Artificial Intelligence: Curse or Blessing

In a previous post, see here, I presented robotics as one of the technological mega-trends that will shape our coming 100 years. what does that mean to us? Should we be scared? Should we be excited? Will street life change with humans and humanoid robots jointly moving around?

Is this what it will look like?

Answering these questions will take much more than one post, which is good, as it will give me more to write about, and us all more exciting technological questions to ponder. Here I will start to give a rough sketch of the relation between robotics, intelligence and artificial intelligence, shortly AI. Then I will evaluate the direction that AI will go: Will it become a menace, as in the movie The Matrix, or will it become a benign technology that will catapult humanity into the future, as e.g. Ray Kurzweil predicts in The Singularity Is Near: When Humans Transcend Biology?

One particularly well-thought through work regarding AI is Luke Muelhauser’s Facing the Intelligence Explosion (just €0.99 in Amazon’s Kindle store).

Intelligence is defined as the ability of an agent (human, animal, machine) to achieve goals in a wide range of environments. This definition of intelligence aims at achieving goals, i.e. instrumental rationality. Further this definition uses two dimensions: Firstly the degree in which goals are achieved, and secondly the width of the environments.

Already we have realized high levels of goal achievement in machines, think of IBM’s specialized Super-Computer Deep Blue’s success is beating Garry Kasparov in a 6-game match in 1997. There are many examples of highly specialized Expert Systems in the areas of medicine (diagnostics), finance and engineering. However, all these examples have a very high goal achievement rate in a very narrow environment. Deep Blue was very good at playing chess, but was utterly useless at recognizing animal pictures, just as an example.

Recently MIT AI system ConceptNet 4 has been tested on IQ, and tested comparable to a 4-year old child (see here). So on the combination of both goal achievement and width of environments AI still has a long way to go. But it is good to realize that we already have come a long way, thinking of how briefly we have been working on this technology.

Why do we develop AI? First of all because scientists try out things. So they try out if they can make an intelligent machine. But more important is that we need AI, we need robotics. As you read this, you are using AI systems: The operating system of your PC contains AI, your car’s electronics contains AI, most Internet Services that you use make extensively use of AI, your doctor does, and so does the mechanic who takes care of your car. AI supports us achieving goals. AI makes people more productive. And that will be badly needed in the coming 100 years, as population growth on Earth is rapidly declining, and the size of the productive population will decline. So we will need technology to take over from us. At least make each and every one of us very much more productive!

I have been using both terms robotics and AI. They are not the same, but are very closely related. AI is the logic governing the robot, which consists of the AI part (the “brain”) and the mechanical part (the “body”). I concentrate in this post on the AI dimension of this technology.

As pointed out, current AI systems match a 4-year old. We expect that the intelligence of these systems will increase, until, at one moment in time, it will surpass the human level of intelligence. Kurzweil expects this to happen around the year 2040, I think it will take a little longer, but the exact timing is irrelevant for this argument. At one moment in the coming 100 years computers will be more intelligent than human beings. What does this mean? It means that computers will be better able to achieve their goals in a broad environment than humans are.

This realization forces the following question on us: Will the goals aimed at by machines by compatible with the goals of humans, or will these goals be conflicting?

What if the goals are conflicting? E.g. computers are mainly consisting of metals. Metals have a tendency to oxidize in an oxygen-rich environment. Humans need oxygen. See here a clear conflict of interest. Some brainstorming will produce many potential conflicts of interest. Currently we, humans, are clearly superior to these machines in achieving our goals. But what happens when the machines surpass us?

We can view this potential conflict in terms of a Darwinian struggle of the fittest. In the past several thousand years humans have been the fittest on Earth, but that position is about to be taken over by machines.

Evolution?

When you talk with people who have been thinking about AI, robotics and the future, you will find optimists and pessimists. The optimists have faith that robots will be good for us, and will support humanity in its conquest of the cosmos. The pessimists paint a picture in which the stronger species will always eat, kill, destroy, obliterate etc the weaker species. Muelhauser makes a strong point that neither of these scenarios is pre-programmed. Different things can happen and do happen in the Universe, and on Earth. Some of these things we consider good, some of these things we consider bad. When an asteroid makes a close pass to the Earth, we consider that a good thing. When it hits the Earth we consider than bad, as it probably wipes out humanity. Both scenarios are possible. Our concept of Good and Bad has no influence on the probability of events!

In the same way, our preferences regarding the aims that robots and AI should pursue, have no impact on whether AI will be benign or evil by the time it surpasses humans in its IQ. That is, unless we drive the development of AI in a desirable direction, unless we start putting as much effort in AI safety as we are putting in AI capability. This is not an easy topic. Again I refer to Asimov who formulated the three universal laws of robotics (see e.g. Asimov’s I, Robot), reality will be much more complicated than this.

But, as Muelhauser argues, whether AI becomes a blessing or the end of humanity depends on the research and engineering effort that we start putting into this issue as of today.

Good Guys?

Advertisements

Review: Raymond Kurzweil’s “The Singularity is Near”

Human life as we know it is about to change dramatically in the coming 30 years. At least that is what inventor and future author Ray Kurzweil claims in his volume The Singularity Is Near: When Humans Transcend Biology, published in 2010.

First of all I would like to recommend to read this book (you can use the link above to order it at Amazon, either in hard-copy or for your Kindle). But be ready to take some time: The volume is thick and the material sometimes a little chewy. Furthermore, not wanting to be too harsh, but Kurzweil’s writing style is quite dry and requires a certain level of endurance. But like endurance sports: The race is not always easy, but you are happy you did it once you crossed the finish line!

The book forces the reader to think out of the box. Actually, by reading the book I started to realize how much more out-of-the-box I will need to think in order to be able to seriously evaluate the possible scenarios for the future of mankind! This effect makes the book by itself a very valuable read, and completely overshadows any further critical comments I will give in the remainder of this review.

For those who hate spoilers: Stop reading here, and come back after you read the book!

The book presents three main mega-trends and claims that these trends show a long-term exponential growth curve. The trends are Genetics, Nanotechnology and Robotics (GNR). The exponential growth curve for each of these three mega-trends is about to reach what Kurzweil calls “the knee“, after which the growth curves become so steep that the development seems to reach endless acceleration.

Developments in genetics will allow us to repair biological errors, grow new organs and renew our DNA for extended vitality of our cells. Human 2.0 is what Kurzweil calls this.

Developments in nanotechnology will create so-called nanobots that will be able to enter the human body by the billions and do their bio-mechanical repair jobs. Or perform a remote sensing job, either on the battle field or on remote exo-planets.

Robots will be based on strong Artificial Intelligence (AI), which requires further increase of our computation capacity per unit volume. Kurzweil demonstrates that we are currently just a few orders of magnitude away from the required computing capacity to develop strong AI. Researchers at the University of Illinois at Chicago have just applied an IQ test to MIT’s ConceptNet 4 artificial intelligence system, and determined it’s about as smart as a somewhat-challenged four-year-old child (see here).

Kurzweil advocates the use of nanobots to do a detailed scan of the brains from the inside, and than use this map for partial re-engineering to create first strong AI brains. These will then be supporting us in further accelerating the development of computation speed and SW solutions.

Exponential growth of technology

With respect to the exponential character of the growth in each of these technology realms Kurzweil presents a richness of material to support his claim that these technologies grow exponentially. Obviously Moore’s law is the best known example of such a growth curve, but the author presents many more examples.

Based on the four points above the author then predicts that in the years 2030 – 2045 the development in these three sectors will reach an explosive speed and power, resulting in radical innovations in our lives: Human 3.0, being an integration of biological and non-biological components of “me”, nanobots all over the place, and robotic intelligence that will outdo our biological intelligence by a factor of well over a billion. Finally, Kurzweil is adamant about the prediction that within his lifetime we will be able to prevent death, first by delaying it until we have the means to completely abolish it (at least from natural causes).

First a few general remarks on Kurzweil’s style of argumentation. In the book he presents an impressive amount of material on the topics he discusses, with lots of references (not all on the level of academic evidence, but good still). Sometimes he has the tendency to drag in too many topics, to the point where he tends to digress, and thus loose the focus. The famous quote: “Sorry my presentation is so long, I didn’t have the time to make a brief one” is certainly valid for this book!

The second general point I would like to make is that Kurzweil weakens the validity of his argument by presenting hard time predictions regarding the different technology break-throughs, and present these not so much as probabilities, but much more are facts. It is a strong and powerful practice in science to present findings as probabilities, with clear indications of the assumptions that underlie the findings. The presentation of the “Higgs-like particle” on July 3rd 2012 by the CMS- and ATLAS-teams at CERN show a very nice and example of this practice (worthwhile to watch here, it was a historic event!) . This way of presenting scientific findings is not so much driven by humbleness, but by experience that making strong statements based on limited evidence is a sure road to embarrassment. It is admitted that Kurzweil explicitly includes the sensitivity of his predictions on the input parameters, but given that not just the parameters are uncertain, but the complete model, this doesn’t change the point.

Then a few words about the explanatory model that Kurzweil uses for the exponential growth rates of the different technologies and the resulting singularity: There is no explanatory model! The author shows the development of a number of trends, and indeed is able to identify rates that seem to be exponential.

By the way, it would not have hurt to do some explicit statistical hypothesis testing on these claims. But that is a detail.

But the reason why these growth rates are exponential, and what determines their parameters, is lacking. Kurzweil recognizes that technology is developed by humans (at least for the time being) and that social processes drive this growth. But how does that work? Are there social processes that can change the direction of these developments? Or even significantly delay or expedite them?

In his inspiring book Cosmos, Carl Sagan argues that the Ionian Awakening, around the 3rd century B.C. in Greece, with the Great Library of Alexandria at its epicenter, was a surge in intellectual and rational development, that was subsequently overrun by worlds of religion, superstition and provincialism. It took the Western World until the days of Galileo to return to the way forward. How could this have happened? What were the social, political and religious forces that halted our progress? It is an interesting exercise to evaluate what-if scenarios! Where would we be today without the stagnation period between 250 BC until Galileo?

Before we can make hard, probabilistic predictions regarding the future development and adaptation of technological mega-trends, we need to put more effort in creating and testing explanatory models of socio-technological development. This is the social science part of futures studies.

With regard to the main mega-trends that Kurzweil selects, I would agree with development and importance of both genetics and robotics, as I have claimed in one of my previous posts, see here. Definitively Kurzweil is an authority in the field of AI, and his description of the developments in this field, combined with the development of computing power, reads as convincing. Definitively also his detailed and thorough discussion on machine intelligence and machine consciousness is important and convinces.

Genetics is in a tremendous maelstrom, bringing new results and new techniques out at an ever faster pace. No wonder that the worlds of politics and of ethics have trouble in keeping up with the developments. Again Kurzweil writes convincing, and with lots of examples of ongoing research. The reader gets a clear impression that the predictions the author makes are within the same universe as we are today, seem challenging but feasible.

Somehow I cannot say the same of his predictions regarding nanotechnology and especially nanobots. A disclaimer regarding this skepticism is that I know very little about nanobots. I know about the fascination developments in the field of nano-tubes. Amazing physical properties are used for surfaces, touch screens and electronics applications, to name a few. But my imagination fails to picture, within my life time, billions of nanobots scanning my brain. Call it a limitation of my mental capabilities, but until I am convinced with more evidence that this is ongoing I will remain a skeptic!

Nano-buts in my Head

It is mainly his predictions regarding to nanobots that make Kurzweil claim that within his lifetime we will be able to abolish natural death. This is the Holy Grail of humanity, literally! That would also bring it within my lifetime. Wouldn’t that be great. However, I just don’t believe it. This is not a very scientific argument, I know, but as before with respect to the nanobots, I believe we are still a long way off!

As an aside, we need to start thinking about the consequences of further dramatic increases of life expectancy. On Earth the resources to sustain biological life are limited, so either we will need to increase the domain from which we derive our resources, and/or we need to dramatically change our procreation habits and ethics. This will not be easy, as procreation is so deeply ingrained in the fundamental fabric of life.

We will be able to extend the life expectancy, probably step-by-step, as we have been doing for the past 100 years. We can reduce the risks of premature death. But we are not yet able to abolish the effects of aging. People who live past their 100th birthday are both physically and mentally not so fit as we would want to be. We must solve that before we can go realistically beyond our current efforts. So I will follow the developments in this area with great interest, have some hope that the pancreatic cancer that both my parents died of will become curable, and commit myself to healthy food and (very) regular physical and mental exercise as my method for life extension.

The End of Aging

A few elements I find lacking attention in the book. These have to do with evolution, with entropy and with expansion of mankind and her technology into the Universe beyond Earth and the Solar System.

Evolution is a simple yet powerful mechanism. First of all you need something that self-replicates. We call this life. Then we need variation in life. This variation happens, over time. For typical biological very much time. Then you need a struggle for limited resources. And then the automatic result will be survival of the fittest. Only those variants that are best at competing for the scarce resources will survive and be able to reproduce. Evolution as as simple and beautiful as that.

Homo Sapiens has been very good at this survival game. Thus far. Through our brain we have been able to adjust our survival strategies faster than genetic adjustments would allow in other species, and therefore we have been so successful in grabbing the food, and the land, and everything.

Friends?

Now we create hyper-intelligent technology, and I believe that will happen, and we will be able to turn that into a benign development. But we will need to prevent introducing a competition for resources with this superior technology. We might not survive that competition. Asimov’s Three Laws of Robotics most certainly will not do the trick (I, Robot).

The second element I find lacking in the book is the understanding of entropy, or the second fundamental law of thermodynamics. Entropy in any closed system always increases. And our Universe is a closed system, according to all current cosmological models. This means that Kurzweil’s prediction that humanity will take over the Universe and prevent it to end in a entropy death cannot be true. Local order is a statistical variation in a universal system that increases it’s amount of disorder. This is a very slow process, but it is happening, and no one and nothing will be able to stop that process.

Finally I believe that Kurzweil gives insufficient attention to the mega-trend of space-travel, in other words the expansion of humanity beyond the boundaries of planet Earth. As I will write about this mega-trend a lot in future posts I will not elaborate on that here.

Second Law of Thermo-dynamics

In summary, The Singularity Is Near: When Humans Transcend Biology is a valuable piece of positive, optimistic Out-of-the-Box thinking. I can highly recommend it. And my reading advice on it is (as with all books): Do not believe just everything what you read, but make it part of your model of the world that we live in. Enjoy!

Technological Mega-trends 2013 – 2113

Proposing mega-trends is more than just looking in a crystal ball. Mega-trends are mostly identified as extrapolations of past and current developments. And there are more mega-trends than only technological ones. For example social and economical mega-trends are profound and have strong mutual interaction with technological mega-trends. After-all technological development requires funding, and funding requires a certain prioritization: The question, for example, if we should spend our money on  the development of space flight or create better medical care is a valid one.

As I have argued in previous posts, there is ample evidence that investments in big science and in big technology have an accelerator effect on economic activities. For every dollar of community investments in NASA the private sector receives back 7 dollars. Part of that is “just” jobs, but also stimulation of e.g. medical sciences and medical methods is evident. Therefore I have argued in favor of spending big bucks on big science and big technology. And that is what we have been doing for quite a while.

In this post I will list the 7 mega-trends that I see, will explain what they are, and what is the rationale behind each of them. In later posts I will dig deeper into each on them, and will aim to post updates on them as time goes by. It will be interesting to see which of these trends actually materializes, and which mega-trends we completely miss. I also would welcome comments and criticism regarding my top-7!

Looking at where we come from I see the following technological mega-trends for the following 100 years, so until 2115.

Genetics

Since DNA was identified in 1944 as the molecule type that carries the blue print for all life of Earth, scientists have come a long way: Watson and Crick discovered in 1953 the double helix structure of DNA and how the code was copied for generation of proteins and for reproduction. Then we learnt the mechanism of meiosis and mitosis. And in the early 21st century the human genome was completely sequenced. Hard work is ongoing to identify which genes are located exactly where on the genome.

Double Helix DNA

We now understand that the genome contains all the information necessary to generate an organisms that is very similar to the parent organism. We also understand that the genome contains lots of non-functional code, and dysfunctional code. Some of that dysfunctional code is specific to certain individuals. These variants will show certain diseases. Other dysfunctional code is in all the instances of the species, and causes e.g. ageing. Actually aging and dying of individuals is probably very functional for a species as a whole, but that is a discussion we will have in a later post in more detail.

The better we understand the mechanisms of reproduction, both on the level of the species and on the level of the individual cells, the better we will be able to make modifications to the genetic code in order to cure the dysfunctional parts. Inherited diseases are being re-engineered as we speak, genetic limitations that e.g. cause ageing are now under the microscope.

Initial ethical doubts will disappear as the general public will adapt to the principle of genetic improvement and enhancement. Resistance will become a fringe phenomenon, just as resistance to inoculation is today.

Genetic engineering will be used, step by little step, to drive at least three of the other mega-trends mentioned below: Life extension and space flight and terra-forming.


Prosthetics

In our drive to improve the quality of life we use all means available. Apart from genetic engineering and “classical” biochemical pharmacy, more and more are we using non-biological methods of solving biological problems. Obviously this is already a very old approach: think of the wooden legs and hooks of old days, but also of spectacles. But today we are able to create very many non-biological devices that support biological life: from hearing aids to pace makers, from artificial hearts to artificial kidneys, all the way to Deep Brain Stimulation to soften the impact of certain brain defects like epilepsy and Parkinson disease.

Prosthetic hand

We start to understand that there is nothing special about biological mechanisms, apart from them being very complex and very delicate. There are basically no limits to enhancing biological functionality with non-biological technology. We will see step by step an increase in these technologies, so that the lame can walk, the blind can see and the deaf can hear. And we will be able to enhance our brains. We already start to understand how the brain works, and what APIs the brain has, how we can connect our electronic processors to communicate with the chemo-electrical circuitry of the brain. This will be a major development, that will make our intellectual capabilities increase exponentially.


Life Extension

Immortality has been the Holy Grail for as long as we have records of human thinking. Probably immortality would be unbearable, but a significant extension of our lives would be something to strive for. And that is what we have done for centuries, with quite some success. The average life expectancy increased from below 40 less than 200 years ago in Western Europe to over 80 in 2010. Current thinking is that without fundamental developments in genetic engineering and other medical techniques the increase in life expectancy will top out somewhere around 100 – 106 years.

But these advanced methods are already being developed, with successful experiments doubling the lifespan of mice. I expect these developments to receive ample attention and funding, with the actual life expectancy increasing to well beyond 100, and with preservation of quality of life.

A phase in-between life extension, prosthetics and robotics is the field of non-biological life. I will not dive too deep into that right now, but this entails transplanting the complete conscious and subconscious footprint of a biological person into a non-biological mechanism. This is a fascinating field that has been used a lot by SciFi writers, just think of the talking space ships with personality in both Dr. Who and in the Culture books by Iain Banks (The Culture Boxed Set: Consider Phlebas, Player of Games and Use of Weapons).


Robotics

Robotics is popular among SciFi fans, and has a futuristic ring. However, robotics is a simple extension of the millennia long quest to create tools to make our life and work easier. Robots are tools that are controlled by artificial intelligence. Note that we live in a world that is full of robots, but we do not recognize them as such because the do not have a humanoid form. Machine intelligence will increase exponentially, both in the number of devices that contain artificial intelligence, and in the level of intelligence these devices have.

Robots

Most of these robots will not look like humans, or dogs, or anything living. The intelligence will reside for example in our cars (that will themselves prevent accidents) and in our houses (that will optimize for energy consumption and comfort, and will order groceries). Ubiquitous intelligence, to serve you and me.

Will we get human-like robots? Maybe. Because it is cool. Or because it would be a generic tool, that can handle all tools that are designed for handling by humans. In the end this will be a cost-benefit decision. Do not expect intelligent artificial men and women to roam our streets any time soon. But do expect year to year an increasing amount of artificial intelligence all around you!


Space Travel

Space was hot in the ’60s and ’70s. Then it cooled down, after the space race was preliminary won by the USA. The International Space Station, the gorgeous photos from the different space telescopes, and the Mars rovers have warmed up the mood for space again, after a cool-down period of ever decreasing budgets. Definitively also the discovery of hundreds of (exo-) planets around nearby stars has stimulated the taste for space again.

The good thing is that now most countries that were competing with each other during the Cold War are now partners in Space: The IIS is a splendid joint effort by NASA, the Russian space agency, ESA (European Space Agency), Canada and Japan. Only China and North-Korea still consider space as a nationalistic effort, with all the waste that this approach causes. Large scientific and technical enterprises like space travel require big budgets, long and stable commitments, and should probably be handled as global efforts, efforts by humanity as a whole.

Space exploration

When we look at the picture “Pale Blue Dot“, the picture taken by Voyager 1 at a distance of 6 billion km, we start to understand what a tiny place our mother Earth is, less than a grain of sand in an almost endless Universe. The fragility of Earth is so evident. Do we want to wait, like all those extinct species before us, until something hits this planet, and also wipes us out? Or do we want to use the brains that have catapulted us ahead of all other species on our planet, and start with some serious spreading of the risks? I guess the answer to this question is evident. Therefore we need to continue our investigation of the Universe, pick up again with manned and unmanned space flight, and start seriously investigating planets and moons. First inside our own solar system. And after that also towards planets accompanying nearby stars. A pretty good further argumentation for (manned) space exploration is Zubrin’s The Case for Mars: The Plan to Settle the Red Planet and Why We Must. I will return to his argumentation is another post.

The technological developments needed for such interstellar travel are magnificent. But we are working on each one of them. It is a matter of having a direction, and then keep working on it!

Terra-Forming

The previous mega-trend already implied trips to other planets. What to do once we are there? Mars as it is does not support human life – or Earth plants and animals, for that matter. The search for exo-planets that contain liquid water, an atmosphere that humans can breath in (so closely resembling our atmosphere), can take a long, long time. But planets that are located in the habitual zone of it’s host star, and that contain, either in it’s atmosphere or in it’s soil or oceans, the needed substances for a useful atmosphere are probably plenty. All we need to do is go there and create a biologically habitual atmosphere using the technologies we have. We are pretty good at changing atmospheres. We have been doing that for many years on Earth. Here we change the atmosphere in the wrong direction. But we can use the same mechanisms to terra-form a planet.

Terraforming

These processes take time, in the order of magnitude of 100 – 300 years. But what is such a period in the history of men? Let alone in the history of life on Earth? Just imagine that in the time of Newton we would have started the terra-forming process of Mars: She would be a green planet by now!

Terra-forming will take time. However, I included it as one of the mega-trends, because terra-forming, or some form of climatic engineering, will be required to make further colonization of space feasible. So I expect that in the coming century this will probably not so much be actually executed, but will be a strong driver for research and development.

Energy

All we do requires energy, lots of energy. This goes especially for space travel. Inside the Solar system we can probably still manage with existing technologies of propulsion, as soon as we want to go beyond we need some serious oomph! We do understand the physics of potential energy creation: fission, fusion and matter-antimatter annihilator, to mention a few. But we will need to be able to engineer these principles in a workable fashion. That requires a sense of purpose. I again refer to the book by Zubrin, mentioned and linked above.

The mega-trends above may appear far-fetched. But let’s just look back in history. Check out what we have accomplished in the last 100 years that would have appeared fantasy in 1915: computing, men on the Moon, robot rovers in Mars, a completely decoded human genome, to name just a few.

No matter how big an effort may seem to realise these mega-trends, they are small compared with other efforts accomplished by humans, and take just a short time in our history.

Seven mega-trends that will liberate men and women. Liberation from the limits of a short life. Liberation from the limits of biological being. Liberation from the need to do all work ourselves in order to be able to survive. And liberate from the limits of our Earth. See here a program for guiding funding, both public and private, to the relevant R&D programs.

I am looking forward to receive your top-mega-trend list, so we can check that we have the right ones. This list will drive a lot of the writing I am planning for months and years to come on this blog.

A Bright Future?