It’s always good to dream big, but sometimes dreams take you in unexpected directions. Growing up with science fiction, I reveled in tales of manned exploration of the Solar System and nearby stars, many of which I assumed would eventually become reality. But I never dreamed about personal computers. You can go through the corpus of science fiction in the first two-thirds of the 20th Century and find many a computer, but there are few tales involving personal computers on the desktop. An exception is Murray Leinster’s short story ‘A Logic Named Joe,’ which ran in the March 1946 issue of Astounding Science Fiction. Leinster invokes something like today’s massively networked computers in a story that anticipates the Internet.
How did science fiction fail to see something as huge as the PC revolution coming down the tracks? Maybe it’s because the future still surprises even those whose business it is to imagine it. I’m musing about all this because of my own desktop PC and the views it’s showing me, not to mention the continuous datastream that updates every mission I’m keeping an eye on. Surely this is a science fictional future as real as Leinster’s.
A widely distributed network lets us see things on demand, one of a kind things like an object that has never before been seen in detail as it slowly swims into focus in our cameras. We’ll have that experience in 2015 as New Horizons arrives at Pluto/Charon, but we’re also getting a taste of it right now as the Dawn spacecraft approaches Vesta. What we’re seeing in the early images has been taken for navigation purposes by Dawn’s framing camera, and as the video shows, we’re already looking at views that are twice as sharp as the best images previously available from the Hubble Space Telescope. Surface details are still a mystery, but Dawn will eventually swing as close as 200 kilometers around the asteroid.
With Vesta and then Ceres ahead of us, let’s also keep a close eye on Pluto, and we don’t have to wait until 2015. In late June, NASA’s Stratospheric Observatory for Infrared Astronomy (SOFIA) — a modified Boeing 747SP carrying a 2.5-meter telescope at high altitude — was able to observe an occultation as Pluto passed in front of a background star. This is a case where SOFIA’s airborne capabilities shone, for Pluto’s shadow fleetingly passed over a mostly empty stretch of the Pacific Ocean. SOFIA was able to position itself in the center of the shadow’s path to make the observations.
You wouldn’t think you could do much with Pluto from a mobile observatory here on Earth, but the science is actually quite rich, says Ted Dunham (Lowell Observatory), who led the team of scientists onboard SOFIA during the Pluto observations:
“Occultations give us the ability to measure pressure, density, and temperature profiles of Pluto’s atmosphere without leaving the Earth. Because we were able to maneuver SOFIA so close to the center of the occultation we observed an extended, small, but distinct brightening near the middle of the occultation. This change will allow us to probe Pluto’s atmosphere at lower altitudes than is usually possible with stellar occultations.”
I have no images of this one, but it’s easy to follow the exploits of SOFIA on the Net. It’s also worth noting that Dunham was a member of the team that originally discovered Pluto’s atmosphere by observing another stellar occultation using the Kuiper Airborne Observatory in 1988. SOFIA’s mobility allowed the scientists to quickly change position when it was learned that the center of the shadow would cross 200 kilometers north of the aircraft’s flight path. A revised flight plan and air traffic control clearance allowed the productive change of course.
Our space-based resources send us things we would never have imagined seeing, as witness ESA’s Mars Express, which was able to perform a special maneuver of its own to observe an unusual alignment of Jupiter and the Martian moon Phobos. The alignment occurred on June 1, when there was a distance of 11,389 kilometers between Mars Express and Phobos, and a further 529 million kilometers to Jupiter.
Image: Three frames from the series of 104 taken by Mars Express during the Phobos-Jupiter conjunction on 1 June 2011. Credits: ESA/DLR/FU Berlin (G. Neukum).
We often lament how the future that was imagined in the 1950s and 60s hasn’t materialized — where are the human missions to the outer planets we thought would be flying now? — but the big surprise that science fiction never showed us was what we could see by staying home. Now we wind up looking at imagery from robotic missions on demand, tapping cameras orbiting Mars and closing on major asteroids. Moreover, a high definition video stream studying Earth down to one-meter resolution from the International Space Station is scheduled to go online in 2012. No humans near Jupiter yet, but the view on our personal screens seems to be getting better all the time.
“The future always arrives too soon and in the wrong order” – John Brunner
This quote is from the first book that I can remember that deals with the cloud computing, personal computer/ hacker/ trojanprograms/ cyberwarfare/ social network-driven revolution-The Shockwave Rider by John Brunner.
The next revolutions will be in personal robotics and personalized genome. This year marked a sea change in the practice of Medicine- several examples of personal diagnosis based on Genome-scale sequencing, not just looking at individual genetic markers. This future will arrive like a night train, hard to see coming but also impossible to ignore.
So we all get front row seats to space exploration as robotic missions explore the solar system. With crowd sourcing we can even participate in discovery! Still- I see a future where we use the best of all these technologies to expand our home range beyond the surface of this one beautiful planet.
Thanks for this. The unintentional irony of seeing people complain over variations of “dude, where’s my jetpack? I was promised a jetpack..” on a pervasive, global, affordable and near-instantaneous information infrastructure never ceases to amaze me.
On a more space-centric note, popular press characterizations of the US space program as being “done” with the retirement of the space shuttle fleet does an incredible injustice to all of the amazing things that continue to be discovered with robotic probes everyday.
Paul, even Arthur C. Clarke was a little late in seeing the PC future. But, in a book called “Imperial Earth”, which came out in 1975, he describes in detail a device called a “Minisec” which was essentially a PDA which connected to the global “Internet”, had wireless data connectivity, built-in encrypted security, voice recognition, etc. On the computer time-scale, he was way ahead of the game on that one.
And no one would know about the advantages of today’s connectivity more than you, Frank. I want to introduce my readers to Frank’s five year journey around the globe:
http://www.tahinaexpedition.com/
Take a look at this dispatch from Fiji:
http://www.tahinaexpedition.com/2011/06/plan-change-and-return-to-savusavu.html
and ponder the glories of sailing around the world on a schedule like this. Frank, great to hear from you, and continued calm seas. I’ll look forward to having coffee with you again when you get back in these parts. Meanwhile, your blog keeps delivering photos and news via my RSS reader. Great for the vicarious traveler…
Great article Paul, as usual. I’ve often wondered the same about the future. Sure it’s gone places we could never predict. Sometimes I wonder if we, as a civilization, have become to complacent and are to afraid to dream big as we used to, due more to a fear of losing what we have gained than a lose of imagination?!
Bruce Sterling made a similar statement on prognostications – that the future is already here, just not widely distributed yet. Networks were IT company things in the 1970s, small localized affairs in the 1980s, world-wide, but nerdy in the 1990s, and now practically ubiquitous and becoming all pervasive. Sequencing DNA and 3-D printing seem to be heading towards the “hobbyist” stage that PCs went through in the 1970s. Where will gene-crafting take us when it takes off? See this week’s “New Scientist” for “The Evolution Machine” to see what I mean.
That video shows why better optics will never replace the need for an interstellar probe. Just imagine how big a Hubble successor would have to be in order to match the resolution we will get when Dawn orbits Vesta.
Actually, I’m a bit disappointed at these emerging views of Vesta. To me it just looks like a bigger asteroid. The old pictures had splotches of blue and brown. But the reality is that its just shades of gray (sigh!!!).
@Adam. I used to work in a small biotech company doing “yeast surrogate genetics” where scientists were very tediously trying to insert functional genes into yeast and selecting the operational cells to develop yeast platforms for pharmaceutical drug testing. I read the “Evolution machine” in NS and was very impressed in how far this technology has got to in less than a decade.
We really are going to see the biotechnology century unfolding rapidly.
Excellent article Paul! This also ties in to our views on interstellar travel and what advanced alien beings may be like.
When Wernher von Braun envisioned manned expeditions to the Moon and Mars, he not only thought big in terms of spaceships but crew numbers as well. He did not forsee computers and other machines doing the work that he assigned humans to for the missions. I also recall that he did not forsee automated probes arriving at those worlds before the human missions to scout things out; that would happen only when the people arrived.
The same situation is happening with interstellar travel. Yes most people now get that the first true interstellar voyagers (ones which will reach another star system in under a few centuries and still be in working order) will be robotic probes with AI brains, but there is still the sense that they will automatically be followed by big starships with human crews, people who somehow won’t be all that different from us despite the centuries and technological advances between now and then.
All interstellar missions may be done by Artilects while our descendants remain in the Sol system watching what goes on (or not). Our descendants may also be very different from us, changed by technology that lets them be wildly different from their biological bases and ancestry.
These paradigms thanks to Star Trek and such will take a while to fade away, perhaps not until we actually do start sending probes into the galaxy. This assumes that our paradigms, biases, and limitations don’t keep them from happening in the first place.
I recall reading a book way back in the late 70s where people used handheld devices to communicate and access a network or computer. I believe they were called “Jokers”, but I can for the life of me not figure out the book or author I remember this from, nor enough other details from the story to intelligibly recount. I don’t think it was Imperial Earth, though.
Should this ring a bell for anyone, please let me know!
It seems to me that when we discuss tomorrow, we think about how much different our progeny will be from us today. Looking back, how much different are we from the people who lived a hundred years ago, or a thousand.
We have better tools, better health care, we may have improved our condition, but in reality I think someone from 1011 or 1911 might be able to find their way in the world today. A door is a door, a car is a horseless cart, food is food, the essence of our daily lives is the same, even if the conditions have dramatically changed.
I think we have found a default condition of humanity, that what makes us human is what made us human a thousand years ago, and what will make us human a thousands years from now. We may live longer, healthier lives, tweaking our genome here and there, but we aren’t going to surrender our essential human features. I just can’t see humankind giving up what makes us human. Even as we leave the cradle, we will keep our humanity, not because it is a limitation, but because it is essential to our nature.
We hope that we move along from our prejudices and tribalism, but I believe that our nature will follow us into the future. We will be who we are, just with better tools. Our humanity will follow us to Olympus Mons, to Titan, and even farther out. Our future is tied to our past, we are, and always will be, plucky hunter gatherers and tool makers, living, loving and fighting as we make our way across the universe. In the end, that may be the only constant in our future.
David, you are correct that if humans do not tamper with themselves mentally and physically, then we will remain the creatures we have been for ages for many more ages to come. Natural evolution is generally a slow process, with the occasional comet or planetoid impactor helping to speed things along, or at least mix it up a bit. As far as we know the dinosaurs never got really smart despite existing for 160 million years.
But we now have technologies and information that have radically and rapidly changed our society and species in a matter of generations, not eons. Look at how quickly our population has soared to seven billion members with another two billion expected by 2050. When I was a kid there were less than 3 billion humans on Earth and I am not that old. :^) This is in no small part thanks to our technological, medical, and social advances. Think about it, two centuries ago most people were rural farmers.
While I don’t know if we will change quite as radically as people like Kurzweil predict in just a few decades, by the time we have real interstellar vessels I do expect many people to have utilized the available technologies to change what nature gave them, namely their IQs, their weight, their gender, and overall appearance. It’s happening now to the limits of what we can do, so imagine what people will have available to transform in another two centuries. And I am not even talking about things like transferring one’s mind into a machine.
This is why I now look at series like Star Trek as nostalgically entertaining but poor predictions of humanity several centuries hence. Look at the original series from the 1960s and you will quickly find things that are archaic in 2011 let alone 2311: Women weren’t allowed to command starships and in one episode it was naturally assumed that when a female Starfleet officer got married (to a man, of course), she would automatically leave the service and go home to have babies. And despite having some rather fantastic gadgets to play with in the 23rd and 24th Centuries, the people in ST are largely unaffected mentally and especially physically.
This is one reason I like the Web site Orion’s Arm. It depicts life ten thousand years in the future and while it may ultimately be as flawed as any other prediction of tomorrow, at least they are unafraid to take evolutionary ideas to places that Star Trek and most other SF until very recent decades rarely went if ever.
We are only now as a culture just starting to let it sink in about what we could do with our potential. It will and does frighten many, but if we stop being so pessimistic we could quite literally surpass our wildest dreams. Needless to say this would have a very positive effect on galactic voyaging, which includes making a multigenerational star colony or ST type starship look as dated as a spaceship from the Buck Rogers era. It would also make star flight in general that much closer to reality.
When we try to predict the future we are biased towards what we already have, not surprisingly we look for more of what we see around us, a bigger, faster today, rather than being able to foresee radically new technologies.
I think it’s also a bit scary for people to even try to contemplate a future in which machine intelligence far exceeds Man’s, and all that that will bring, though that future seems inevitable now.
Several long-range historical developments have indeed been touched upon by science fiction:
1. Life extension by suspension of senescence.
This will require a biological understanding profound enough to lead to
2. Brain expansion and augmentation.
Cortical expansion of prefrontal ‘thinking areas’, as well as addition of an outer cortical layer comprising neurons specialized for pulsed optical input from transparent skull- microchannels. (Electronic input would involve foreign bodies with risk of rejection.)
Only the immortalized old would be doing this, since it would probably require 20+ years of skull growth and brain integration (doing it to a kid would definitely be child abuse).
Once numerous brain-networked people can share thoughts directly, an even more intense communciation medium will be born. So too will:
3. Collective AI (perhaps based on giant organic brains)
Psychological experiments with our sense of self seem to indicate that enough people sharing enough thoughts could lead to a kind of Mental Internet.
Can we spell B-O-R-G? (Brain-ORganized Gerontocracy)
Don’t say you weren’t warned.
Life extension makes sense only when there is something to do. Having my life extended so that I can be an office drone would be closer to hell on Earth. As far as “improving” the human wetware of our frontal cortex, that is where I have to disagree. It would seem to be a radical change of what we are. If we “wire” ourselves to a network, then our ultimate freedom, to be alone in our thoughts, would disappear. I just don’t see how a species that has been optimized for individual consciousness would surrender that for collective consciousness.
Maybe I am an old coot, but it would appear that we are already too connected as is, where we jump at every beep, bleep and warble our electronic devices throw at us. I really believe that our technology to improve our selves will make us healthier, and longer lived. I just don’t see us surrendering our humanity.
One of the things that I think Trek in all it’s forms got right is that humans will be humans. We might have starships, but humans will always be human, we will eat, drink and shag the hottie space princess if we get the chance. If we weren’t terribly put off by our Neanderthal cousins, could a shapely blue skinned alien be to far out of round for our appetites?
In all seriousness, in my humble opinion, humans will be exactly what we expect them to be. They will be venal and lustful, noble and daring, brilliant and disappointingly stupid, but they will be human.
ljk, largely agree, silicon based AI has an even greater advantage over humans for interstellar work than robotic probes have over crewed missions for sol system work. The “Rise of the Machines” indeed. But philosophically (from a human, watery carbon based life form perspective), what would be the point?
Suppose that 5% c probes can be launched, suppose that there is a habitable world just 20 LY away. Ok, fine. Now we are talking about waiting 400 years for a report from the AI, about a place “we” (biological entities) could not reach. This is no longer a robotic extension of our senses like Dawn and New Horizons, but a complete replacement for human experience. Would we even care to know about unreachable oases for biological life? And would the AI have any curiosity about such things? Why? Maybe the AIs would rather do their own thing with a rubble belt around a bright A star like Sirius …
Assuming that AI is possible (and I see no laws of physics against it), humans are in an evolutionary race which we may very well lose. Agree that normal humans have very little hope of ever reaching the stars. But there are many paths for heroic bioengineering to create biological successors to H sapiens who might be up to the job. We have living monotremes with us now, genomes already under study. One can imagine creating a egg laying human form species. Or alternatively, creating a giant echidna with opposable thumbs, a humanized brain, and a humanized larynx. Such a species could travel to the stars in the form of cryopreserved amniote eggs, young to be suckled by a “lactating” robot. May not be as appealing to current sensibilities as Star Trek. But, from a bio-chauvanist point of view, beats the heck out of ceding space to Hal.
Joy, many humans may not like the idea of our machines exceeding us one day, but that does not mean it isn’t possible or won’t happen. In fact since a good part of evolution is about changing to adapt to new environments, a sophisticated AI (aka Artilect) within a strong, long-lasting. and adaptable “body” makes sense if you want the most efficient way to function in space or deal with a world where the conditions would kill an unprotected human in seconds (everywhere but Earth in this solar system at least).
It is entirely possible that humans evolved to make the Artilects a reality. Read the works by Hugo de Garis, who does not sugarcoat our potential future with such intelligences. He currently resiedes in China, where that government has given him millions to conduct his AI research, after failing to get continued support in various Western nations. Guys like Kurzweil tend to focus on the aspects of AI and nanotechnology that will largely benefit us, such as putting our minds in perfect android bodies so we can live forever (Athena Andreadis expertly takes this idea apart and shows why it will probably remain science fiction in a recent article on her blog in Starshiip Reckless); however, with seven billion people on this planet and rising every second, who and how many will get to live like that (we already know the answer to those questions)?
We are not the end result of human evolution, whether the next stage is biological or artificial. Just as we want our biological children to exceed us with their lives, we should consider a future where our “Mind Children” as robotics expert Hans Moravec calls them will be better suited for the reality beyond our little planet Earth. This includes dealing with alien intelligences, who may also be the artificial result of their biological creators.
Our species and society really is on the edge between greatness and extinction, or irrelevance if we don’t start shaping up and doing what is truly best for ourselves and our descendants.
You are right but you missed something out – Social Networking. Thanks to that you can get tweets from Astronaunts on the ISS and pictures that they have just taken – No waiting for NASA to approve or upload but uncensored, direct to the PC from the ISS as taken a few minutes ago.
Plus you get tweets from mission teams of whats going on. I’ve been following the new horizons check out process and admitedly, some of the things went over my head but it was fascinating to follow it.
And it really doesn’t get much better than that to feel truly connected to whats going on in space. Wow. truly, just wow.
Sometimes the future is under our noses but we miss it. All the AI cultist prognostications are simply extrapolations in cybernetic technology with the major assumption that artificial intelligence is possible. Similarly, the modification of human form via DNA tech is as old as our Homo Erectus ancestors first applying ocre and other substances to change their appearance and animal fur to change their cold weather function.
Real game changers are sorta unpredictable. I’d hazard a swag that something like really understanding what’s up with quantum probability, cosmological origins and the dual wave particle nature of the electron where we are able to experience an infinity of multiverses would be to say the least unexpected and produce quite a different future.
@ljk
I do understand that the silicon might win in the end. But until my personal genome succumbs to entropy, this die hard former biologist will champion the DNA based stuff. We are not all that fragile. Google “tardigrades”, which can survive vacuum, dehydration, freezing, and radiation. We metazoans have a gigayear of evolutionary elegance and self reproduction going for us, as well as a huge untapped potential. There is no obvious physical law preventing the engineering of human descendants as tough as tardigrades.
Eloquently put @Joy.
For my money I think we miss the fact, by its imperceptible banality and near universality, that we’re already half machine or our machines are part human. We’re all cyborgs and the merger of bio & silico could accelerate in ways we can’t imagine. Just because we can currently turn the machines off and put them away doesn’t make the fact they’re a part of our identity any less true.
But the other aspect is the ecology of ideas that we are but small parts in. We dimly perceive the super-organisms we’re mere cells of. What happens when they become self-aware?
Joy, while I agree that there are certainly some tough little buggers like the tardigrade on this planet, these creatures are not attempting to colonize the Sol system or build an interstellar vessel (so far as I know anyway). And while humans are tough and smart in our own way, we are still quite fragile when it comes to environments like space or any other world that we know of at present. We need a lot of technology and resources just to survive out there. We haven’t even tried living more than a few days on the Moon and a year or so in Earth orbit, let alone years on another world.
As for genetically engineering humans to survive and thrive in space and on those other worlds, I have read that it is more difficult than we imagine. Even if we do accomplish this, there will be plenty of ethical and moral issues to deal with too that often seem to be conveniently forgotten in SF stories on this matter and by those who assume such things will just happen one day.
If we do make Homo spacus and they are as smart as us if not smart and far more adaptable, who says they will want to pursue the goals of their creators from Earth? I have brought up the same issue before regarding AIs in starships.
Philw, I find it interesting that you dismiss AI which at least exists in some form now, but accept multiverses for which we have theories but no evidence.
The fact is that any real starship, crewed or otherwise, is going to need a smart machine to run things. Perhaps we can mimic intelligence without needing consciousness, but we need real AI experts to pursue the answers to these ideas. A real interstellar vessel will need advanced AI for a successful mission.
Just personal preference, I would rather have Homo spacus or even Terragens spacus, (with or without AI partners) replace us than be replaced by AIs alone. Nothing could be more abhorrent, immoral, or unethical to me personally that to surrender the cosmos (and likely our only home world as well) to silicon.
I would rather give birth to a giant lagomorph with a humanized brain, if that was what was required. Living descendants would know love, have dreams, experience the full range of biological emotions, they would be our spiritual descendants. AI could never be.
Also the point always glossed over by AI supremacists, AI are more like gametes than organisms. What I mean by that is that living soma are complex disposable bioartifacts created by gametes for the purpose of making more gametes. Send functioning adult organisms to an environment where life is possible, and they will proliferate.
Not so with AI. No matter how brilliant, an AI would have to create an extensive industrial infrastructure to make another copy of itself, or travel with a factory ship and mining equipment, non trivial in the extreme. The soma of an AI is not the CPU or even the robot chassis, it is the mining equipment, the ore refineries, and the factories. So AIs as an emulation of life would not be small things, the mass of their soma would be measured in kilotons at least, probably megatons. What does a semiconductor fabrication plant mass? I know they cost billions of USD.
Yet a single living female animal, with a canister of frozen sperm, could begin the population of a planet
PS: even backwards, rural New Zealand has manufactured dairy goat/human hybrids already, the moral/ethical debate is over and the biologists are moving on
Joy:
This is a very important insight. In life as we know of, there is no connection between the gamete and the mind that we know of. We cannot read our genome, nor change it. Our memories are stored in a completely different way (synaptic connections) from our genome (DNA). This would be different for mechanical life. Mechanical life forms would store their blueprints in the same medium that they store their logs. Some form of flash memory chip, perhaps.
If they were smart enough, they could read, understand, and change their blueprints at will. However, intelligence is not a prerequisite for life, and it is quite possible to imagine self-replicating factories that are dumber than bugs. In fact, I think that is exactly what we will want to build.
You are also right about the mechanical “soma” and its complexity. However, regarding size, I think you do not consider the possibility of miniaturization enough. Reducing the dimensions of all machinery by 10 (toy size) reduces its weight by 1000, along with its resource and energy requirements. Reduce it by 100 (watch size), and you get a million-fold reduction. Ultimately, you may get all the way down to the molecular level, as envisioned by Feynman, Drexler and others. At that point the machines would be far more compact and efficient than their biochemical counterparts.
Personally, I think the first successes on the way to self-replicating machines will come at the toy size level, because it really helps if you can work on things in a garage with screwdriver and soldering iron, using parts available in hobby stores, before the machine is ready to build itself.
Why not?
One last point, strictly for fun: Carbon is not the most common element in the universe. The most common elements besides hydrogen and helium are metals, silicon, and oxygen. It almost seems like all of it is destined to be converted into metal, glass, and ceramics for machinery. The moon, for one, completely barren for carbon based life, would appear like solid cheese to mechanical life forms.
Eniac,
Appreciate all of your thoughts. To me the most glaring defect solid state self replication at any scale is that it has never been demonstrated to be energetically possible. It is still quite controversial as to whether photovoltaic cells produce enough energy during their operating lives to replace the energy expended in their construction, much less create a surplus. This doesn’t even go into acquiring the rare dopants commonly used, such as indium.
Accordingly, if a photovoltaic powered AI factory (of any size scale) were deposited on Luna, it would wear out before it could replicate itself. Stirling cycle solar power might be more promising, but on a body with no native gases, the liberation of oxygen from silicates to produce a working fluid is so energy intensive that this is also a losing strategy. On the megastructure strategy scale, Luna is poor in fissionables, no concentrated ores as on Earth. If fusion existed, of course there is He-3 on Luna, but there is no way that the creation of large lasers or tokamak magnets would be net energy positive. So no, I don’t think that Luna would be cheese to AI.
On the nano scale, ribosomes exist, and work beautifully, assembling organic compounds in aqueous solutions. Drexler’s solid state assemblers remain in the category of wild speculations at best and have been critiqued at length over the last 20 years by people working in the nanotechnology field.
There might not be any law of information theory preventing a solid state emulation of the functioning of the mammalian limbic system. But designing an emotional version of the Turing test to determine if the goal of emulating human subjective experience had been achieved would be guesswork. On the other hand, that our pets and other mammals experience emotions similar to ours is a certainty.
Ljk, there is little doubt that Star Trek will appear a failure if used a predictor of the future, but I’m not as sure as you that its projection of social mores in that future is its bad point. At the time the series was written all of the obvious indicators pointed to an eventual future of total sexual equality, so it is notable that they pointedly disregarded these trends.
Today we have a little more data, and it indicates a quite different future. While we should still predict a future where male and female have equal legal status, and enshrined equal prospects if they wish to peruse a career, the social pressures do not now look as if they will also reflect this equality, nor do the genetics and physiology.
Firstly to differences that look physiological. Current women are displacing men from middle management roles, and this seems to be due to an inherent greater ability to manage people than abstract entities such as organisations (or star ships). Risk aversion seems to also play a role. In the very few western jobs that still entail a high risk of death there remains a very low penetration of women into the workforce. Note that this barrier does not seem nearly so high for jobs that just entail hard physical labor.
Secondly to the likely social background. In the western world fertility rates are way below replacement and continue to plummet. Some, perhaps all, of this trend is due to women perusing careers. Currently the western world can make up the shortfall by immigration from the third world, but this would not be a suitable answer in the world of prosperity envisioned in the Star Trek universe. Surely defaulting to a totalitarian state that has biological control of us is also a poor answer. At the moment, the most moderate answer to this far future problem that springs to mind, is simply to run massive advertising campaigns, and inculcate them in girls schooling, that it is a wonderful (even preferential) option to become a homemaker.
Perhaps you should have written on the 29th , that the original series looks ahead of its time as far as gender roles are concerned, even if it gets all else wrong. Confident predictions (including mine here) tend to make fools of us all.
I’m pleased that gametes have come up in the above carbon v silicon debate, because it is central to an issue that has always bothered me. Any machine that has developed through natural selection has had self-preservation and promotion of its genetic line above all other priorities written into every fiber of its being. With AI we have to implant its priorities and motivations de novo. Why does absolutely everyone assume that the priorities given will entail humanity coming second. Perhaps someone involved in AI could explain to me why you must do this?
Call me a fool, Rob Henry, but I am confident that the real Milky Way galaxy is not like the one depicted in Star Trek. Maybe in some alternate reality assuming there are an infinity of those, but not this one.
And just as I often sense fear when it comes to the topic of advanced alien life forms, I see the same with Artilects. This just shows that the human species has a long way to go before it can successfully interact with an alien mind. That is why I think Artilects will be the ones who do this successfully, in part due to the possibility that some ETI are Artilects too.
Eniac, leaving aside Joys energy considerations, what would be your estimate of the number of parts contained in a minimal ‘solid state’ machine? I’m just wondering since to me it would seem to be astronomical.
For that matter what of the promised nano-revolution. The smaller you make parts the higher the friction, so I have a horrible feeling that the minimum part size for a self replicating machine might be way larger than desirable. Of cause, parts can be made so small that thermal agitation enables their functioning rather than provides friction, but surely this is then life not a nanobot machine??
Joy, these are good points.
I have to call you, though, on the claim that PV solar cells have a negative energy return. I just don’t think it is true. What is your basis for this assertion? From what I have seen, returns are thought to be somewhere around 5-10, which would be quite sufficient. And we are not done improving them, either.
You say mechanical self-replication is not demonstrated. I think our own industry is a pretty good demonstration. We have all the processes worked out that can take us from raw materials to complex machinery. True, we cheated, having used wood, coal and oil which we stole from the biosphere, and still continue to use. I think that we will rapidly replace those by renewable and nuclear sources once they become too expensive, but that part still has to be demonstrated. Remember, though, not long ago all our energy came from wood and animals, sources we have outgrown completely in less than a hundred years.
We can also ask the question: Is mechanical “metabolism” easier or harder than the biochemical version? In some respects, biochemical life has it harder, because the carbon needed for material is much scarcer on Earth than metals. Plants subsist on trace amounts of CO2 in the atmosphere, which is quite a feat. Biological solar energy conversion is also much less efficient than photovoltaics. Despite these odds, the balance works out in the end. I don’t see why it should not work at least as well for metals.
Mechanical “life” also has the opportunity for substantial shortcuts. Much of the energy needed is in the form of process heat. Heat can be directly derived from sunlight at very high efficiency using simple polished aluminum mirrors.
For materials, many parts can be made directly from glass and ceramics, reducing the need for splitting metal oxides (dirt). Biochemistry does not have this luxury. All the biomass on this planet is made from carbon and hydrogen painstakingly split from CO2 and H2O at high energy cost. The primary source of this energy is solar power, converted at very poor efficiency.
Of course, there won’t be proof it is possible until we have done it, but to me things are looking quite promising.
Joy,
I believe emotions are actually much easier to model than intellectual ability, and the Turing test covers both. After all, you are allowed to ask any questions, and test for emotional as well as intellectual responses. If AIs have no emotions, it will not be because they can’t for some fundamental reason (what would that be, anyway?) , it will be simply because we want it that way.
Judging from the current robotics industry, there will be some of each: Dumb robots with strong arms to work in factories, smart disembodied voices and faces to work in customer service, and cute electric animals with apparent emotions who are sold at toy stores. Come to think of it, the future is already happening…
Rob:
A biochemical life form can make do with a few thousand different proteins, plus the usual assortment of small molecules which number in the hundreds, perhaps thousands. Then of course there is the DNA (a single molecule, but very complex), and various RNA species, the most numerous of which are the mRNAs, around 1 per protein. Under 10,000 is probably a decent estimate for the number of different molecules making up the simplest autotrophs. Some non-replicating machines have a much higher parts count, I believe, such as airliners and the space shuttle, where I remember hearing the parts count is in the millions.
Keep in mind, though, that complexity does not imply a high parts count. For example, you can build incredibly complex machinery from just a few different types of Legos. You could also view atoms as parts rather than molecules, making the parts count of a bacterium at most a dozen or two.
For self-replicating machines a low parts count is of obvious advantage, much more so than for airplanes, so my guess would be that we would build them from a relatively small number of very versatile parts, with nuts and bolts of a limited number of different sizes being the most prevalent on one end of the scale, and a handful of different microchips that are highly programmable on the other, with a continuum of more or less specialized parts in between.
As I understand it, it is not higher friction that is the problem. There have been several fundamental issues raised, such as that on the molecular level there is no such thing as friction, which would be bad because machines rely on it. There are energy/heat considerations and the fat fingers problem, all of which have been raised and rebutted, and I don’t think there is much of a consensus on their validity at this point.
In my view the biggest problem is somewhat less fundamental. It is is that with scaling, things change. Electric motors become inefficient at smaller scales, but other ways to convert electricity to motion become better. And so it goes with a whole host of other components. Practically, what that means is if we scale it too much, we have no idea what we are doing and have to design everything from scratch. Plus, we cannot build it manually, as our fingers are too fat, literally. If we stay at the toy scale, we know how to do things, because we are good at making toys. This is not a small advantage, I would estimate the design from scratch to be several orders of magnitude more difficult than building on existing tools and technologies. This is why I think the replicator revolution will be of the clanking type and arrive well ahead of nanotechnology. With luck it will also coincide with (be prompted by and enable?) space industrialization.
Eniac,
A lot of my pessimism about energy generation and optimism about biology was amplified by trying to maintain a sailboat for years. Mold and algae grow everywhere, even in the dark, even in very clean fresh water made by RO, even in the diesel tank, not to mention on the marine ecology on the hull of the boat. Meanwhile, the plastics are UV degraded, the aluminum, bronze, and stainless steel all corrode. The means of generating electricity (wind turbine or photovoltaic) are expensive, and the storage batteries don’t last very long.
As inefficient as life is, we know we can scatter seeds in a sterilized hydroponic container and they will grow in sunlight, and ultimately produce more seeds than were sown. Assuming that there are some non-toxic liquid water planets without incompatible life, a reasonably sized starship could establish a simple terrestrial ecosystem.
I am not convinced that the same will ever be true for silicon. I remember back in the 1970s that cheap thin film PV cells were supposed to be just around the corner then too. The PV industry claims of high EROEI of their products do not pass the smell test, if this were true, why are subsidies needed in this very mature industry? Why can’t even China’s near slave labour system make dirt cheap PV cells in PV powered factories in their deserts (aka bootstrap)? Jeff Vail is actually a lawyer by training rather than a physicist, but published a very cogent critique of the EROEI of PV a few years ago:
http://www.jeffvail.net/2006/11/energy-payback-from-photovoltaics.html
Most of the energy mavens who blog on theoildrum agree with Jeff that the EROEI of PV (if all inputs were counted) is probably circa 1:1. A full PV system EROEI (including storage batteries) is probably < 1.1 due to the much shorter service life of batteries.
The inefficiency of biological usage of light keeps recurring here. This is seemly based on analyses that measure the ‘dark reaction’ efficiency of photosynthesis. Am I the only one that thinks that this is unfair? Wouldn’t it be more accurate to say that life is inefficient at carbon capture?
A thought occurred to me tonight:
the fundamental difference between biological intelligence and machine intelligence can all be explained by the origins: biological intelligence , along with the hardware it runs on, has evolved by direct competition in the environment, while machine calculations have developed via a surrogate: their ability to serve their creators and sponsors namely us. Man is the obligate judge in the evolution of machine : we are the ones who buy the iPADS and pay for clicks on Google, we connect servers into the internet and set the specifications for the next generation of low power computational devices. They in turn evole to meet our needs, we only evolve to survive. Even a complex program like windows 7 will not directly lead to windows 8 – each generation of software has legacy but not inheritance. Until this changes it does not matter if machines are used to aid design of the next generation of chips, if new programs have parts copied form old programs or even if machines are programed to reproduce themselves. To paraphrase- as long as our needs are the “gods” of AI, then they will be trapped in the image of our needs and desires. Ultimate if man can download their personalities into machine they will still be serving us, and if there is then completion between silicon driven personalities, then we will be evolving on a new plane.
I agree with jkittlejr but wish to address a point from above that he did not.
Baring calamities, one day it will be AI’s that create AI’s rather than humanity. As cited above, if they reproduced through general gametes, their priorities and motivations might be thought of as being similar to that of real life forms, yet AI are pointedly different. Surely the question then becomes will they continue to inculcate the motivations and priorities that we gave them in their own AI creations?
If not we may be in trouble, if so we might end up like the mollycoddled humans in Wall-E who had to fight and outsmart their AI before they were allowed to take enough risks to make their lives fulfilling.
Joy,
Jeff Vail makes an argument that there are many things not considered in typical EROEI calculations, but then makes no attempt to show that any of them would change the results significantly. Shipping, for example, is so cheap that we can buy $0.99 toys that came all the way from China. It will hardly make much difference for high value items such as solar panels.
He then proceeds to make the argument that EROEI can be approximated by economic viability (in fact, economic viability is a much stronger requirement than EROEI, as it includes non-energy inputs such as people’s time). Even with that, he just makes it to 1:1, based on 2006 prices. Since then, panels have become a lot cheaper (see for example here: http://www.guardian.co.uk/environment/2011/jun/20/solar-panel-price-drop), and they can be expected to keep getting cheaper some more.
I think I will stick with 5:1 as a conservative estimate, which is good enough.
Rob:
Good point. The losses in photosynthesis are broken down roughly here:
http://en.wikipedia.org/wiki/Photosynthetic_efficiency
and carbon capture probably plays a large role in one of the steps. Overall, however, it seems that carbon capture is not the most significant factor in the low efficiency of photosynthesis. On the other hand, carbon is really much rarer than light in a typical plant environment, so perhaps light energy efficiency has not been selected for as strongly as one might think.
From the reference given, photosynthesis can be 28.2% efficiency at absorbing all solar radiation (if we also consider very hard to use infrared light) when there is no requirement to incorporate new carbon. You will be amazed how efficient life is normally at utilising ATP, so to see its efficiency to dive to less than a third this level when carbon capture is needed to make sugar is a mystery that no one as yet seems to have found adequate explanation for. Unfortunately, life is not optimised for carbon capture on our contemporary Earth, but there is speculation that is was so from the environment in which it first evolved and then somehow became ‘frozen in’ and too complex to change.
Well, I think the article misses quite a number of steps. After the 28% are “collected by chlorophyll”, there are several electron transfer steps, then protons are translocated through the membrane, then ADP phosphorylated. There are bound to be losses associated with each one of these steps, which is glossed over in the article as a collective 68% loss, which is not really bad at all considering the large number of steps involved.
It’s refreshing to see you spread the blame ENIAC, every other man and his dog focuses on RuBisCO these days. This carbon dioxide capturing enzyme is so non-specific that it wastes much of our energy capturing oxygen molecules. I bet your thinking that such a poor mechanism must have instead been selected for its amazingly high activity, but that’s also wrong. RuBisCO has such low activity rates that (by some estimates) it needs to be the most abundant protein in our biosphere. The puzzle then is why its sequence is so highly conserved. So desperate is our plight that attempts have been made to beam it out into space, presumably so some friendly ETI will desperately race to our rescue with something better.
I truth your worries over the complicity of electron transport and phosphorylation are justified but so very much harder to evaluate, especially in regard to finding their peak efficiency in a hypothetical optimised in vivo setting.
I think the question of why RuBisCO is so inefficient is answered quite readily: It is a direct consequence of the fact that carbon availability is the limiting factor for plant growth. If RuBisCO was any better at fixing carbon, it would immediately get to work doing so. This would lower the CO2 content in the atmosphere until CO2 is so rarefied that the new, extra strength RuBisCO would work just as badly as the old one.
I really hope that efforts to improve RuBisCO fail, because now all the plants that still use the old (natural) RuBisCO will starve, leaving only the one lucky species we picked for engineering. Not good for biodiversity. Also, the consequences of lowered atmospheric CO2 may very well include another ice age. Luckily, the enzyme has been so extensively optimized by the billion year struggle to take every last bit of carbon out of the air that it seems very unlikely that we can take it any further, but who knows….
ENIAC, your last comments caused the scales to fall from my eyes. Experimental investigation implies that RuBisCO resists further optimisation, a fact you seem well aware of, but if evolution could find the next step if necessary as you also posit, this would be a wonderful boon for the plants that posses it. As you also foretold they will then sequester a far higher portion of our CO2. If so when all settles down and biodiversity returns, all would look good except for one thing. CO2 levels would be so low that rock weathering could no longer moderate our climate.
ENIAC, our biosphere only happens to posses a single system that couples carbon capture to phototrophy. I am no longer in any doubt that we are staring straight at the anthropic principle here, and change would be a disaster. RuBisCO has made intelligent life possible.
To take this a little further (after the audience has long since left :-) ), there may be a stochastic homeostasis here that has allowed a rich biosphere over billions of years. Early on there was a cool sun, inefficient carbon capture (or none, very early), and a nice balmy temperature due to the greenhouse effect. Every once in a while (stochastically) a genetic innovation leads to better carbon capture, reducing CO2 levels and cooling the planet. If the innovation is too good, temperatures fall too much (Snowball Earth?), biodiversity suffers and with it the pace of genetic innovation. The ever warming sun sooner or later brings back the balmy climate, biodiversity returns and another improvement in carbon capture becomes likely.
I don’t think there is much anthropocentrism involved here, it would work this way with or without us. It would be interesting to test the hypothesis by correlating RuBisCO sequence evolution with climate swings and mass extinctions. We could sequence fossilized protein and/or we could infer ancestral sequence from a large sample of present organisms using phylogenetic techniques. Both, I believe is possible: the former back to perhaps a few million years, the latter over all of genetic history.
It makes me feel sorry for those poor misguided saps who would propose to use technological means to capture carbon from the air. They’ll have a very hard time getting something to work as well as my front lawn, never mind the cost.
Oops, I should have said “our sort of intelligent life” above not “intelligent life” per se
ENIAC, as usual your analysis is spot on, except that you still do not seem able to contextualise RuBisCO against other enzymes. You have done well in analysing the forward context in an evolutionary setting, but the inspiration your comments have given me comes from the back one.
Today’s RuBisCO in abnormally inefficient even in a high CO2 low O2 setting, but if we re-optimise it for those primordial conditions, I now predict that it will have such high efficiency and activity rates as to stagger its current researchers. I will now go out on a limb and say that we could have such a new RuBisCO soon and that the agricultural portions of O’Neil habitats will all use unbreathable gas mixes.
I agree it would be not be difficult to create a less selective RuBisCO that works faster and more efficiently in a high CO2 environment, and it would be harmless. It may not be worth it, though, as the elimination of the CO2 bottleneck will only reveal other bottlenecks that nature has not bothered to optimize away because they were not rate-limiting. It may be better to search for natural high light/high CO2 habitats (not sure if there are any) and see what grows there….
Phobos-Grunt is in trouble:
http://www.planetary.org/blog/article/00003252/