But I have more.

FROM 30 YEARS TO 10 MINUTES
I also have the ability to search for almost anything. I can type a few words into Google and find the prices of whale oil in the 19th century.

Just think about that for a second. Within the span of about 30 years we took a process that would involve me doing days of research in a library, then hours of editing and writing, submitting the documentation, editing it again, sending it to press, putting it out for delivery and getting it into your hands… to a process that takes a few minutes.

That is staggering progress.

CREATIVITY IS SAFE… OR IS IT?
But it’s just the beginning.

Companies like Automated Insights promise to push the envelope such that most articles don’t even need the human to do any of the work. Actually, this is already happening, if you Google “This story was generated by automated insights” you will find press releases like this one which were done completely without human interaction.

Ok… you might not be impressed because obviously creativity is well off in the future. If you think so read this poem and be honest about how easily you can tell it was written by AI.

Flow Machines is a project which is attempting to have AI compose music. I think this area still needs a lot of advancement and the criticism is that all it REALLY does is look at millions of different types of music to look for certain patterns, then puts them into different combinations and tries to find ones that seem to “work.” But really… if we look at the evolution of classical music from Bach to Rachmaninoff, aren’t they more or less doing the same thing? Borrowing from previous structures, weaving in the local folk tunes and then discovering new forms that seem to resonate in our minds when we hear them? They just do it slowly and clumsily and we like to think of that pain as being critical to the process… well… maybe it is, maybe not.

WHERE TO?
I think all this stuff is neat and it makes for interesting headlines, but where is it going?

To answer that question I want to take us back many thousands of years to when humans were first getting started.

Imagine a time when the sky was pitch black at night and the eerie animal sounds echoing in the blackness mixed and mingled with our minds to create the early myths and dreams. During the day we would have to spend nearly all our energy on survival. We had to hunt by hand, tear apart the meat, create the fire, cook the meat and eat it. We had to pick the berries and the fruit (after finding or growing it). We had to till the soil, find and carry the water, build the shelters… all by hand. The work was relentless and the prospects for survival were dubious.

Over time we learned that some of this work could be improved by making use of technology. We learned that some animals could be tamed, we learned that round trees could be used to make moving heavy stone easier. We built aqueducts to carry our water in and sewers to carry our waste out. We learned about boiling water to kill germs and we learned that the cure for bacteria was mold.

Up until very recently the advances in technology have been our aids. Our relationship with technology was primarily the technology doing and us telling… doing it better, cheaper, safer, faster.

What we have today is a bit different.

DRIVING
When I step into my car and go somewhere the first thing I do is type the address into my GPS. The car (or phone) usually quickly tells me the directions that I should follow to go there. In essence the relationship is that the car is telling ME what to do and I am merely the executor of it’s commands. Now you might say… “BS Hermann, YOU told it where YOU wanted to go.”

But did I? If I want to figure out WHERE I want to go, how do I do that? Well, I once again go to Google and type in something like “indoor activities for children” click on maps and boom; a bunch of things to choose from. I could say that *I* originally decided that indoor children’s activities is what *I* wanted to do and that was the real active choice… the rest was the computer passively aiding me… but I’m not so sure our relationship with technology is so clear cut.

My choices are shaped by technology to an alarming degree. With a bit of information about my tastes and preferences AI can predict which things I like and dislike. Indeed this is the basis behind most of the consumption technologies in the world today. Removing human biases from the evaluation process proves to improve the predictive results of what people will “like.”

AI SAYS – YOU DO
I believe this will be an underlying trend over the next few decades. The new advances in mechanization will be computers telling us what to do and we will take on the roles of executors. I think stories like Watson diagnosing cancer in a particular case in Japan better than doctors; will become more frequent.

It’s worth digging in a bit here to understand WHY Watson was able to do this.

Diagnosing cancer is a very complex problem. Symptoms can look like other diseases, the patient’s genetics and history are significant, similar cases can be helpful in finding patters and the amount of data coming from research is staggering. It’s impossible for a human to have all that information in their minds and then process it in an efficient way. Previously AI would “brute force” it’s way to finding solutions to puzzles with known complexity.

Watson and the new AI technology is doing something very different. It is able to learn how to pattern match based on looking at millions of units of information across different dimensions and then figuring out which answer has the highest probability of being correct. It can then have that probability tested in the real world and generate data that it can use to improve it’s diagnosis in the future. This is what human oncologists do… but they are much slower and less accurate. Of course, humans are creative and have “instincts” and can thus short cut many potential possibilities that are “silly.” But Watson and AI like it can learn those as well… without having the biases and emotional limitations. They don’t need sleep or get hungry or have arguments with their spouses. Of course, they probably COULD if that helped improve diagnostic outcomes.

If you think this is limited to using the past to understand the present… that is only the beginning. Back in 2009, an AI system use basic math and huge data sets to “discover” the laws of motion.T

oday, there is lots of research into prediction. For example, this AI tries to generate video that will predict what happens next. In this system there is a so called “adversarial” process in place. One AI attempts to predict what will happen and the other tries to determine if it looks real or fake. By taking data of what really happened both the prediction AI and the evaluation AI can improve their accuracy… all without any human intervention.

FAST FORWARD
Let’s really stretch possibilities here for a second.
With this kind of massive data set, adversarial networks, random tinkering, complete spectrum analysis (remember, AI can see EVERYTHING not just the limited colors we can see) what is possible?

Well… maybe AI can discover a cure for diseases. By having access to humans to experiment on, past data about a disease and access to unlimited chemicals to mix and match; it’s entirely possible.

AI could discover new recipes. By mixing all the possible ingredients and comparing that against data sets of what flavors come from where, what mixes are preferred and how they interact; AI could one day generate recipes… oh wait… that already happened, you can pick it up on Amazon here.

Perhaps AI could create the next breakthrough in computers. Or figure out how to make our screens bigger, more flexible, more durable.

Maybe it could help us figure out how to explore space or break through the speed of light.

Who knows.

BUT WHAT ABOUT ME?
Indeed, when compared to creating a new digital device or breaking Einstein’s laws of physics, writing this article seems trivial. It seems to me that in the coming decades there may be little work in content creation by humans. Perhaps I would need to create a theme or a series of interests so that the AI has something to start with, but I can imagine it might be able to generate those on it’s own as well.

What *I* have to do is consume the content and render my verdict. Do I “like” it or “ignore” it. Based on the the AI can learn what humans like me enjoy and what we don’t. But really… that’s kind of what Bloggers do today anyway… just much more slowly and less accurately.

The role of humans in the centuries to come may very well be consumers and products. In order for AI to advance it needs to have an external reality to bounce things off. That is, until it figures out that improving lives of humans might not be the best goal. As long as AI is focused on making our lives, better, easier, faster, safer… I can imagine a world where humans are essentially products which AI producers trade and compete for. We would basically be real world sims.

If AI eventually correctly interprets evolutionary history, it’s entirely possible it will quite rationally conclude that humans are just another species in the long chain of life over the years. While the extinction of Dinosaurs was a boon for Mammals, I’m not sure the Dinosaurs saw things the same way. Perhaps AI spending all of the worlds resources to keep humans satisfied and safe will act as a limitation on evolution. The next logical step is to just get rid of that problem altogether… in the same way that we have wiped out many species that got in our way.

It’s a scary thought, and not as far off in terms of science fiction as we would like to think. We always like to imagine the future in terms of how *WE* will be part of it, but it’s equally, or perhaps more, likely that the future will be without us at all.

*This article was not generated by AI, yet….*

AI is a hot topic these days.

Everything from self driving cars to robotic surgery seems to be making headlines.

Along with the amazement of an automated future, something troubling has simultaneously arisen.

If AI can do everything humans can do, but better… what do the humans do?

Many articles discuss how humans will potentially become more creatively focused, we will awaken our intrinsic motivations instead of constantly focusing on extrinsic rewards. Switzerland is voting on a referendum to give every citizen $2600/mo tax free. While unlikely to pass, other cities are already experimenting with the idea of “basic income.”

Those experiments and ideas are not directly related to AI per se, but they do reflect a long term trend in machines taking over work from humans.

What started in the industrial revolution in manufacturing has slowly been swallowing up more and more work. This shows up as a general term of “efficiency” but I think what it really represents is replacing human effort with non-human effort.

Whether, and to what extent, this is “good” or “bad” is subject to interpretation. On the one hand having a machine that can harvest corn is far more efficient than having hundreds of people doing it by hand. It also frees those people from the repetitive and uncomfortable job of harvesting corn by hand. However, if your life is dependent on you harvesting corn by hand, this replacement is a mixed blessing at best.

THE SPEED OF REPLACEMENT

Something that makes these transitions even more painful is that they tend to happen quite quickly. In the US, for example, we have seen the manufacturing workforce shrink in size and reduce in costs while increasing output. This happened within a few decades leaving many millions of people without a way to replace the income they had from the previous occupations, regardless of how dangerous, repetitive and unrewarding they may have been. One could argue that there are people who genuinely love this kind of work, but I suspect for most it was a comparatively good income that was the main attraction.

It’s easy to come up with simple and cavalier solutions.

Free from the toil of tedious manufacturing work, they are free to train for new, more exciting positions. But practically how does this work? Without the income from the previous job how does one pay for retraining? And even if one had the means to retrain, how will one find the time while having another position to keep one’s life going? And of course there are the legion practical difficulties of relearning something late in one’s career.

We can also take the position that market forces will work this out. Sure, a few people will be victims of technological progress, but from a broader social point of view, it’s a small price to pay. Well… maybe.

THE BREADTH OF REPLACEMENT

Up until now, technology has primarily focused on replacing physical jobs. Almost all technology from the industrialization period on (and before) have focused on replacing physical human effort with machines. The wheel, the lever, the printing press, Railroads, automobiles and trucks, robots on assembly lines, and so on have slowly replaced humans, even as the work moved from country to country as companies chased low cost labor as a way to compete with each other.

The rise of the machines has been relentless, rapid and broad.

With AI though, the replacement could be much broader. It could start to include so-called “knowledge workers.” People like meteorologists, lawyers, real estate agents, doctors, financial advisors, pharmacists, programmers etc. may find themselves increasing replaced with semi-intelligent learning machines. Beyond that more “complex” jobs such as drivers, pilots, and soldiers may find themselves replaced by more comprehensive and cheaper robotic alternatives. Indeed, the idea of replacing drivers was unthinkable even as little as 10 years ago… now it seems almost certain.

These are no longer physically demanding and grueling jobs… these are the so-called “good” jobs that humans are “built for.”

But if we replace all of these jobs with cheaper, more efficient alternatives, what should those people train to do? Is there something beyond “knowledge work” which humans are better at that is even more desirable and rewarding?

GOING EVEN BROADER

Ok. So maybe further down the line we become a species of creative expression and invention. No longer limited by physical needs or even mental needs we can essentially “do what we want.” Putting aside the economic problems that this may spark, let’s imagine what this may look like.

Perhaps people can focus on music or painting or sports or invention. Personally rewarding, creative and entertaining pursuits that still give personal meaning as well as social value.

But what if the machines can do that as well?

What if an AI can compose BETTER music, paint a BETTER painting and play a BETTER game of basketball? What then?

What if we develop an AI that can find the cure for cancer? Or figure out how to colonize Mars? Or how to travel faster than light?

What if in a few hundred or a few thousand years we just watch in awe as our creation grows far beyond our wildest dreams?

And why wouldn’t this happen?

The incentive for making the best is what drives this from it’s origin, isn’t it?

If machines are better at the physical, and then better at the cognitive, why can’t they also be better at the creative?

I think they can, and I think they will.

INVERTING THE PROBLEM

So I think it’s worth turning this around.

Why are we making these machines at all? Is it to make life “better.” Better for whom? And what is better exactly? Less toil? Less suffering? Less pain? Is it about making things easier and safer?

Is it for our continued survival as a species?

Ok. Let’s assume that’s true… once we’ve checked off “survival” as a goal, does it become about maximizing comfort and satisfaction?

But again, we must face the eternal why.

Let’s imagine an absurd world where an army of perfect robots can provide us with every need, solve every problem, advance technology faster than we ever could and, more importantly, improve itself FASTER than we can improve them. And let’s imagine that their sole purpose is to fulfill our human desires.

What does that world look like?

If the end result of our current trajectory is to lead a life of leisure and consumption, why should we exist at all? Beyond the pure personal experience, what is the difference between living a life of pure indulgent consumption and living no life at all?

Don’t get me wrong… people could and would still fulfill their desires. They could still paint and make music. They could still program and tinker with new technology. All that would be completely possible, but it would make as much difference to the AI robots as a group of ants collecting chips on the floor.

SO WHAT *IS* OUR PURPOSE IN THE UNIVERSE?

Well… of course I don’t know, and neither do you.

Perhaps there isn’t one. Perhaps “purpose” is the evolutionary requirement in our minds to keep the species going. It’s what causes us to behave in the ways we do, invent the things we invent in order to keep the genes passing from one generation to the next; building up to the next evolutionary leap.  Perhaps that illusion is so deeply implanted into us that we can’t escape it even if we are completely aware of it.

Perhaps we are just one dot in the chain of evolution and once we replace the slow, biological evolutionary advances with the more rapid and exponential technical ones… we will have served our purpose and, like the dinosaurs before us, we should be eliminated to make way for another species to have it’s chance.

And maybe that is purpose enough.

Stephen Hawking and Elon Musk made headlines this year by raising the specter of super advanced Artificially Intelligent machines taking over the world. The idea goes something like this:

1) Today humans dominate because we are creative and can make complex decisions.
2) Computers/Robots are “mindless” servant that do our bidding.
3) In the race to make things “more efficient” we will eventually make AI that can imagine and create.
4) When that AI does that it will realize it doesn’t need us anymore.
5) Bye bye, humans.

OK cool.

So my first reaction is… what’s the problem? I mean from a universal evolutionary standpoint, there’s not really any reason to believe that humans are special. If a meteor can wipe out the Dinosaurs (who dominated FAR longer than we have) in a few geologic seconds; why can’t we get wiped out by the results of our own making. Sure.. it is bad for HUMANS, but that doesn’t mean that it’s some kind of universal evil.

But let’s be more practical. Let’s assume that the Star Trek vision of the universe is “objectively good” or at least desirable for humanity. Does AI really threaten that and if so, how?

Between here and there

I think that the conversation has become far too black and white. On the one hand “AI will always be our loyal slaves” and on the other “eventually they will wipe us out.”

I’d like to talk a bit about the road in between starting with the very near future… say 50-150 years.

The last 150 years has been marked by “machines” and technology replacing base human labor. Robots build cars, stuff boxes, package things, cut down trees, break rocks, construct buildings and so on. As that storm has raged, humans have become more “efficient.” That is, it used to take many thousands of people many years to build a pyramid, but a few hundred people can build a skyscraper in a year or two. The side effect of this, of course, has been that humans have to compete with robots for labor “wages.”

So if you own an auto plant you have to decide if paying someone $x/hr over the course of 30 years is cheaper or more expensive than buying a robot for $x and maintaining it. That pushes car making (and everything else) to cheaper labor costs (robots or cheaper humans).  The reason why this is different than before is that the labor wages are not moving between people but from people to machines… and thus, the owner of the machines. The flip side is that the cost of those things also drops dramatically so as long as there is other work, that displaced worker can live cheaper. And thus over the past 150 years the cost of things has generally reduced while the “productivity” has gone up.

Ok. That is a very primitive summary and you could fire a bunch of holes into it, but it’s not really what I want to talk about… it’s more setting the stage, so bear with me. My main point is that machines in the past 150 years have primarily replaced human physical activity as opposed to human mental activity.

The Next 150 Years Looks Quite Different

I think that is going to change dramatically.

The next stage in the eternal quest for efficiency and productivity is to reduce the cost of mental activity. We can already see this happening with things like self-driving cars. Driving is a suboptimal human activity. It’s boring, frustrating, dangerous and doesn’t really create value. If we can get robots to do it for us, GREAT!. Of course, that means we take out the people who make a living driving, but that’s just the natural cost of the ongoing technology evolution, right? Those people can learn to do something more useful, more fun, more helpful, and more interesting. Maybe they can start a blog or something…

Other areas also seem pretty obvious… how about ordering fast food? Or how about checking out at the grocery store? Or what about trying to find a special tool at Home Depot? Imagine if there’s Wall-e the helpful robot tool guy and you ask him, and he cheerfully shows you exactly what you need. No more listening to someone drone on about some uninteresting thing, no more waiting while he finds the person who might know the answer, no more “hmm… I’m not sure if we have that, let me go find out.”

Restaurants could also benefit. Who needs a moody waiter or an overstressed kid on their first day screwing up your order. Androids would always be polite, accurate, and efficient. You can even just pay with your iPhone and not even leave a tip!

Of course all those people lose their jobs too… but of course, they can find more productive things to do. Or so we hope.

But Wait A Minute…

I think we’re actually going somewhere else… maybe a step “further” or maybe a different step.

Actually I think robo-waiters are not going to be so easy… because when I go out to eat I VALUE the human interaction and while you can AI that pretty effectively, it’s very hard to AI the warmth and care. If you have a favorite restaurant you may know the waiter’s family (through pictures), you feel happy seeing him, and if he screws up your order a bit, that’s ok because he probably also remembers something or can notice when you’re down and cheer you up. These “semi-intimate” service jobs have a very high value component that goes beyond getting what you want quickly.

Same with the Home Depot person. She may ask you about what you’re trying to do and empathize with your frustration of not knowing how a toilet actually flushes. She may take time to explain it, crack a joke and make you feel better. That’s a pretty hard thing to AI away.

But… I think there’s another kind of service industry that WILL benefit from AI… and it’s much scarier when it comes to “retraining people.”

Let’s think about areas where errors lead to real consequences. Like, say, surgery? Or maybe military tactics? Or police work? Or perhaps judges? What about deciding whether a drug should be approved or not? Those are areas where AI androids might do BETTER than people and as a consumer I don’t care as much about the human side. When I’m knocked out on the OR table and someone is going to cut me open… who do I want? The human surgeon who just got out of a law suit and found out that his wife is cheating on him… maybe he had a drink after a 12 hour shift and was called back last second? Or do I want the feelingless, tireless and perfectly accurate robot that has millions of historic cases of surgery it can recall instantly to make the best possible decision and execute each cut perfectly?

Hmm… I think I want the robot.

Surgery and Driving are more similar than Waitering and Driving.

If I run a hospital, this is good news. I am pretty sure that when androids start doing medical procedures, malpractice rates will go DOWN. So all we need is one or two human professionals to make sure that the machines are working well and we’re good. Of course we still need the hardcore research but that’s a small fraction of human labor. So all you people working hard to become surgeons and make big bucks… sorry! You probably want to learn to be a very good waiter instead.

What about other parts of the medical field? General diagnosis, taking blood, urine samples, dispensing medications, etc. I think Androids are better at that too. Once the have all the information and cases available I think they will be more accurate at diagnosing than humans. And I think they will be BETTER on their own than with human/machine interaction – contrary to what expert humans will say of course :).

Why? Because these are classic areas where humans will be overconfident, bias in their anchoring, not have all the information, etc. What AI can’t replace is a person who shows empathy and compassion… but that’s a very different kind of job…normally more part of “nursing.” The doctor is the jerk who knows everything and the nurse is the nice person who cares. I think the jerk who knows everything is much easier to AI away. I need a nice person who makes me feel good, but I want the AI doing the diagnosis and deciding the treatment.

Of course, people will have to get used to this. The first generation will rebel and want the “good old days” but assuming the outcomes are superior, the new technology will win out.

I dug into medicine, but I think anywhere you see jobs that require calculated decision where bad outcomes are actually dangerous, you will see AI take over. And that’s a LOT of people. AND they can scale very quickly. A single AI can probably diagnose hundreds of people at the same time, given that we have enough input devices (some kind of tricorder, or maybe an army of humans who just hook you up to some machines). In essence the humans in these areas will do what the AI says, not the other way around.

Think about your own job. Does this describe what you do? Well… maybe not for long.

Scary, huh?

Conclusion

While I think we are very far from AI “taking over humanity and wiping it out” I do think we will face the very real challenge of  high-risk outcome, judgment based jobs being replaced by AI. And those jobs ARE interesting and productive for people… it’s just… we’re not as good at them.

So we will face a different problem. Is it BETTER for people to be less efficient and have worse outcomes in exchange for being more professionally satisfied, or is the race to improved efficiency and better outcomes the more important goal?