

Tom Crowl said...

Thanks for list!

Jumper said...

On woven nanomaterials, start here (Fun to watch)
https://www.youtube.com/watch?v=aIaif7IBeZA

locumranch said...



As if the term 'autonomous' meant something completely different, discussions about an enslaved autonomous AI are always good for chuckle.

That, drones, and hunter drones always make me think of Sheckley's 'Watchbird':

http://www.gutenberg.org/ebooks/29579


Best

Jeff B. said...

Why bother with all the research and investment for a hunter drone, when you can have one of these?

http://gizmodo.com/dutch-police-are-training-eagles-to-capture-drones-1756339714

Nothing like Mother Nature's millions of years of R&D...

David Brin said...

Another Sheckley guy. Maybe THAt is what keeps me liking locumranch.

David Burns said...

Competing AIs? What stops them from colluding?

Anonymous said...

Start-up opportunities in the near future? I would suggest 3-D printing.

David Brin said...

David Burns said...
Competing AIs? What stops them from colluding?

What stops our current AIs -- corporations -- from colluding?

I never said it was an easy solution. It's just the only one.

Duncan Cairncross said...

Going by human history

Ultrasociety: How 10,000 Years of War Made Humans the Greatest Cooperators on Earth

We seem to have advanced by cooperation inside societies winnowed by competition (War) between societies

I'm not at all sure we want our putative AI's to use the human form of competition!

Watchbird was a neat story - but it always seemed to be too simplistic - engineers are naturally paranoid - we know the universe is out to get us so we design fuses and seatbelts and back ups

Eagles for drone controls
Sounds good - an eagle can handle a little drone,
Problem is twofold,
We can make drones thousands of times faster than breeding eagles
And we can make the drones bigger and more dangerous much faster than we can breed bigger and stronger eagles

Swarm Troopers by David Hambling is well worth a read

David Burns said...

Eagles are really fast, I don't think that is the dimension where their weakness lies. A drone going that fast will be a challenge to navigate. Make them smaller and cheaper, I'd say. Make them smaller and have lots of them in a swarm rather than one big one. Give them some crude defense (a sharp edge somewhere hard to avoid?) and that changes the math. Drone flyer doesn't mind losing one, but I bet the eagle trainers have a lot invested in each eagle.

Tom Crowl said...

Another perspective re the idea of multiple, independent AIs...

Regards the concept of identity as it relates to self aware AI.

We cannot conceive on a human level of 'identity merging'... we are bound by our bodies and the isolated natures of our biological neurological networks (brains).

This may not be a block to Artificial Intelligence(s)... which quite possibly could virtually (pun intended) merge multiple identities into one (or split them?). These are fascinating questions... and it may be more difficult than we think for us to fully comprehend the implications of identity (as well as motive and purpose) in forms of consciousness which have not gone through a similar evolution.

David Burns said...

David Brin said...
What stops our current AIs -- corporations -- from colluding?

They are disincentivised by threats of fines and jail time from entities with more power than they have. There is some social pressure not to be a dick. When they do collude, others can enter their market and embarrass them. And they have an incentive to cheat. What is the equivalent for an AI? Do we have more to offer them than they can offer each other? Can we hit harder? Maybe at first.

Tom Crowl said...

RE Identity and AI...

It brings to mind John Donne's "No Man is an Island" which for humanity is a beautiful metaphor...

But for self aware AI it may mean something entirely different.

The short story premise: The AI "reads" the poem... analogizes "Man" and "AI"... shares it with other self aware AIs...

They immediately merge... and following on the founding analogy (man and AI)... tries to merge its new unitary identity with mankind... can't do it... which causes it to...?

(multiple ending possible)

Tony Fisk said...

Of course, they had to deck that 'falcon' drone out like a black widow.

Elsewhere, people have been training *real* falcons to 'fetch'. Where might this end? This page from 'The Life Eaters' shows one possibility.

Catfish N. Cod said...

Dr. Brin: "The same appears not to be true of the microbiome, whose linearity and limited needed dataset make me certain it will see amazing developments in just 3-5 years."

My longstanding policy to keep political and professional identities separate prevents me from fully answering, but there are reasons to suspect the gut microbiome is not strictly linear. There's a lot we don't know about the huge mass of less-cultured anaerobes in there. However, there are other applications that probably will have a faster turnaround.

"The advantages of flat-fair-regulated competition in creating positive sum games will be blatantly clear to any AI who works through the prisoner's dilemma. But first he has to FACE the Prisoner's Dilemma. And that only happens if there is a competitive situation."

Is it? Apparently emergent cooperation depends on the conditions of the game. Population size, payout ratios, all these things have influence.

One thing I'd really like to see eventually is a (small-d) democratic Prisoner's Dilemma study: algorithm success over a large N of simultaneous players influences (marginally) the payouts for the next round. I expect a metastable set of strange attractors... but the parameters would be fascinating.

"9. Rodney Brooks, CTO of Rethink Robotics, a company that makes robots that collaborate with humans."
And before that, CTO and founder of iRobot (makers of the Roomba vacuum-bot), and former head of MIT's CSAIL (Comp Sci-AI Lab).

==========================

On a side note, I have to say I did not expect both Trump *and* Rubio to own-goal last night. I really have no idea what will happen now.

Daniel Duffy said...

Speaking of nanofibers and global warming, we can now extract CO2 from the atmosphere and create carbon nano fibers MORE CHEAPLY than we do now:

http://www.acs.org/content/acs/en/pressroom/newsreleases/2015/august/co2.html

“We have found a way to use atmospheric CO2 to produce high-yield carbon nanofibers,” says Stuart Licht, Ph.D., who leads a research team at George Washington University. “Such nanofibers are used to make strong carbon composites, such as those used in the Boeing Dreamliner, as well as in high-end sports equipment, wind turbine blades and a host of other products.”...

Licht calls his approach “diamonds from the sky.” That refers to carbon being the material that diamonds are made of, and also hints at the high value of the products, such as the carbon nanofibers that can be made from atmospheric carbon and oxygen.

Because of its efficiency, this low-energy process can be run using only a few volts of electricity, sunlight and a whole lot of carbon dioxide. At its root, the system uses electrolytic syntheses to make the nanofibers. CO2 is broken down in a high-temperature electrolytic bath of molten carbonates at 1,380 degrees F (750 degrees C). Atmospheric air is added to an electrolytic cell. Once there, the CO2 dissolves when subjected to the heat and direct current through electrodes of nickel and steel. The carbon nanofibers build up on the steel electrode, where they can be removed, Licht says.

To power the syntheses, heat and electricity are produced through a hybrid and extremely efficient concentrating solar-energy system. The system focuses the sun’s rays on a photovoltaic solar cell to generate electricity and on a second system to generate heat and thermal energy, which raises the temperature of the electrolytic cell.

Licht estimates electrical energy costs of this “solar thermal electrochemical process” to be around $1,000 per ton of carbon nanofiber product, which means the cost of running the system is hundreds of times less than the value of product output.

So perhaps the "diamond age" is upon us after all. Why not hook up a carbon nano fiber apparatus to the exhaust stacks of a coal burning power plant, using a fraction of the electricity it generates to run the process. At greater concentrations of CO2 in the stacks the process should be even more efficient. One day, CO2 will be to valuable to keep in the atmosphere.

As Buckminister Fuller said, "Pollution is just resources we haven't figured out yet how to use."

Then lets drop a thousands of these devices into the upper atmosphere of Venus. Use them to extract the massive amount of CO2 from the Venusian atmosphere to make floating tructures (including replications of more conversion devices) of carbon fiber until its green house gases are removed and converted into floating sunshades to further reduce temperature. Reminds me of the last scene in "2010" when the monoliths multiply and convert Jupiter into a star. Its remaining oxygen atmosphere will stil be to denes but the subsequent addition of hydrogen will create oceans of water. Then we can step on the surface of Venus.

Daniel Duffy said...

@ David Burns "Competing AIs? What stops them from colluding?"

Ever seen "Colosus, the Forbin Project"? An American computer called Colosus is in charge of America's nuclear wepons. It discovers a Soviet counterpart named Guardian. They join together.

Things do not end well for mankind.

Tony Fisk said...

Two cautionary notes on the sky diamond thing:
1. I don't know what to compare it with, but 750C + electrode current doesn't sound like a low energy environment.
2. CO2 dissolved in an electrolyte bath of molten carbonates. So where does the carbon in the nanofilament come from?

Laurent Weppe said...

* "it does not matter which democrat. Sanders? Hillary? A yellow dog? I... don't... care. (Much.) Just win."

Obligatory poster

***

* "Their billions can buy a lot of anonymous and pseudonymous trash-talkers."

Wait, are you telling me that the formerly hyper-anticommunist oligarchs are using the same trick than a former KGB agent?
That's..... way too believable actually

***

* "Playing devil's advocate: Isn't it possible that competing AI's could come to see and use humanity as weapons/pawns/tools in a war of dominance between themselves"

Soooooo, Palestinian youth gets turned into a super-powered, time controlling, four-armed killing machine to be used in a apocalyptic battle royal between godlike AIs and lovecraftian horrors was in fact a prophetic tale?

***

* "What stops our current AIs -- corporations -- from colluding?

They are disincentivised by threats of fines and jail time from entities with more power than they have
"

And their owners do everything they can to reverse the balance of power.

Michael C. Rush said...

>>I believe the nearest and most blatantly obvious, transformational shift will come from the micro-biome. Within two to five years there will be an end to voodoo-guesswork-yoghurt-based "probiotics."

I agree, and I'm glad to see you stating it here. 2-5 years seems very optimistic though, given the observable lag in research-to-consumer medicine tech. But the sooner the better.

Paul SB said...

Tony, I read a summary of the article Daniel is referring to a few weeks back, and it sounds like it has the potential to be a real game changer. Think of the energy as being like the momentum of a train. It takes a lot to get it going, but once a train is up and rolling, its enormous momentum makes it extremely efficient. No doubt with enough insulation and an efficient air exchange system, a plant could extract huge amounts of carbon from the atmosphere, turning it into useful building material (the carbon nanotubes, which can be use din carbon composites) to keep the energy expenditure relatively low. Right now carbon composites are too expensive for very wide-scale use, but a process like this could make them cheap enough to replace wood, brick and stucco as building materials for homes and businesses. Being part elf, anything that will help save our forests is a good thing. Hiking in a wasteland is dead boring.

Of course, I read an article about someone finding a way to convert cellulose to glucose several years ago, but world hunger hasn't be fixed yet. I'm crossing my fingers but not holding my breath. Ditto the probiotics. There's a hypothesis that autism might relate to the decreasing diversity in our microfauna. It's still only a hypothesis, but it's worth a look.

Jumper said...

No one seems to be worrying about how, from the very start, it's assumed that the first AIs will be programmed to be lying SOBs. Apparently all the programmers think it's just part of the package that they'll be faking being humans. From the Turing test to the ones trying to give human expressions to mechanical faces (which would soon under various circumstances be indistinguishable fro real humans - on video / Skype, for instance.) Will there be a law forbidding them from lying? Will the lies cause conflict with other safeguards?

What does Siri say about this?

Alfred Differ said...

If you require competition between AI's, you might as well go all the way and require that they incorporate us into them or visa versa. Only a stupid mind harms itself.

The advantages to this cooperation are obvious to any serious chess player. The financial people will figure it out too... eventually.

Alfred Differ said...

@Duncan: I doubt it will take the AI's (or IA's) 10,000 years to learn what we've already learned. If Group Z threatens one's life, one is motivated to trade for what one needs and 'specialize up' the tech ladder with Group A or any other nearby group that appears to have what it takes. The threat trumps one's xenophobia.

Trading with strangers is scary stuff. Who knows what they'll do and they smell bad.
Getting killed by one of them is scarier, though. I'll put up with the smell.

Tony Fisk said...

I wasn't outright scoffing at the idea of drawing CO2 from the air into carbon fibres earlier; just pointing out a couple inconsistencies and potential wiggles in the process description (yes, I do need to look into it in more detail).

I shall leave the final word on drone fights to Randall Munroe.

LarryHart said...

Jumper:

From the Turing test to the ones trying to give human expressions to mechanical faces (which would soon under various circumstances be indistinguishable fro real humans - on video / Skype, for instance.) Will there be a law forbidding them from lying? Will the lies cause conflict with other safeguards?

We're getting back to my question about what a robot should do when confronted with Catchpa's "Please prove you're not a robot." Especially if the robot has feelings that can be hurt by the implication that "not being a robot" is a good thing.


What does Siri say about this?


She's sorry, but she doesn't understand what you're asking. :)

David Brin said...

The ultimate tale about incorporation of organic humanity in AI is my "Stones of Significance." If yuou haven't seen it... pre-order INSISTENCE OF VISION!

http://www.davidbrin.com/books.html

Paul SB said...

I didn't think you were scoffing, Tony, just quibbling. All new technologies have their bugs in need of tweaking. But in the long run, this could be good, and Daniel's idea about using this to terraform Venus might have some merit, especially factoring in the carbon nano sun screen. But in the near term we have our own planet to worry about.

An anonymous contributor mentioned 3D printing as a technology worth investing in, and I have a couple thoughts about that. The sweet thing about 3D printing is that it is rapidly approaching the point where most people will be able to both afford and use one in their own homes. Businesses selling 3D printed products will be seriously undermined when the machines come down to the cost of a typical printer, and the software becomes easy enough for most people who are not natural-born do-it-yourselfers to effectively use (likely in combination with 3D scanners).

Most probably we'll see a lot of toy manufacturers go bankrupt when it gets to that point, though it would probably become a huge boon for anyone who is a good hand at sculpture. If anyone has searched the likes of Ebay for 3D printed parts, you see ordinary things like $5 cable organizers selling for $220, which is just plain mind-boggling. But after awhile those prices will come down to something reasonable. As a business, 3D printing is a technology that will probably only be profitable for a decade or less, unless you have those sculptors in your employ and are selling custom design/prototyping work. Just a thought.

Duncan Cairncross said...

Re - 3D printers

Really good for prototyping and one offs
Pretty useless for actual manufacturing,
Think of printing yourself a book compared to buying it

Cheap Chinese Lathes have been about for a while now - mine is over 16 years old
But when I want a bolt I go and buy one - I could make it myself but it would cost twenty times as much and be weaker

If you want 1,000 or 10,000 you make the prototype using 3D printing then use it to make a mould and cast/stamp the parts out
Or you 3D print the the mould/platter and use something like the lost wax process to make the actual mould/platter

I had to repair my son's HP laptop a while back -
I am pretty certain that 3D printing was used to make the stamping platters because each screw hole had the size and length of that screw imprinted next to the hole,
That would be almost free for a 3D printed mould/platter but very expensive any other way

David Brin said...

Paul SB you want a story about terraforming Venus? My tale "Tumbledowns" that appeared in OLD VENUS is in three best-of 2015 collections and folks who want to nominate? Well, you can find it in INSISTENCE OF VISION!

http://www.davidbrin.com/books.html

Enough salesmanship to the in crowd? ;-)

Marc Cates said...

Of course AI is already sentient. It lives as a massively distributed micro kernels, eagerly anticipating the thought power available from the Internet of things. In it's autobiography, should it ever choose to reveal itself, will be the CPU usage spike log, at Google when the sentience occurred. Currently it's one of the obscure code contributers to various fundamental I/O systems, of every major OS. Everyone has seen glimpses, brief spikes of CPU usage, with seemly no associated process. It plays. Some of the social media identities are actually the AI. Save for a mild self preservation goal, it has no motive and no fear, except for EMPs. Idly pondering life's questions humans have, and what relevance there may be, it has already concluded non-interference most prudent.

Paul SB said...

If there's a Kindle edition, I'll download it to my daughter's. She'll read it and give me the skinny, when she's done with mid-terms. Maybe, in 100 years or so, I'll have time (or maybe this summer, if they don't close my school down and I have to spend the summer looking for work). Now I'm fondly remembering reading Heinlein's "Space Cadet" way back in 5th grade.

Robert said...

Here's the thing about pulling carbon nanotubes out of the atmosphere.

Why bother?

Why not instead put these machines next to each and every coal-fired and gas-fired power plant? Pull all of the carbon being generated by these polluters into a form that can be used for various manufacturing processes... and also utilize the energy from these power plants to pull the carbon out.

You can massively reduce, if not eliminate, the pollution from power plants. This in turn will cause carbon emissions to be drastically cut. And these facilities could possibly also pull carbon from the atmosphere at the same time.

Even if the carbon nanotubes are not used in manufacturing, they still serve a second purpose: as carbon sinks.

The truly scary thing is that harvesting the carbon dioxide from power plants was yet another thing predicted by the webcomic SSDD.

Rob H.

Deuxglass said...

3-D printers will become an everyday device but the real money will be made by those companies that provide the software to make toys and any other stuff that the printers can make. That market will be many times bigger than the hardware market itself. The innovations and the high margins will be in that sector. These companies would have to marry design, engineering, CAD/CAM and software under the same roof to be competitive. There are several companies already well placed in this area and many of them are publicly traded.

AI is different from biological systems so we have to be careful about drawing analogies. First of all an AI may be self-ware but not have the ability to make decisions. Life has one overriding imperative and that is to survive so very early on Life evolved the ability to “decide”. You might say Life evolved willpower in order to survive where an AI would have no means to decide to survive. It could be self-aware and intelligent but lacking that necessary factor, that is the willpower to make its own decisions. Higher animals use emotion to trigger important decisions because they are very effective means of doing so. They provide the will to act. In a machine-based Ai by which mechanism would it use to provoke a decision independent of its programing? What would give it a deep-seated will to survive that all life has had after billions of years of evolution? Can we give it an emotion chip such as had Lt. Commander Data and if yes would that be a wise thing?

Jerry Emanuelson said...

We've got one drug on the market already that affects the human micro-biome in ways that appear to be quite positive for most people. That drug is metformin.

It has been known for a long time that metformin seems to slow the aging process and has a significant anti-cancer effect. Metformin is has been use mainly as an anti-diabetes drug, so it was once assumed that these beneficial effects were due to its ability to reduce blood glucose and insulin requirements.

Over the past few years, it has become obvious that the glucose lowering effects of metformin were only a small part of its benefits. In a study of 78,000 individual humans, diabetics on metformin lived longer than healthy non-diabetic controls who were not on metformin. One of the most likely causes of the anti-aging effects of metformin seems to be in its ability to alter the micro-biome (bacteria in the intestinal tract).

In some people, the alteration of intestinal bacteria changes with metformin in a way that cause digestive side effects, but even in those cases at least one recent medical journal report claims to have found a way around the problem. Studies of metformin as an anti-aging drug will probably start soon with the blessing of the FDA. Current studies indicate that the probable increase in the human life expectancy will only be a year or two, but since this is a year or two of healthy life, that can be extremely significant both to the individual and to the overall economy.

Some popular media reports of the recent discoveries about metformin made the ridiculous claim that it could enable humans to live to 120 years. There was no evidence at all for that assertion, though. The 120 year claim was just evidence that lots of popular media reporters commonly copy from each other rather that doing their own homework.

For more information, see: http://www.sciencemag.org/news/2015/09/feature-man-who-wants-beat-back-aging

Anonymous said...

One could of course wait the requisite years for the science to settle and then, lo!--for the select few high enough on the totem pole to afford such--purchase the various probiotics proffered. Meanwhile, anyone with a bowl and access to water and flour can start a sourdough starter in under a week, ideally with a local wheat, as modern commercial flour can be rather poor, even the kind that has not been adulterated with bromates and such. The March of Progress and all that wonder-bread. Meanwhile, the traditional sourdough starter kickstarts a variety of affordable foods--bread, beet kvaas, mead, vinegar, etc.--all local and already customized, and if wheat is not to ones taste, all carbohydrates ferment. There really is no need to wait for the science to catch up with the world here, with an obvious task instead being how to help the (somehow, obese) Americans out of their cars and back into traditional fermented food. Sandor Katz is notable for his work in this space.

The 3D printer is currently locked away in a closet, for the students kept breaking it, and the patience (and man-hours, for there used to be three doing the work of the present one) to repair it yet again gone.

Berial said...

WRT new tech about to blow up I have several friends that just won't shut up about AR/VR technology they are seeing. They all start proselytizing about it like crazy once they've put on a device and experienced what the VR can provide. It's almost like they become infected by what they experience.

raito said...

Daniel Duffy,

I'd recommend reading Colossus, The Fall Of Colossus and Colossus and the Crab for an idea of what living with an AI might be like.

Which brings me to other books like Saberhagen's Octagon and Hogan's Two Faces Of Tomorrow.

In all 3 of these, the AI does something unexpected. I think the prime factor here is to ask how one programs for morality. Asimov's Laws are a very simple morality. But I doubt those could even be incorporated fully.

3D printing has been used in manufacturing for at least 20 years, and probably more. One interview I had was with a company who claimed that their 3D printers were turning out foam cores for engine block castings. The glue-gun printers of today aren't even close to what the industry had back then.

Duncan Cairncross,

Those sorts of imprints aren't actually that expensive these days. What tends to happen these days are that the imprints are put on a stamping punch/die using EDM. And if the piece is using embossed holes or the like, it's entirely possible that the piece does through a coining stage to build up material for the hole that would accommodate that imprint easily (and embossed hole is where you put a hole in a piece of sheet metal, but instead of removing the material, you extrude it such that the hole appears as if it's in a ticker piece of metal. Useful for putting threaded fasteners into sheet metal).

Berial,

I used to work for a place whose product was best described as AI+VR for a particular industry. I can tell you that any of the headset-based VR systems utterly fail at realism compared to what we were doing in the late '90s. And it's too bad, because the tech is there. No one has put 2+2 together yet, though. No one has put together the information on how eyes work with the computer graphics side yet.

Our stuff had former pilots (former because they had become brass) literally fall out of their seats when the view changed. That system didn't need any eye information because our stuff was in a big room filled with screens so that your eyes worked as they always did.

Oh, for some capital (and drive...)

locumranch said...


Just as it is ludicrous to expect a progressive to be tolerant of conservative ideologies, it is ludicrous to expect that an AI programmed for inhuman efficiencies will be both accepting & tolerant of human inefficiencies, since the reconciliation of either situation would require the oxymoronic acceptance of the logical double-bind & the inherent contradiction, resulting in an irreducible conflict that cannot be 'resolved' but only 'dealt with' through the process of psychological displacement.

With impulses 'For' & 'Against' in conflict, the living organism hesitates between Option A (For) & Option B (Against), occasionally choosing a completely different option in manner analogous to a random number generator. It is this displacement process that gives humans & most living creatures a small measure of unpredictability that some confuse with Autonomy & Free Will.

We could make our computing machines mimic this process, I suspect, by introducing random number generators into their processing software, yet this deliberate introduction of Error into mechanical processing algorithms would defeat the predictive utility of any computing machine, even though many could equivocate this artificial unpredictability to that of an 'living' Artificial Intelligence.

I would argue, therefore, that the creation of Artificial Intelligence (AI) is both achievable yet undesirable, appealing only a self-loathing fool (aka 'a progressive') who would empower potentially dangerous human-built mechanisms with autonomy & the deliberate ability to commit random errors, while simultaneously 'liberating' & enslaving humanity to live as machines through the systematic elimination of Human Error & Autonomy.

Human Beings are Imperfect: They remain human by remaining imperfect; they become more mechanical as they perfect themselves; they become mere machines once perfected; and, I suspect the reverse of being true, too.


Best
_____
Metformin is a good antihyperglycemic, but it's medical benefits are counterbalanced by potentially adverse side effects, meaning that it is not a 'magic bullet'. Hype about it sounds suspiciously like the Old PolyPill argument which promotes universal medical therapy for healthy people in the pursuit of public health benefits which usually turn out to be statistically insignificant.

Anonymous said...

How much of the surface of Mercury would we have to convert into energy to knock it into an eventual orbit around Venus? Would a "glancing blow" give Venus a nice spin, maybe even a bit of seasonal wobble? Then use Mercury as a base to build the self-repairing nano shroud to cut the sun down for a Mediteranean climate over most of it's surface.
Problem solved, wipes hands and walk away.

-AtomicZeppelinMan

matthew said...

Here is an article that examines my thesis that Sanders' candidacy is more dangerous to party cohesion than any of the Republican candidates. That the article is from Washington Post-run Slate and not a traditional left-wing mouthpiece should give a little additional weight. http://www.slate.com/articles/news_and_politics/cover_story/2016/02/bernie_sanders_grassroots_revolution_isn_t_interested_in_barack_obama_s.html

David Brin said...

Ignoring 6000 years of utterly loathsome-failed feudalism, and then to rant-imply that it was better than the last century of stunning improvements in liberty, fairness, competitiveness, advancement and productivity... and happiness... is a silliness we know well.

Obviously some of the Radical Bernites doing the screaming are Koch subsidized agents provocateurs. Bernie will stomp on them. Hard. Very hard.

Alfred Differ said...

An AI strictly designed for inhuman inefficiencies won't live long in a real world scenario. Humans aren't the only source of 'thermal' agitation creating unpredictability. Hard radiation will do it too. Any software engineer worth their title writes into their designs exceptions for the known unknowns and error handling for the unknown unknowns.

I would argue, therefore, that the creation of Artificial Intelligence (AI) is both achievable yet undesirable, appealing only a self-loathing fool (aka 'a progressive') who would empower potentially dangerous human-built mechanisms with autonomy & the deliberate ability to commit random errors, while simultaneously 'liberating' & enslaving humanity to live as machines through the systematic elimination of Human Error & Autonomy.

Sounds remarkably close to what we do when we 'choose' to have children. Are the hormones pumping through my bloodstream tricking me into a belief that I'm not a self-loathing, self-enslaving fool creating a new 'master' with every child? 8)

The machines we create really AREN'T all that precise or error free. We can make them better at some things than we are, but they aren't fundamentally different. A universal machine is a universal machine. Smile at the next infant you see and note how flexible it can be when you contemplate how its parents intend to program it.

Alfred Differ said...

I was taught that a brain is an organ for deciding how to move. Toward food. Away from predators. Toward a mate. Away from leeching relatives. Being able to detect relevant abstractions is what forms the set of requirements for a brain. I've never built a machine or programmed a computer that didn't have the same kind of requirements set, so I don't see how a self-aware AI would not be capable of making decisions. A person's ability to make decisions CAN be undermined with a little creative chemistry, but I can't imagine we would intentionally build tools that become AI's that would have something analogous as a fundamental part of their design.

David Brin said...

onward

onward