Duality and computation in the MCU

I  took my kindergartener to see Avengers: Endgame recently. My son was a little disappointed, complaining that the film had too much talking and not enough fighting. To me, the immense popularity of the Marvel Cinematic Universe series and so-called science fiction/fantasy in general is an indicator of how people think they like science but really want magic. Popular science-fictiony franchises like MCU and Star Wars are couched in scientism but are often at odds with actual science as practiced today. Arthur C Clarke famously stated in his third law that “Any sufficiently advanced technology is indistinguishable from magic.” A sentiment captured in these films.

Science fiction should extrapolate from current scientific knowledge to the possible. Otherwise, it should just be called fiction. There have been a handful of films that try to do this like 2001: A Space Odyssey or more recently Interstellar and The Martian. I think there is a market for these types of films but they are certainly not as popular as the fantasy films. To be fair, neither Marvel nor Star Wars (both now owned by Disney) market themselves as science fiction as I defined it. They are intended to be mythologies a la Joseph Campbell’s Hero’s Journey. However, they do have a scientific aesthetic with worlds dominated by advanced technology.

Although I find the MCU films not overly compelling, they do bring up two interesting propositions. The first is dualism. The superhero character Ant-Man has a suit that allows him to change size and even shrink to sub-atomic scales, called the quantum realm in the films. (I won’t bother to discuss whether energy is conserved in these near instantaneous size changes, an issue that affects the Hulk as well). The film was advised by physicist Spiros Michalakis and is rife with physics terminology and concepts like quantum entanglement. One crucial concept it completely glosses over is how Ant-man maintains his identity as a person, much less his shape, when he is smaller than an atom. Even if one were to argue that one’s consciousness could be transferred to some set of quantum states at the sub-atomic scale, it would be overwhelmed by quantum fluctuations. The only self-consistent premise of Ant-Man is that the essence or soul if you wish of a person is not material. The MCU takes a definite stand for dualism on the mind-body problem, a sentiment with which I presume the public mostly agrees. 

The second is that magic has immense computational power. In the penultimate Avengers movie, the villain Thanos snaps his fingers while in possession of the complete set of infinity stones and eliminates half of all living things. (Setting aside the issue that Thanos clearly does not understand the the concept of exponential growth. If you are concerned about overpopulation, it is pointless to shrink the population and do nothing else because it will just return to its original size in short time.) What I’d like to know is who or what does the computation to carry out the command. There are at least two hard computational problems that must be solved. The first is to identify all lifeforms.  This is clearly no easy task as we to this day have no precise definition of life. Do viruses get culled by the snap? Do the population of silicon-based lifeforms of Star Trek get halved or is it only biochemical life? What algorithm does the snap use to find all the life forms? Living things on earth range in size from single cells (or viruses if you count them) all the way to 35 metre behemoths, which are comprised of over 10^{23} numbers of atoms. How do the stones know what scales they span in the MCU? Do photosynthetic lifeforms get spared since they don’t use many resources? What about fungi? Is the MCU actually a simulated universe where there is a continually updated census of all life? How accurate is the algorithm? Was it perfect? Did it aim for high specificity (i.e. reduce false positives so you only kill lifeforms and not non lifeforms) or high sensitivity (i.e. reduce false negatives and thus don’t miss any lifeforms). I think it probably favours sensitivity over specificity – who cares if a bunch of ammonia molecules accidentally get killed. The find-all-life problem is made much easier by proposition 1 because if all life were material then the only way to detect them would be to look for multiscale correlations between atoms (or find organic molecules if you only care about biochemical life). If each lifeform has a soul then you can simply search for “soulfulness”. The lifeforms were not erased instantly but only after a brief delay. What was happening over this delay. Is magic propagation limited by the speed of light or some other constraint? Or did the computation take time? In Endgame, the Hulk restores all the Thanos erased lifeforms and Tony Stark then snaps away Thanos and all of his allies. Where were the lifeforms after they were erased? In Heaven? In a soul repository somewhere? Is this one of the Nine Realms of the MCU? How do the stones know who is a Thanos ally? The second computation is to then decide which half to extinguish. The movie seems to imply that the choice was random so where did the randomness come from? Do the infinity stones generate random numbers? Do they rely on quantum fluctuations? Finally, in a world with magic, why is there also science? Why does the universe follow the laws of physics sometimes and magic other times. Is magic a finite resource as in Larry Niven’s The Magic Goes Away. So many questions, so few answers.

Catch-22 of our era

The screen on my wife’s iPhone was shattered this week and she had not backed up the photos. The phone seems to still be functioning otherwise so we plugged it into the computer to try to back it up but it requires us to unlock the phone and we can’t enter in the password. My wife refused to pay the 99 cents or whatever Apple charges to increase the disk space for iCloud to automatically back up the phone, so I suggested we just pay the ransom money and then the phone will back up automatically. I currently pay both Apple and Dropbox extortion money. However, since she hadn’t logged onto iCloud in maybe ever, it sent a code to her phone under the two-factor authentication scheme to type in to the website, but of course we can’t see it on her broken screen so that idea is done. We called Apple and they said you could try to change the number on her iCloud account to my phone but that was two days ago and they haven’t complied. So my wife gave up and tried to order a new phone. Under the new system of her university, which provides her phone, she can get a phone if she logs onto this site to request it. The site requires VPN and in order to get VPN she needs to, you guessed it, type in a code sent to her phone. So you need a functioning phone to order a new phone. Basically, tech products are not very good. Software still kind of sucks and is not really improving. My Apple Mac is much worse now than it was 10 years ago. I still have trouble projecting stuff on a screen. I will never get into a self driving car made by any tech company. I’ll wait for Toyota to make one; my (Japanese) car always works (my Audi was terrible).

Missing the trend

I have been fortunate to have been born at a time when I had the opportunity to witness the birth of several of the major innovations that shape our world today.  I have also managed to miss out on capitalizing on every single one of them. You might make a lot of money betting against what I think.

I was a postdoctoral fellow in Boulder, Colorado in 1993 when my very tech savvy advisor John Cary introduced me and his research group to the first web browser Mosaic shortly after it was released. The web was the wild west in those days with just a smattering of primitive personal sites authored by early adopters. The business world had not discovered the internet yet. It was an unexplored world and people were still figuring out how to utilize it. I started to make a list of useful sites but unlike Jerry Yang and David Filo, who immediately thought of doing the same thing and forming a company, it did not remotely occur to me that this activity could be monetized. Even though I struggled to find a job in 1994, was fairly adept at programming, watched the rise of Yahoo! and the rest of the internet startups, and had friends at Stanford and Silicon Valley, it still did not occur to me that perhaps I could join in too.

Just months before impending unemployment, I managed to talk my way into being the first post doc of Jim Collins, who just started as a non-tenure track research assistant professor at Boston University.  Midway through my time with Jim, we had a meeting with Charles Cantor, who was a professor at BU then, about creating engineered organisms that could eat oil. Jim subsequently recruited graduate student Tim Gardner, now CEO of Riffyn, to work on this idea. I thought we should create a genetic Hopfield network and I showed Tim how to use XPP to simulate the various models we came up with. However, my idea seemed too complicated to implement biologically so when I went to Switzerland to visit Wulfram Gerstner at the end of 1997,  Tim and Jim, freed from my meddling influence, were able create the genetic toggle switch and the field of synthetic biology was born.

I first learned about Bitcoin in 2009 and had even thought about mining some. However, I then heard an interview with one of the early developers, Gavin Andresen, and he failed to understand that because the supply of Bitcoins is finite, prices denominated in it would necessarily deflate over time. I was flabbergasted that he didn’t comprehend the basics of economics and was convinced that Bitcoin would eventually fail. Still, I could have mined thousands of Bitcoins on a laptop back then, which would be worth tens of millions today.  I do think blockchains are an important innovation and my former post-bac fellow Wally Xie is even the CEO of the blockchain startup QChain. Although I do not know where cryptocurrencies and blockchains will be in a decade, I do know that I most likely won’t have a role.

I was in Pittsburgh during the late nineties/early 2000’s in one of the few places where neural networks/deep learning, still called connectionism, was king. Geoff Hinton had already left Carnegie Mellon for London by the time I arrived at Pitt but he was still revered in Pittsburgh and I met him in London when I visited UCL. I actually thought the field had great promise and even tried to lobby our math department to hire someone in machine learning for which I was summarily dismissed and mocked. I recruited Michael Buice to work on the path integral formulation for neural networks because I wanted to write down a neural network model that carried both rate and correlation information so I could implement a correlation based learning rule. Michael even proposed that we work on an algorithm to play Go but obviously I demurred. Although, I missed out on this current wave of AI hype, and probably wouldn’t have made an impact anyway, this is the one area where I may get a second chance in the future.

 

 

The end of (video) reality

I highly recommend listening to this Radiolab podcast. It tells of new software that can create completely fabricated audio and video clips. This will take fake news to an entirely different level. It also means that citizen journalists with smartphones, police body cams, security cameras, etc. will all become obsolete. No recording can be trusted. On the other hand, we had no recording technology of any kind for almost all of human history so we will have to go back to simply trusting (or not trusting) what people say.

The robot human equilibrium

There has been some push back in the media against the notion that we will “soon” be replaced by robots, e.g. see here. But absence of evidence is not evidence of absence. Just because there seem to be very few machine induced job losses today doesn’t mean it won’t happen tomorrow or in ten years. In fact, when it does happen it probably will happen suddenly as have many recent technological changes. The obvious examples are the internet and smartphones but there are many others. We forget that the transition from vinyl records to CDs was extremely fast; then iPods and YouTube killed CDs. Video rentals became ubiquitous from nothing in just a few years and died just as fast when Netflix came along, which was then completely replaced a few years later by streaming video. It took Amazon a little longer to become dominant but the retail model that had existed for centuries has been completely upended in a decade. The same could happen with AI and robots. Unless you believe that human thought is not computable, then in principle there is nothing a human can do that a machine can’t. It could take time to set up the necessary social institutions and infrastructure for an AI takeover but once it is established the transition could be abrupt.

Even so that doesn’t mean all or even most humans will be replaced. The irony of AI, known as Moravec’s Paradox (e.g. here), is that things that are hard for humans to do, like play chess or read X-rays, are easy for machines to do and vice versa. Although drivers and warehouse workers are destined to be the first to be replaced, the next set of jobs will likely be highly paid professionals like stock brokers, accountants, doctors, and lawyers. But as the ranks of the employed start to shrink, the economy will also shrink and wages will go down (even if the displaced do eventually move on to other jobs it will take time). At some point, particularly for jobs that are easy for humans but harder for machines, humans could be cheaper than machines.  So while we can train a machine to be a house cleaner, it may be more cost effective to simply hire a person to change sheets and dust shelves. The premium on a university education will drop. The ability to sit still for long periods of time and acquire arcane specialized knowledge will simply not be that useful anymore. Centers for higher learning will become retreats for the small set of scholarly minded people who simply enjoy it.

As the economy shrinks, land prices in some areas should drop too and thus people could still eke out a living. Some or perhaps many people will opt or be pushed out of the mainstream economy altogether and retire to quasi-pre-industrial lives. I wrote about this in quasi-utopian terms in my AlphaGo post but a dystopian version is equally probable. In the dystopia, the gap between the rich and poor could make today look like an egalitarian paradise. However, unlike the usual dystopian nightmare like the Hunger Games where the rich exploit the poor, the rich will simply ignore the poor. But it is not clear what the elite will do with all that wealth. Will they wall themselves off from the rest of society and then what, engage in endless genetic enhancements or immerse themselves in a virtual reality world? I think I’d rather raise pigs and make candles out of lard.

 

 

 

 

Talk at SIAM Annual Meeting 2017

I gave an invited plenary talk at the 2017 Society of Applied and Industrial Mathematics Annual Meeting in Pittsburgh yesterday. My slides are here. I talked about some very new work on chaos and learning in spiking neural networks. My fellow Chris Kim and I were producing graphs up to a half hour before my talk! I’m quite excited about this work and I hope to get it published soon.

During my talk, I made an offhand threat that my current Mac would be the last one I buy. I made the joke because it was the first time that I could not connect to a projector with my laptop since I started using Mac almost 20 years ago. I switched to Mac from Linux back then because it was a Unix environment where I didn’t need to be a systems administrator to print and project. However, Linux has made major headway in the past two decades while Mac is backsliding. So, I’m seriously thinking of following through. I’ve been slowly getting disenchanted with Apple products over the past three years but I am especially disappointed with my new MacBook Pro. I have the one with the silly touch screen bar. The first thing the peeves me is that the activate Siri key is right next to the delete key so I accidentally hit and then have to reject Siri every five minutes. What mostly ties me to Mac right now is the Keynote presentation software, which I like(d) because it is easy to embed formulas and PDF files into. It is much harder to do the same in PowerPoint and I haven’t found an open source version that is as easy to use. However, Keynote keeps hanging on my new machine. I also get this situation where my embedded equations will randomly disappear and then reappear. Luckily I did a quick run through just before my talk and noticed that the vanished equations reappeared and I could delete them. Thus, the Keynote appeal has definitely diminished. Now, if someone would like to start an open source Keynote project with me… Finally, the new Mac does not seem any faster than my old Mac (it still takes forever to boot up) and Bard Ermentrout told me that his dynamical systems software tool XPP runs five times slower. So, any suggestions for a new machine?

AlphaGo and the Future of Work

In March of this year, Google DeepMind’s computer program AlphaGo defeated world Go champion Lee Sedol. This was hailed as a great triumph of artificial intelligence and signaled to many the beginning of the new age when machines take over. I believe this is true but the real lesson of AlphaGo’s win is not how great machine learning algorithms are but how suboptimal human Go players are. Experts believed that machines would not be able to defeat humans at Go for a long time because the number of possible games is astronomically large, \sim 250^{150} moves, in contrast to chess with a paltry \sim 35^{80} moves. Additionally, unlike chess, it is not clear what is a good position and who is winning during intermediate stages of a game. Thus, any direct enumeration and evaluation of possible next moves as chess computers do, like IBM’s Deep Blue that defeated Gary Kasparov, seemed to be impossible. It was thought that humans had some sort of inimitable intuition to play Go that machines were decades away from emulating. It turns out that this was wrong. It took remarkably little training for AlphaGo to defeat a human. All the algorithms used were fairly standard – supervised and reinforcement backpropagation learning in multi-layer neural networks1. DeepMind just put them together in a clever way and had the (in retrospect appropriate) audacity to try.

The take home message of AlphaGo’s success is that humans are very, very far away from being optimal at playing Go. Uncharitably, we simply stink at Go. However, this probably also means that we stink at almost everything we do. Machines are going to take over our jobs not because they are sublimely awesome but because we are stupendously inept. It is like the old joke about two hikers encountering a bear and one starts to put on running shoes. The other hiker says: “Why are you doing that? You can’t outrun a bear.” to which she replies, “I only need to outrun you!” In fact, the more difficult a job seems to be for humans to perform, the easier it will be for a machine to do better. This was noticed a long time ago in AI research and called Moravec’s Paradox. Tasks that require a lot of high level abstract thinking like chess or predicting what movie you will like are easy for computers to do while seemingly trivial tasks that a child can do like folding laundry or getting a cookie out of a jar on an unreachable shelf is really hard. Thus high paying professions in medicine, accounting, finance, and law could be replaced by machines sooner than lower paying ones in lawn care and house cleaning.

There are those who are not worried about a future of mass unemployment because they believe people will just shift to other professions. They point out that a century ago a majority of Americans worked in agriculture and now the sector comprises of less than 2 percent of the population. The jobs that were lost to technology were replaced by ones that didn’t exist before. I think this might be true but in the future not everyone will be a software engineer or a media star or a CEO of her own company of robot employees. The increase in productivity provided by machines ensures this. When the marginal cost of production goes to zero (i.e. cost to make one more item), as it is for software or recorded media now, the whole supply-demand curve is upended. There is infinite supply for any amount of demand so the only way to make money is to increase demand.

The rate-limiting step for demand is the attention span of humans. In a single day, a person can at most attend to a few hundred independent tasks such as thinking, reading, writing, walking, cooking, eating, driving, exercising, or consuming entertainment. I can stream any movie I want now and I only watch at most twenty a year, and almost all of them on long haul flights. My 3 year old can watch the same Wild Kratts episode (great children’s show about animals) ten times in a row without getting bored. Even though everyone could be a video or music star on YouTube, superstars such as Beyoncé and Adele are viewed much more than anyone else. Even with infinite choice, we tend to do what our peers do. Thus, for a population of ten billion people, I doubt there can be more than a few million that can make a decent living as a media star with our current economic model. The same goes for writers. This will also generalize to manufactured goods. Toasters and coffee makers essentially cost nothing compared to three decades ago, and I will only buy one every few years if that. Robots will only make things cheaper and I doubt there will be a billion brands of TV’s or toasters. Most likely, a few companies will dominate the market as they do now. Even, if we could optimistically assume that a tenth of the population could be engaged in producing goods and services necessary for keeping the world functioning that still leaves the rest with little to do.

Even much of what scientists do could eventually be replaced by machines. Biology labs could consist of a principle investigator and robot technicians. Although it seems like science is endless, the amount of new science required for sustaining the modern world could diminish. We could eventually have an understanding of biology sufficient to treat most diseases and injuries and develop truly sustainable energy technologies. In this case, machines could be tasked to keep the modern world up and running with little need of input from us. Science would mostly be devoted to abstract and esoteric concerns.

Thus, I believe the future for humankind is in low productivity occupations – basically a return to pre-industrial endeavors like small plot farming, blacksmithing, carpentry, painting, dancing, and pottery making, with an economic system in place to adequately live off of this labor. Machines can provide us with the necessities of life while we engage in a simulated 18th century world but without the poverty, diseases, and mass famines that made life so harsh back then. We can make candles or bread and sell them to our neighbors for a living wage. We can walk or get in self-driving cars to see live performances of music, drama and dance by local artists. There will be philosophers and poets with their small followings as they have now. However, even when machines can do everything humans can do, there will still be a capacity to sustain as many mathematicians as there are people because mathematics is infinite. As long as P is not NP, theorem proving can never be automated and there will always be unsolved math problems.  That is not to say that machines won’t be able to do mathematics. They will. It’s just that they won’t ever be able to do all of it. Thus, the future of work could also be mathematics.

  1. Silver, D. et al. Mastering the game of Go with deep neural networks and tree search. Nature 529, 484–489 (2016).