Some 31 years ago, all the way back in 1986, the futurist Eric K. Drexler published a new book – Engines of Creation, in which he wrote about his vision of molecular assemblers. Those machines – each too small to be seen by the naked eye – were supposed to take atoms and put them together in a myriad of shapes and combinations. While others, like Nobel laureate Richard Feynman, discusssed this idea before, Drexler was the first to fully consider its implications, and has ignited the imagination – and motivation – of millions around the world.
Why are such molecular assemblers so important? Because everything is made of atoms: the earth, our food, our medications and even you and me. That means that these molecular assemblers can be the ultimate recycling machines.
According to Drexler’s vision, the molecular assemblers of the future would be able to –
Take apart feces and urine, and reconstruct them into steaming-hot steaks;
Capture carbon dioxide (a greenhouse gas) from the atmosphere, and use the atoms to create usable plastic or even oil;
Create sheets of perfect diamond (one of the simplest structures in nature – ‘just’ carbon atoms attached to each other;
Create all forms of medication, at everyone’s house.
And as Drexler himself beautifully put it –
“COAL AND DIAMONDS, sand and computer chips, cancer and healthy tissue: throughout history, variations in the arrangement of atoms have distinguished the cheap from the cherished, the diseased from the healthy. Arranged one way, atoms make up soil, air, and water; arranged another, they make up ripe strawberries. Arranged one way, they make up homes and fresh air; arranged another they make up ash and smoke.”
The possibilities and potential are practically endless. Only thing is – thirty one years after Drexler’s burst of genius, molecular assemblers are still a vision which we are striving to fulfill.
The upside is that such assemblers are more than just a figment of the imagination. At some point in the future, we will be able to create them.
How do I know that? Because they already exist.
The Living Assemblers
Consider the cells in your body. Each of them is a wondrous and highly complex machine with one sole purpose: to break down certain molecules into simpler molecules and atoms, and to reconstruct those basic building blocks into more complex molecules that your body can use. These cells, in a very real sense, are molecular assemblers.
And not only that, but some cells can even create incredibly sophisticated molecular structures. Consider, for example, the Surirella spiralis – a unicellular (one-cell) organism, that secretes silica in a highly specific and accurate fashion, and thus creates a unique armor around itself. The size of this ‘spaceship’ that you see in the picture below is around 70 micro-meters, which is about the diameter of a single human hair.
How do cells know to create these wondrous structures, or to place molecules so precisely around themselves? They are programmed to do so, and the program is in their DNA – their genetic code.
Some biochemists are trying to genetically engineer cells to instruct them what to do. But others are trying a different approach, which has gained substantial attention these last two years, especially since three of the leaders in the field have won the 2016 Nobel prize in chemistry for creating actual molecular machines.
The Molecular Machines
Three researchers won the 2016 Nobel prize in chemistry: Jean-Pierre Sauvage, J. Fraser Stoddart, and Bernard L. Feringa. Put simply, those scientists all focused on creating molecular machines: machines the size of molecules, that contain several parts that move in relation to each other.
Sauvage created a molecular machine, one part of which could revolve around the other when fed with energy.
Stoddart created a ‘molecular lift’, that could raise itself above a surface, and later an developed a molecular ‘artificial muscle’ that could even bend a (very) thin gold sheet.
Finally, Feringa built molecular motors, that actually powered a molecular nano-car. The wheels of the car really did revolve, and the entire construct could race across a gold surface, which is about a cool a thing as could be.
Why are all these developments important? Make no mistake: they’re not going to impact our lives in the next decare or so, but as the Nobel committee itself declared–
“In terms of development, the molecular motor is at about the same stage as the electric motor was in the 1830s, when researchers proudly displayed various spinning cranks and wheels in their laboratories without having any idea that they would lead to electric trains, washing machines, fans and food processors.”
In the future – maybe twenty years away from us, but probably more – the evolved versions of these molecular motors will be everywhere. And yes, they’ll be part of the molecular assemblers whose existence Drexler forecast thirty one years ago. We are nowhere near the end of the journey into molecular engineering.
In fact, a new invention from 2017 shows that we’ve already created a working molecular assembler – even though it’s a pretty limited one. A group of biochemists led by Simone Pisano, created a molecular ‘robot’ (as they described it themselves), that could selectively pick up very specific molecules, attach them to certain sites on other molecules, and repeat the action again and again. This is a very small step towards molecular assemblers, but it is a step nonetheless, in an area in which even the tiniest developments require years of painstaking research.
Conclusion: the Molecular Assembelers Are Coming
When we look at the developments in the field, and at the great achievements created by the blind watchmaker – semi-random evolutionary processes – it seems practically a certainty that at some point in the future, we will have molecular assemblers. Many science fiction authors write about that future with great confidence that it will arrive – maybe twenty or thirty years into the future, maybe a hundred or more. But arrive it will, and when it does, the human race will never lack for resources again. Quite literally, we will be able to transmute air to edible meat, and our waste into breathable air, drinkable water and edible food. It will be a period of abundance unlike any in human history before.
And all that began with the detailed vision of one futurist – Eric K. Drexler – thirty one years ago.
If you like reading about future technologies and their impact on society, you can find more posts and answers like this one in my Quora content (where this answer also appeared originally).
Many economists and philosophers are trying to figure out today about the future of work. What will people do once robots and autonomous systems can perform practically all tasks in the workplace better than human beings can?
Well, here’s a little well-known secret: many of us are already unemployed. Many, many more than governmental statistics indicate. We just haven’t realized it yet.
Why? Because plenty of people today are working in bullshit jobs, in the words of anthropologist David Graeber. Here’s what he has to say about bullshit jobs –
“…more and more employees find themselves… working 40 or even 50 hour weeks on paper, but effectively working 15 hours… since the rest of their time is spent organizing or attending motivational seminars, updating their facebook profiles or downloading TV box-sets.”
Think of your own job. If you work at a desk or at an office, some of your days probably look approximately like this:
You come into the office in the morning.
You chat with your co-workers for 15 minutes.
You open your computer and chat with your friends on Facebook for another hour.
You feel compelled to do some work. You open a document you began work on yesterday, work on it for ten minutes, then excuse yourself to check your e-mails, then your facebook again, then read answers on Quora (try my content for more future-related answers!), then play just one game of Solitaire…
…and two hours later, you return to reality and realize you haven’t done any significant work today. You resolve to work harder, immediately after lunch.
Lunch takes an hour.
And then you’re drowsy for yet another hour. Luckily, that’s the time for the weekly departmental motivational seminar, during which you can safely sleep while nodding your head vigorously at the same time and grunting affirmatively.
Finally, you realize with a shuddder that it’s almost the end of the workday. You feel guilty and ashamed, and so, in a concentrated effort of 1–2 hours, you actually SIT DOWN AND WORK.
And the amazing thing is that in those 1–2 hours of work, you actually complete an amount of work that used to require an entire office of secretaries to perform a few decades ago. That’s because you’re using smart and automated tools like Microsoft Office Word, Excel and Powerpoint. These tools increase productivity, so that a single person who is proficient in using them can do more in a shorter period of time.
So why do so many of us still work for eight hours a day? Why do so many people work at jobs that they know are ineffective, and in which they waste their time?
Simply put, because human beings need the illusion of being useful, or at least of doing something with their lives. They need to preserve a veneer of action – even though much of that action throughout the workday is almost entirely fictional.
Now, obviously, many of us do not work at a bullshit job… yet. But bullshit jobs form when productivity increases dramatically, which basically describes any form of work in which automation is going to have an impact. And that means that many of our jobs will become much more… bullshitty… in the future.
So – what would happen when robots take over all of our jobs? My guess is that mankind would just inflate the old jobs so that the work that can be done in ten minutes, will still engage workers for a full day. In short, we’ll all ‘work’ at bullshit jobs.
Here, I made a diagram of what it’ll look like. And you know where I did it? That’s right – at work, while answering questions on Quora.
Meet Omer, my five years old son (in the picture above). He will be remembered for as long as humanity exists.
That’s pretty neat, isn’t it?
Let me explain why. Think of the great inventors, leaders and scientists of ages past: people like Alexander the Great, Isaac Newton, Plato and others. Most of them did not have a personal biographer looking over their shoulders, to record their great deeds. Even for those who did hire such a personal biographer, we only know today what they wanted us to know.
Now consider Omer. He is growing up in a period of time in which he is being monitored continuously. All the pictures I took of him, almost since the very moment he was born, are stored in Google’s and Facebook’s servers, and are being maintained and looked after continuously, so that they will be preserved for a very long time indeed. Every purchase I made for him using my credit card, has been recorded somewhere by a data merchant, and the information was sold to other companies.
As my son grows up, his smartphone will record his activities and health, his electronic devices will keep a close watch over him, and aerial drones in the sky will be able to record his movements on the ground. All of this information will be gathered effortlessly, and will be easily analyzed by AI engines to construct a picture of my son’s life.
So – in the future, we will all be remembered and recognized. Maybe not for our great inventions or prowess in battle, but for our personal, small and intimate stories and achievements. My son will know me – his father – as a real human being, full of nuances and quirks. He will know what I did tonight before going to bed, which websites I visited (yes, even if I used incognito mode – the data is still being retained by my internet service provider and Google), and what made me the man I was. And his kids – my grandchildren – will know my son’s story even better than he will know mine. And so on and on, into future generations.
In a way, I will never die for my son, and neither will you. Our stories will remain here to teach our children the lessons we’ve learned over our lives.
A few years ago I lectured in a European workshop about global risks. Before me lectured one of the World Health Organization (WHO) chief officers, who presented a very interesting graph.
What he showed was basically that life expectancy is expected to keep on rising all over the world, so that by the year 2100 it’s going to reach 85–90 years in high-income countries.
Well, I was pretty astounded about that forecast, which seemed to me extremely pessimistic. I talked with him over lunch, and asked whether this forecast included all of the technologies currently being developed in university labs. I asked how the forecasts would be affected by –
The development of nano-robots that could hold back cancer, coronary thrombosis (heart attack), strokes and other diseases from inside the body;
Sophisticated techniques for genetic engineering, that could produce vaccines against cancer and other diseases;
Tissue engineering techniques that could repair entire tissues – sometimes while they’re still in the body;
Artificial intelligence engines that would provide real-time medical monitoring and consultation much more accurate than that of today’s best medical doctors;
I’m paraphrasing his answer a little, since it all happened a few years ago, but the gist of what he said was –
“No, we can’t take all that into account. The model can’t acknowledge medical breakthroughs. We know that such breakthroughs will have a dramatic impact, but we just don’t know when they’ll emerge from the lab. But I can tell you that if even 15% of the research currently being done in biomedical labs succeeds, then the forecasts will change dramatically.”
So – there is simply no good forecast that will answer the basic question of how long we’re supposed to remain alive in this century. It is entirely conceivable – indeed, even likely, as that WHO official admitted – that sometime in the next few decades, a ‘perfect storm’ of medical breakthroughs will work together to dramatically halt aging and put a stop to most old-age diseases.
Let’s start with a little challenge: which of the following tunes was composed by an AI, and which by an HI (Human Intelligence)?
I’ll tell you at the end of the answer which tune was composed by an AI and which by an HI. For now, if you’re like most people, you’re probably unsure. Both pieces of music are pleasing to the ear. Both have good rhythm. Both could be part of the soundtrack of a Hollywood film, and you would never know that one was composed by an AI.
And this is just the beginning.
In recent years, AI has managed to –
Compose a piece of music (Transits – Into an Abyss) that was performed by the London Symphony Orchestra and received praise from reviewers. [source: you can hear the performance in this link]
Identify emotions in photographs of people, and create an abstract painting that conveys these emotions to the viewer. The AI can even analyze the painting as it is being created, and decide whether it’s achieving its objectives [source: Rise of the Robots].
Create a movie trailer (it’s actually pretty good – watch it here).
Now, don’t get me wrong: most of these achievements don’t even come close to the level of an experienced human artist. But AI has something that humans don’t: it’s capable of training itself on millions of samples, and constantly improve itself. That’s how Alpha Go, the AI that recently wiped the floor with Go’s most proficient players, got so good at the game: it played a few million games against itself, and discovered new strategies and best moves. It acquired an intuition for the game, and kept rapidly evolving to improve itself.
And there’s no reason that AI won’t be able to do that in art as well.
In the next decade, we’ll see AI composing music and even poems, drawing abstract paintings, and writing books and movie scripts. And it’ll get better at it all the time.
So what happens to art, when AI can create it just as easily as human beings do?
For starters, we all benefit. In the future, when you’ll upload your new YouTube clip, you’ll be able to have the AI add original music to it, which will fit the clip perfectly. The AI will also write your autobiography just by going over your Facebook and Gmail history, and if you want – will turn it into a movie script and direct it too. It’ll create new comic books easily and automatically – both the script and the drawing and coloring part – and what’s more, it’ll fit each story to the themes that you like. You want to see Superman fighting the Furry Triple-Breasted Slot Machines of Pandora? You got it.
That’s what happens when you take a task that humans need to invest decades to become really good at, and let computers perform it quickly and efficiently. And as a result, even poor people will be able to have a flock of AI artists at their beck and call.
What Will the Artists Do?
At this point you may ask yourselves what all the human artists will do at that future. Well, the bad news is that obviously, we won’t need as many human artists. The good news is that those few human artists who are left, will make a fortune by leveraging their skills.
Let me explain what I mean by that. Homer is one of the earliest poets we know of. He was probably dirt poor. Why? Because he had to wander from inn to inn, and could only recite his work aloud for audiences of a few dozen people at the time, at most. Shakespeare was much more succesful: he could have his plays performed in front of hundreds of people at the same time. And Justin Bieber is a millionnaire, because he leverages his art with technology: once he produces a great song, everyone gets is immediately via YouTube or by paying for and downloading the song on iTunes.
Great composers will still exist in the future, and they will work at creating new kinds of music – and then having the AI create variations on that theme, and earning revenue from it. Great painters will redefine drawing and painting, and they will teach the AI to paint accordingly. Great script writers will create new styles of stories, whereas the old AI could only produce the ‘old style’.
And of course, every time a new art style is invented, it’ll only take AI a few years – or maybe just a few days – to teach itself that new style. But the human creative, crazy, charismatic artists who created that new style, will have earned the status of artistic super-stars by then: the people who changed our definitions of what is beautiful, ugly, true or false. They will be the people who really create art, instead of just making boring variations on a theme.
The truly best artists, the ones who can change our outlook about life and impact our thinking in completely unexpected ways, will still be here even a hundred years into the future.
I was asked on Quora how the tanks of the future are going to be designed. Here’s my answer – I hope it’ll make you reflect once again on the future of war and what it entails.
And now, consider this: the Israeli Merkava Mark IV tank.
It is one of the most technologically advanced tanks in the world. It is armed with a massive 120 mm smoothbore gun that fires shells with immense explosive power, with two roof-mounted machine guns, and with a 60 mm mortar in case the soldiers inside really want to make a point. However, the tank has to be deployed on the field, and needs to reach its target. It also costs around $6 million.
Now consider this: the Israeli geek (picture taken from the Israeli reality show – Beauty and the Geek). The geek is the one on the left, in case you weren’t sure.
With the click of a button and the aid of some hacking software available on the Darknet, our humble Israeli geek can paralyze whole institutions, governments and critical infrastructures. He can derail trains (happened in Poland), deactivate sewage pumps and mix contaminated water with drinking water (happened in Texas), or even cut the power supply to tens of thousands of people (happened in Ukraine). And if that isn’t bad enough, he could take control over the enemy female citizens’ wireless vibrators and operate it to his and/or their satisfaction (potentially happened already).
Oh, and the Israeli geek works for free. Why? Because he loves hacking stuff. Just make sure you cover the licensing costs for the software he’s using, or he might hack your vibrator next.
So, you asked – “how will futuristic tanks be designed”?
I answer, “who cares”?
But Seriously Now…
When you’re thinking of the future, you have to realize that some paradigms are going to change. One of those paradigms is that of physical warfare. You see, tanks were created to do battle in a physical age, in which they had an important role: to protect troops and provide overwhelming firepower while bringing those troops wherever they needed to be. That was essentially the German blizkrieg strategy.
In the digital age, however, everything is connected to the internet, or very soon will be. Not just every computer, but every bridge, every building, every power plant and energy grid, and every car. And as security futurist Marc Goodman noted in his book Future Crimes, “when everything is connected, everything is vulnerable”. Any piece of infrastructure that you connect to the internet, immediately becomes vulnerable to hacking.
Now, here’s a question for you: what is the purpose of war?
I’ll give you a hint: it’s not about driving tanks with roaring engines around. It’s not about soldiers running and shooting in the field. It’s not even about dropping bombs from airplanes. All of the above are just tools for achieving the real purpose: winning the war by either making the enemy surrender to you, or neutralizing it completely.
And how do you neutralize the enemy? It’s quite simple: you demolish the enemy’s factories; you destroy their cities; you ruin your enemy’s citizens morale to the point where they can’t fight you anymore.
In the physical age, armies clashed on the field because each army was on the way to the other side’s cities and territory. That’s why you needed fast tanks with awesome armanent and armor. But today, in the digital age, hackers can leap straight over the battlefield, and make war directly between cities in real-time. They can shut down hospitals and power plants, kill everyone with a heart pacemaker or an insulin pump, and make trains and cars collide with each other. In short, they could shut down entire cities.
So again – who needs tanks?
I’m not saying there aren’t going to be tanks. The physical aspect of warfare still counts, and one can’t just disregard it. However, tanks simply don’t count as much in comparison to the cyber-security aspects of warfare (partly because tanks themselves are connected nowadays).
Again, that does not mean that tanks are useless. We still need to figure out the exact relationships between tanks and geeks, and precisely where, when and how needs to be deployed in the new digital age. But if you were to ask me in ten years what’s more important – the tank or the geek – then my bet would definitely be on the geek.
If this aspect of future warfare interests you, I invite you to read the two papers I’ve published in the European Journal of Futures Research and in Foresight, about future scenarios for crime and terror that rely on the internet of things.
I’ve done a lot of writing and research recently about the bright future of AI: that it’ll be able to analyze human emotions, understand social nuances, conduct medical treatments and diagnoses that overshadow the best human physicians, and in general make many human workers redundant and unnecessary.
I still stand behind all of these forecasts, but they are meant for the long term – twenty or thirty years into the future. And so, the question that many people want answered is about the situation at the present. Right here, right now. Luckily, DARPA has decided to provide an answer to that question.
DARPA is one of the most interesting US agencies. It’s dedicated to funding ‘crazy’ projects – ideas that are completely outside the accepted norms and paradigms. It should could as no surprise that DARPA contributed to the establishment of the early internet and the Global Positioning System (GPS), as well as a flurry of other bizarre concepts, such as legged robots, prediction markets, and even self-assembling work tools. Ever since DARPA was first founded, it focused on moonshots and breakthrough initiatives, so it should come as no surprise that it’s also focusing on AI at the moment.
Recently, DARPA’s Information Innovation Office has released a new Youtube clip explaining the state of the art of AI, outlining its capabilities in the present – and considering what it could do in the future. The online magazine Motherboard has described the clip as “Targeting [the] AI hype”, and as being a “necessary viewing”. It’s 16 minutes long, but I’ve condensed its core messages – and my thoughts about them – in this post.
The Three Waves of AI
DARPA distinguishes between three different waves of AI, each with its own capabilities and limitations. Out of the three, the third one is obviously the most exciting, but to understand it properly we’ll need to go through the other two first.
First AI Wave: Handcrafted Knowledge
In the first wave of AI, experts devised algorithms and software according to the knowledge that they themselves possessed, and tried to provide these programs with logical rules that were deciphered and consolidated throughout human history. This approach led to the creation of chess-playing computers, and of deliveries optimization software. Most of the software we’re using today is based on AI of this kind: our Windows operating system, our smartphone apps, and even the traffic lights that allow people to cross the street when they press a button.
Modria is a good example for the way this AI works. Modria was hired in recent years by the Dutch government, to develop an automated tool that will help couples get a divorce with minimal involvement from lawyers. Modria, which specializes in the creation of smart justice systems, took the job and devised an automated system that relies on the knowledge of lawyers and divorce experts.
On Modria’s platform, couples that want to divorce are being asked a series of questions. These could include questions about each parent’s preferences regarding child custody, property distribution and other common issues. After the couple answers the questions, the systems automatically identifies the topics about which they agree or disagree, and tries to direct the discussions and negotiations to reach the optimal outcome for both.
First wave AI systems are usually based on clear and logical rules. The systems examine the most important parameters in every situation they need to solve, and reach a conclusion about the most appropriate action to take in each case. The parameters for each type of situation are identified in advance by human experts. As a result, first wave systems find it difficult to tackle new kinds of situations. They also have a hard time abstracting – taking knowledge and insights derived from certain situations, and applying them to new problems.
To sum it up, first wave AI systems are capable of implementing simple logical rules for well-defined problems, but are incapable of learning, and have a hard time dealing with uncertainty.
Now, some of you readers may at this point shrug and say that this is not artificial intelligence as most people think of. The thing is, our definitions of AI have evolved over the years. If I were to ask a person on the street, thirty years ago, whether Google Maps is an AI software, he wouldn’t have hesitated in his reply: of course it is AI! Google Maps can plan an optimal course to get you to your destination, and even explain in clear speech where you should turn to at each and every junction. And yet, many today see Google Maps’ capabilities as elementary, and require AI to perform much more than that: AI should also take control over the car on the road, develop a conscientious philosophy that will take the passenger’s desires into consideration, and make coffee at the same time.
Well, it turns out that even ‘primitive’ software like Modria’s justice system and Google Maps are fine examples for AI. And indeed, first wave AI systems are being utilized everywhere today.
Second AI Wave: Statistical Learning
In the year 2004, DARPA has opened its first Grand Challenge. Fifteen autonomous vehicles competed at completing a 150 mile course in the Mojave desert. The vehicles relied on first wave AI – i.e. a rule-based AI – and immediately proved just how limited this AI actually is. Every picture taken by the vehicle’s camera, after all, is a new sort of situation that the AI has to deal with!
To say that the vehicles had a hard time handling the course would be an understatement. They could not distinguish between different dark shapes in images, and couldn’t figure out whether it’s a rock, a far-away object, or just a cloud obscuring the sun. As the Grand Challenge deputy program manager had said, some vehicles – “were scared of their own shadow, hallucinating obstacles when they weren’t there.”
None of the groups managed to complete the entire course, and even the most successful vehicle only got as far as 7.4 miles into the race. It was a complete and utter failure – exactly the kind of research that DARPA loves funding, in the hope that the insights and lessons derived from these early experiments would lead to the creation of more sophisticated systems in the future.
And that is exactly how things went.
One year later, when DARPA opened Grand Challenge 2005, five groups successfully made it to the end of the track. Those groups relied on the second wave of AI: statistical learning. The head of one of the winning groups was immediately snatched up by Google, by the way, and set in charge of developing Google’s autonomous car.
In second wave AI systems, the engineers and programmers don’t bother with teaching precise and exact rules for the systems to follow. Instead, they develop statistical models for certain types of problems, and then ‘train’ these models on many various samples to make them more precise and efficient.
Statistical learning systems are highly successful at understanding the world around them: they can distinguish between two different people or between different vowels. They can learn and adapt themselves to different situations if they’re properly trained. However, unlike first wave systems, they’re limited in their logical capacity: they don’t rely on precise rules, but instead they go for the solutions that “work well enough, usually”.
The poster boy of second wave systems is the concept of artificial neural networks. In artificial neural networks, the data goes through computational layers, each of which processes the data in a different way and transmits it to the next level. By training each of these layers, as well as the complete network, they can be shaped into producing the most accurate results. Oftentimes, the training requires the networks to analyze tens of thousands of data sources to reach even a tiny improvement. But generally speaking, this method provides better results than those achieved by first wave systems in certain fields.
The Achilles heel of second wave systems is that nobody is certain why they’re working so well. We see artificial neural networks succeed in doing the tasks they’re given, but we don’t understand how they do so. Furthermore, it’s not clear that there actually is a methodology – some kind of a reliance on ground rules – behind artificial neural networks. In some aspects they are indeed much like our brains: we can throw a ball to the air and predict where it’s going to fall, even without calculating Newton’s equations of motion, or even being aware of their existence.
This may not sound like much of a problem at first glance. After all, artificial neural networks seem to be working “well enough”. But Microsoft may not agree with that assessment. The firm has released a bot to social media last year, in an attempt to emulate human writing and make light conversation with youths. The bot, christened as “Tai”, was supposed to replicate the speech patterns of a 19 years old American female youth, and talk with the teenagers in their unique slang. Microsoft figured the youths would love that – and indeed they have. Many of them began pranking Tai: they told her of Hitler and his great success, revealed to her that the 9/11 terror attack was an inside job, and explained in no uncertain terms that immigrants are the ban of the great American nation. And so, a few hours later, Tai began applying her newfound knowledge, claiming live on Twitter that Hitler was a fine guy altogether, and really did nothing wrong.
That was the point when Microsoft’s engineers took Tai down. Her last tweet was that she’s taking a time-out to mull things over. As far as we know, she’s still mulling.
This episode exposed the causality challenge which AI engineers are currently facing. We could predict fairly well how first wave systems would function under certain conditions. But with second wave systems we can no longer easily identify the causality of the system – the exact way in which input is translated into output, and data is used to reach a decision.
All this does not say that artificial neural networks and other second wave AI systems are useless. Far from that. But it’s clear that if we don’t want our AI systems to get all excited about the Nazi dictator, some improvements are in order. We must move on to the next and third wave of AI systems.
Third AI Wave: Contextual Adaptation
In the third wave, the AI systems themselves will construct models that will explain how the world works. In other words, they’ll discover by themselves the logical rules which shape their decision-making process.
Here’s an example. Let’s say that a second wave AI system analyzes the picture below, and decides that it is a cow. How does it explain its conclusion? Quite simply – it doesn’t.
Second wave AI systems can’t really explain their decisions – just as a kid could not have written down Newton’s motion equations just by looking at the movement of a ball through the air. At most, second wave systems could tell us that there is a “87% chance of this being the picture of a cow”.
Third wave AI systems should be able to add some substance to the final conclusion. When a third wave system will ascertain the same picture, it will probably say that since there is a four-legged object in there, there’s a higher chance of this being an animal. And since its surface is white splotched with black, it’s even more likely that this is a cow (or a Dalmatian dog). Since the animal also has udders and hooves, it’s almost certainly a cow. That, assumedly, is what a third wave AI system would say.
Third wave systems will be able to rely on several different statistical models, to reach a more complete understanding of the world. They’ll be able to train themselves – just as Alpha-Go did when it played a million Go games against itself, to identify the commonsense rules it should use. Third wave systems would also be able to take information from several different sources to reach a nuanced and well-explained conclusion. These systems could, for example, extract data from several of our wearable devices, from our smart home, from our car and the city in which we live, and determine our state of health. They’ll even be able to program themselves, and potentially develop abstract thinking.
The only problem is that, as the director of DARPA’s Information Innovation Office says himself, “there’s a whole lot of work to be done to be able to build these systems.”
And this, as far as the DARPA clip is concerned, is the state of the art of AI systems in the past, present and future.
What It All Means
DARPA’s clip does indeed explain the differences between different AI systems, but it does little to assuage the fears of those who urge us to exercise caution in developing AI engines. DARPA does make clear that we’re not even close to developing a ‘Terminator’ AI, but that was never the issue in the first place. Nobody is trying to claim that AI today is sophisticated enough to do all the things it’s supposed to do in a few decades: have a motivation of its own, make moral decisions, and even develop the next generation of AI.
But the fulfillment of the third wave is certainly a major step in that direction.
When third wave AI systems will be able to decipher new models that will improve their function, all on their own, they’ll essentially be able to program new generations of software. When they’ll understand context and the consequences of their actions, they’ll be able to replace most human workers, and possibly all of them. And why they’ll be allowed to reshape the models via which they appraise the world, then they’ll actually be able to reprogram their own motivation.
All of the above won’t happen in the next few years, and certainly won’t come to be achieved in full in the next twenty years. As I explained, no serious AI researcher claims otherwise. The core message by researchers and visionaries who are concerned about the future of AI – people like Steven Hawking, Nick Bostrom, Elon Musk and others – is that we need to start asking right now how to control these third wave AI systems, of the kind that’ll become ubiquitous twenty years from now. When we consider the capabilities of these AI systems, this message does not seem far-fetched.
The Last Wave
The most interesting question for me, which DARPA does not seem to delve on, is what the fourth wave of AI systems would look like. Would it rely on an accurate emulation of the human brain? Or maybe fourth wave systems would exhibit decision making mechanisms that we are incapable of understanding as yet – and which will be developed by the third wave systems?
These questions are left open for us to ponder, to examine and to research.
That’s our task as human beings, at least until third wave systems will go on to do that too.
I was asked on Quora what Google will look like in 2030. Since that is one of the most important issues the world is facing right now, I took some time to answer it in full.
Larry Page, one of Google’s two co-founders, once said off-handedly that Google is not about building a search engine. As he said it, “Oh, we’re really making an AI”. Google right now is all about building the world brain that will take care of every person, all the time and everywhere.
By 2030, Google will have that World Brain in existence, and it will look after all of us. And that’s quite possibly both the best and worst thing that could happen to humanity.
To explain that claim, let me tell you a story of how your day is going to unfold in 2030.
2030 – A Google World
You wake up in the morning, January 1st, 2030. It’s freezing outside, but you’re warm in your room. Why? Because Nest – your AI-based air conditioner – knows exactly when you need to wake up, and warms the room you’re in so that you enjoy the perfect temperature for waking up.
You go out to the street, and order an autonomous taxi to take you to your workplace. Who programmed that autonomous car? Google did. Who acquired Waze – a crowdsourcing navigation app? That’s right: Google did.
After lunch, you take a stroll around the block, with your Google Glass 2.0 on your eyes. Your smart glasses know it’s a cold day, and they know you like hot cocoa, and they also know that there’s a cocoa store just around the bend which your friends have recommended before. So it offers to take you there – and if you agree, Google earns a few cents out of anything you buy in the store. And who invented Google Glass…? I’m sure you get the picture.
I can go on and on, but the basic idea is that the entire world is going to become connected in the next twenty years. Many items will have sensors in and on them, and will connect to the cloud. And Google is not only going to produce many of these sensors and appliances (such as the Google Assistant, autonomous cars, Nest, etc.) but will also assign a digital assistant to every person, that will understand the user better than that person understands himself.
I probably don’t have to explain why the Google World Brain will make our lives much more pleasant. The perfect coordination and optimization of our day-to-day dealings will ensure that we need to invest less resources (energy, time, concentration) to achieve a high level of life quality. I see that primarily as a good thing.
So what’s the problem?
Here’s the thing: the digital world suffers from what’s called “The One Winner Effect”. Basically it means that there’s only place for one great winner in every sector. So there’s only one Facebook – the second largest social media network in English is Twitter, with only ~319 million users. That’s nothing compared to Facebook’s 1.86 billion users. Similarly, Google controls ~65% of the online search market. That’s a huge number when you realize that competitors like Yahoo and Bing – large and established services – control most of the rest ~35%. So again, one big winner.
So what’s the problem, you ask? Well, a one-winner market tends to create soft monopolies, in which one company can provide the best services, and so it’s just too much of a hassle to leave for other services. Google is creating such a soft monopoly. Imagine how difficult it will be for you to wake up tomorrow morning and migrate your e-mail address to one of the competitors, transfer all of your Google Docs there, sell your Android-based (Google’s OS!) smartphone and replace it with an iPhone, wake up cold in the morning because you’ve switched Nest for some other appliance that hasn’t had the time to learn your habits yet, etc.
Can you imagine yourself doing that? I’m sure some ardent souls will, but most of humanity doesn’t care deeply enough, or doesn’t even have the options to stop using Google. How do you stop using Google, when every autonomous car on the street has a Google Camera? How do you stop using Google, when your website depends on Google not banning it? How do you stop using Google when practically every non-iPhone smartphone relies on an Android operating system? This is a Google World.
And Google knows it, too.
Google Flexes it’s Muscles
Recently, around 200 people got banned from using Google services because they cheated Google by reselling the Pixel smartphone. Those people woke up one morning, and found out they couldn’t log into their Gmail, that they couldn’t acess their Google Docs, and if they were living in the future – they would’ve probably found out they can’t use Google’s autonomous cars and other apps on the street. They were essentially sentenced to a digital death.
Now, public uproar caused Google to back down and revive those people’s accounts, but this episode shows you the power that Google are starting to amass. And what’s more, Google doesn’t have to ban people in such direct fashion. Imagine, for example, that your website is being demoted by Google’s search engine (which nobody knows how it works) simply because you’re talking against Google. Google is allowed by law to do that. So who’s going to stand up and talk smack about Google? Not me, that’s for sure. I love Google.
To sum things up, Google is not required by law to serve everyone, or even to be ‘fair’ in its recommendations about services. And as it gathers more power and becomes more prevalent in our daily lives, we will need to find mechanisms to ensure that Google or Google-equivalent services are provided to everyone, to prevent people being left outside the system, and to enable people to keep being able to speak up against Google and other monopolies.
So in conclusion, it’s going to be a Google world, and I love Google. Now please share this answer, since I’m not sure Google will!
Note: all this is not to say that Google is ‘evil’ or similar nonsense. It is not even unique – if Google takes the fall tomorrow, Amazon, Apple, Facebook or even Snapchat will take its place. This is simply the nature of the world at the moment: digital technologies give rise to big winners.
I was recently asked to write a short article for kids, that will explain what is “The Singularity”. So – here’s my shot at it. Let me know what you think!
Here’s an experiment that fits all ages: approach your mother and father (if they’re asleep, use caution). Ask them gently about that time before you were born, and whether they dared think at that time that one day everybody will post and share their images on a social network called “Facebook”. Or that they will receive answers to every question from a mysterious entity called “Google”. Or enjoy the services of a digital adviser called “Waze” that guides you everywhere on the road. If they say they figured all of the above will happen, kindly refer those people to me. We’re always in need of good futurists.
The truth is that very few thought, in those olden days of yore, that technologies like supercomputers, wireless network or artificial intelligence will make their way to the general public in the future. Even those who figured that these technologies will become cheaper and more widespread, failed in imagining the uses they will be put to, and how they will change society. And here we are today, when you’re posting your naked pictures on Facebook. Thanks again, technology.
History is full of cases in which a new and groundbreaking technology, or a collection of such technologies, completely changes people’s lives. The change is often so dramatic that people who’ve lived before the technological leap have a very hard time understanding how the subsequent generations think. To the people before the change, the new generation may as well be aliens in their way of thinking and seeing the world.
These kinds of dramatic shifts in thinking are called Singularity – a phrase that is originally derived from mathematics and describes a point which we are incapable of deciphering its exact properties. It’s that place where the equations basically go nuts and make no sense any longer.
The singularity has risen to fame in the last two decades largely because of two thinkers. The first is the scientist and science fiction writer Vernor Vinge, who wrote in 1993 that –
“Within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended.”
The other prominent prophet of the Singularity is Ray Kurzweil. In his book The Singularity is Near, Kurzweil basically agrees with Vinge but believes the later has been too optimistic in his view of technological progress. Kurzweil believes that by the year 2045 we will experience the greatest technological singularity in the history of mankind: the kind that could, in just a few years, overturn the institutes and pillars of society and completely change the way we view ourselves as human beings. Just like Vinge, Kurzweil believes that we’ll get to the Singularity by creating a super-human artificial intelligence (AI). An AI of that level could conceive of ideas that no human being has thought about in the past, and will invent technological tools that will be more sophisticated and advanced than anything we have today.
Since one of the roles of this AI would be to improve itself and perform better, it seems pretty obvious that once we have a super-intelligent AI, it will be able to create a better version of itself. And guess what the new generation of AI would then do? That’s right – improve itself even further. This kind of a race would lead to an intelligence explosion and will leave old poor us – simple, biological machines that we are – far behind.
If this notion scares you, you’re in good company. A few of the most widely regarded scientists, thinkers and inventors, like Steven Hawking and Elon Musk, have already expressed their concerns that super-intelligent AI could escape our control and move against us. Others focus on the great opportunities that such a singularity holds for us. They believe that a super-intelligent AI, if kept on a tight leash, could analyze and expose many of the wonders of the world for us. Einstein, after all, was a remarkable genius who has revolutionized our understanding of physics. Well, how would the world change if we enjoyed tens, hundreds and millions ‘Einsteins’ that could’ve analyzed every problem and find a solution for it?
Similarly, how would things look like if each of us could enjoy his very own “Doctor House”, that constantly analyzed his medical state and provided ongoing recommendations? And which new ideas and revelations would those super-intelligences come up with, when they go over humanity’s history and holy books?
Already we see how AI is starting to change the ways in which we think about ourselves. The computer “Deep Blue” managed to beat Gary Kasparov in chess in 1997. Today, after nearly twenty years of further development, human chess masters can no longer beat on their own even an AI running on a laptop computer. But after his defeat, Kasparov has created a new kind of chess contests: ones in which humanoid and computerized players collaborate, and together reach greater successes and accomplishments than each would’ve gotten on their own. In this sort of a collaboration, the computer provides rapid computations of possible moves, and suggests several to the human player. Its human compatriot needs to pick the best option, to understand their opponents and to throw them off balance.
Together, the two create a centaur: a mythical creature that combines the best traits of two different species. We see, then that AI has already forced chess players to reconsider their humanity and their game.
In the next few decades we can expect a similar singularity to occur in many other games, professions and other fields that were previously conserved for human beings only. Some humans will struggle against the AI. Others will ignore it. Both these approaches will prove disastrous, since when the AI will become capable than human beings, both the strugglers and the ignorant will remain behind. Others will realize that the only way to success lies in collaboration with the computers. They will help computers learn and will direct their growth and learning. Those people will be the centaurs of the future. And this realization – that man can no longer rely only on himself and his brain, but instead must collaborate and unite with sophisticated computers to beat tomorrow’s challenges – well, isn’t that a singularity all by itself?
A few months ago I wrote in this blog about the way augmented reality games will transform the face of the gaming industry: they’ll turn the entire physical world into a gaming arena, so that players would have to actually walk around streets and cities to take part in games. I also made a forecast that players in such games will be divided into factions in order to create and legitimize rivalries and interesting conflicts. Now Pokemon Go has been released, and both forecasts have been proven true immediately.
By combining the elements of augmented reality and creating factions, Pokemon Go has become an incredibly successful phenomenon. It is now the biggest mobile game in U.S. history, with more users than Twitter, and more daily usage time than social media apps like WhatsApp, Instagram or Snapchat. One picture is worth a thousand words, and I especially like this one of a man capturing a wild Pidgey pokemon while his wife is busy giving birth.
But is the game here to stay? And what will its impact be on society?
Pokemon Go in the Long Haul
It’s no wonder Pokemon Go has reached such heights of virality. Because of the game’s interactions with the physical world, people are being seen playing it everywhere, and in effect become walking commercials for the game. Pokemon Go also builds on the long history – almost twenty years – of pokemon hunting which ensures that anyone who’s ever hunted pokemon just had to download the app.
Will the game maintain its hype for long? That’s difficult to answer. Dan Porter, one of the creators of Draw Something – a game that garnered 50 million downloads in just 50 days – wrote a great piece on the subject. He believes, in short, that the game is a temporary fad. It may take a year for most people to fall off the bandwagon, so that only a few millions of the hardcore gamers will remain. That’s still an impressive number, but it’s far from the current hype. As he says –
“For the casual Pokemon Go player, the joy of early play I believe will eventually be replaced by gyms that are too competitive and Pokemon that are too hard to find.”
I agree with his analysis, but it does depend on one important parameter: that the game does not undergo evolution itself, and continually readapt itself to different groups of users. Other social games, like World of Warcraft, have successfully undergone this transition to maintain a large user base for more than a decade. Niantic may be able to do that, or it may not. In the long haul it doesn’t matter: other, more successful, AR games will take over.
Pokemon Go is bringing in a lot of revenues right now, with the estimates ranging from $1 to $2.3 million a day. Some analysts believe that the game could pull in a billion dollars a year once it is launched worldwide. That’s a lot of money, and every half-decent gaming company is going to join the race for AR very soon. It could be Blizzard that will recreate Starcraft’s fame in an AR fashion, with teams running around buildings, collecting virtual resources and ambushing each other. Or maybe Magic the Gathering or Hearthstone, with players who can collect thousands of different cards in Hearthstops around the world much like Pokemon Go, and use them to build decks and fight each other. Heck, I’d play those, and I bet so would the tens of millions of gamers whose childhood was shaped by these games. The dam gates, in short, have been broken open. AR games are here to stay.
And so we must understand the consequences of such games on society.
A Whole New World
Pokemon Go is already starting to change the way people interact with each other. I took the following picture from my house’s window a few days ago, depicting several people walking together, eyes on their phones, without talking with each other – and yet all collaborating and being coordinated with each other. They were connected via the layer of augmented reality. In effect, they were in a world of their own, which is only tenuously connected to the physical world.
In Australia, a hastily advertised Pokemon Go meeting has brought together 2,000 players to a single park, where they all hunted pokemon together. And coffee shops-turned-gyms around the world have suddenly found themselves buzzing with customers who came for the win – and stayed for the latte. And of course, the White House has been turned into a gym, with all three Pokemon Go teams competing over it.
The game has made people to go to places they would not ordinarily go to, in their search for pokemon. As a result, at least two dead bodies have been discovered so far by players. If you watch players walking on the streets, you’ll also notice their peculiar pattern of movement: instead of following the road, they’ll periodically stop, check their smartphones and change course – sometimes making a U-turn. They’re not following the infrastructure in the physical world, but rather obeying a virtual infrastructure and entities: pokestops and pokemon.
And that’s just a sign of what’s coming, and of how power – the power to influence people and their choices – is starting to shift from governments to private hands.
The Power Shift
What is power? While many philosophers believed that governments had power over their citizens because of their ability to mobilize policemen, the French philosophers Louis Althusser and Michel Foucault realized that the power control mechanisms are inherent in society itself. Whenever two people in a society exchange words with each other, they also implicitly make clear how each should behave.
Infrastructure has the same effect over people. And has been used since time immemorial as a mechanism for directing the populace. For a very long time now, governments used to control the infrastructure in urban places. Governments paved roads, installed traffic lights and added signs with streets names. This control over infrastructure arose partly because some projects, like road paving, are so expensive but also because things like traffic lights and signs have an immense influence over people’s behavior. They tell us where we’re allowed to go and when, and essentially make the government’s decisions manifest and understandable for everyone. There’s a very good reason that I couldn’t erect a new traffic sign even if I wanted to.
But now, with Pokemon Go, the gaming industry is doing just that: it’s creating an alternative virtual reality that has new rules and different kinds of infrastructures, and merges that virtual reality with our physical one so that people can choose which to obey.
Is it any wonder that authorities everywhere are less than happy with the game? It has fatwas being issued against it, religious leaders wanting to ban it, Russian politicians speaking against it, and police and fire explaining to citizens that they can’t just walk into jails and fire stations in their search for pokemon.
In the long run, Pokemon Go and AR in general symbolizes a new kind of freedom: a freedom from the physical infrastructure that could only be created and controlled by centralized governments. And at the very same time, the power to create virtual infrastructure and direct people’s movement is shifting to the industry.
What does that mean?
In the short term, we’re bound to see this power being put to good use. In the coming decade we’ll see Pokemon Go and other AR games being used to direct people where they could bring the most good. When a kid will get lost in the wilderness, Niantic will populate the area with rare pokemon so that hundreds and thousands of people will come search for them – and for the child too. Certain dangerous areas will bear virtual signs, or even deduct points from players who enter them. Special ‘diet’ pokemon will be found at the healthy food sections in stores.
In the long run, the real risk is that the power will shift over to the industry, which unlike the elected government does not have any built-in mechanisms for mitigating that power. That power could be used to send people to junk food stores like McDonald’s, which as it turns out is already in partnership with Pokemon Go. But more than that, AR games could be used to encourage people to take part in rallies, in political demonstrations, or even simply to control their movement in the streets.
This power shift does not necessarily have to be a bad thing, but we need to be aware of it and constantly ask what kind of hidden agendas do these AR games hold, so that the public can exercise some measure of control over the industry as well. Does Pokemon Go encourage us to visit McDonald’s, even though it ultimately damages our health? Well, a public outcry may put a stop to that kind of collaboration.
We’ve already realized that firms that control the virtual medium, like Facebook, gain power to influence people’s thinking and knowledge. We’ve also learned that Facebook has been using that power to influence politics – although in a bumbling, good-natured way, and seemingly without really meaning to. Now that the physical world and the virtual world become adjoined, we need to understand that the companies who control the virtual layer gain power that needs to be scrutinized and monitored carefully.
Pokemon Go is not going to change the world on its own, but it’s one of the first indicators that can tell us how things are about to change when physical reality is augmented by virtual ones. The critical question we must ask is who controls those added layers of reality, and how can we put constraints on the power they gain over us. Because we may end up controlling all the pokemon, but who will gain control over us?