The Sentient Machine

The Coming Age of Artificial Intelligence

Amir Husain

• Publisher: Scribner
• Year: 2017
• ISBN:
  9781501144691
• Topic: Filosofia
• Subtopic: Filosofia ed etica nella scienza

"A must-read for anyone looking to understand how artificial intelligence is poised to transform human society and life." —Paul Scharre, Author of Four Battlegrounds: Power in the Age of Artificial Intelligence The future is now. Acclaimed technologist and inventor Amir Husain explains how we can live amidst the coming age of sentient machines and artificial intelligence—and not only survive, but thrive. Artificial "machine" intelligence is playing an ever-greater role in our society. We are already using cruise control in our cars, automatic checkout at the drugstore, and are unable to live without our smartphones. The discussion around AI is polarized; people think either machines will solve all problems for everyone, or they will lead us down a dark, dystopian path into total human irrelevance. Regardless of what you believe, the idea that we might bring forth intelligent creation can be intrinsically frightening. But what if our greatest role as humans so far is that of creators?Amir Husain, a brilliant inventor and computer scientist, argues that we are on the cusp of writing our next, and greatest, creation myth. It is the dawn of a new form of intellectual diversity, one that we need to embrace in order to advance the state of the art in many critical fields, including security, resource management, finance, and energy. "In The Sentient Machine, Husain prepares us for a brighter future; not with hyperbole about right and wrong, but with serious arguments about risk and potential" (Dr. Greg Hyslop, Chief Technology Officer, The Boeing Company). He addresses broad existential questions surrounding the coming of AI: Why are we valuable? What can we create in this world? How are we intelligent? What constitutes progress for us? And how might we fail to progress? Husain boils down complex computer science and AI concepts into clear, plainspoken language and draws from a wide variety of cultural and historical references to illustrate his points. Ultimately, Husain challenges many of our societal norms and upends assumptions we hold about "the good life."

PART ONE

What Is AI?

The first time it happened, I was at the University of Texas in Austin as a student. It was one of my first weeks of school and I woke up in my bedroom in such a confusing paroxysm of pain that I was convinced I was dying. Nothing in my life leading up to that moment had prepared me for this pain: as if a knife were slicing through my right eye socket.

Somehow I stumbled out onto the street where I was able to find my way to the university health center. They handed me an Advil and directed me to lie down in a dark, quiet room until the pain subsided. But the pain did not subside. For two long days, I played mind games with myself in that dark room. Would you rather . . . Would you rather stick your hand in an open flame or feel this particular pain in the right side of your head? Would you rather have a bullet go through your body or continue to experience this throbbing in your eye? These were not abstract questions. They were very real negotiations as I navigated the incomprehensible level of suffering. Eventually, when the throbbing continued unabated, the game simplified: Would you rather be enduring this torture or be dead?

There is a reason my ultimate diagnosis—I was suffering from the first of many cluster headaches—is referred to colloquially as a “suicide headache.” Many headache doctors who have treated patients with chronic cluster headaches have had at least a few of their patients choose death over the pain.

After my first episode with them at the age of eighteen, I didn’t have them again until I was twenty. Both of those attacks were relatively short—lasting only a few days. Over time, however, the duration of the attacks increased. They went from days to months. I couldn’t find words or numbers to define this type of pain. Simply put, it was impossible to even remain conscious through it. Eventually, I almost always passed out.

Two years ago, after a remission of several years, I was revisited by my old demons. One night, I experienced an attack so extreme that I knew I was in trouble. I had gone way beyond my weekly dosage of the prescribed sumatriptan, so I reached for another painkiller in desperation. My vision got cloudy; the last thing I remember is telling my wife to call 911.

When I woke up, I was checked into the hospital and the neurologist on staff, Dr. Reddiah Mummaneni, was sitting with me. I could immediately sense that he was kind and bright and engaged. Dr. Mummaneni had done some research on my condition and he handed me a medical paper he had found using a query on his research repository.

“There is only one thing I’ve found that might help right now,” he told me. “It’s a really drastic treatment and only one or two incidents are described in this paper. We can try a really high dose of a steroid, methylprednisolone, administered through an IV. It’s a common treatment for multiple sclerosis because it acts like a shock to the immune system to calm it down and it helps with the inflammation.” He explained that he hoped the anti-inflammatory capabilities of the steroid—at such an extremely high level—could suppress the unexplained inflammation and give a jolt to the mechanisms in my brain that were causing the clusters. After “jolting” my brain for three days, we would move down to an oral steroid for a month and then taper off from there. Dr. Mummaneni said, “I want you to know that this treatment could have very serious repercussions. Are you okay with the risks?” I read the paper and signed off immediately. Considering my current quality of life, there was not a single doubt in my mind. The course of treatment was so severe, however, that the staff was repeatedly instructed to check my sugar, blood pressure, pulse, and heart rate. They followed up with an EKG before, at last, the IV drip was put in place. Suddenly, as quickly as it had started, the attack abated. After four months, the pain was gone. I was finally free. For the time being, at least.

• • •

I am a computer scientist, a technologist, and an inventor. I hold two dozen awarded patents and have dozens more pending. I love machines and I have a deep affinity for how they work and what they can do. Although the story I just shared certainly makes a case for compassionate medical care, more than anything, it illustrates to me our own limitations in the field of healthcare. When humans succeed at sorting through all the data involved in an obscure medical treatment, it is often the result of happenstance: I happened to get the right doctor assigned to me on that night in the hospital and that doctor happened to get the right combination of words for the query to pull up just the right paper. And then he happened to have the time to read it and process it. When I use the word happenstance, I don’t discredit all the wisdom, discernment, and training that go into being a highly skilled—and in this case, an incredibly effective—medical practitioner. Happenstance is simply a reflection of the probabilities in search queries that now exist for anyone in the medical field. Medline, one of the most popular online research repositories for doctors, reported that more than 6 million journal article references have accumulated since 1965 and their database is growing at a rate of 300,000 per year. Human intelligence is simply too limited to process this amount of information. If we happen to find the essential data point, we find it. If we happen to miss it, we miss it.

But there is a different kind of intelligence much better suited to solving these types of challenges; this is artificial intelligence. Today, AI is being applied to all kinds of health data, allowing us to assess previously invisible, seemingly immeasurable aspects of our lives. Why did my cluster headaches go into remission for two years and then one year and then five years? Why did they sometimes occur at the same time twice a day and sometimes only once a day? Why did they strike in the fall sometimes and at other times in the summer? What is really happening? Can artificial intelligence help me solve my own personal mysteries? Can I use my intimate knowledge of computer science and technology to end my suffering?

Cluster headaches are just one of countless ailments that cause unbearable pain but don’t warrant major research or attention because not enough people suffer from them. Our human intellect is a finite resource and it is currently—and rightly—directed at problems that have the biggest impact on humanity. But there are lots of diseases that will never receive the top-end human capital and expertise because they simply don’t affect enough people. For that reason, many people who experience the long tail of suffering will never receive the quality of intellect they need to get rid of their pain. Humans simply can’t keep up with all the diseases. But machines can. And they must—if we want to improve our lives.

This isn’t just happening in healthcare. This is happening in every domain in our society. This is happening in astronomy, chemistry, materials science, manufacturing, financial services. The list goes on and on. This is progress, and I believe it is imperative we embrace it.

WHAT ARE WE SO AFRAID OF?

I was speaking on artificial intelligence at the popular technology conference South by Southwest—SXSW—in my hometown of Austin. Soon after, protesters chanted “I say robot; you say no-bot,” and their signs read “Stop the Robots!” and “Humans Are the Future!” The year was 2015 but it might as well have been 1980, 1967, 1950, or even the late 1800s. Throughout all of these periods in history, technological advances have left us both optimistic and also deeply uneasy about the role of machines in our culture.

These ebbs and flows of our greater historical mood toward machine automation can be tracked to the technological advancements of the age. More than two centuries ago, the invention of the steam engine sent society into a tailspin as cultural commentators, prognosticators, and businesses tried to parse out the “machinery question.” Writers associated with the Young England movement aligned themselves with laborers in sentimentalizing the work of the hand as a patriotic toil. Machines like the steam engine and the power loom became emblematic as dehumanizing forces that threatened the national character.

Merrily went the click-clack, the hammer, and the plough

And honest men could live by the sweat of their brow.

In light of Brexit, this yearning for “merry old England” feels particularly resonant. The destabilizing forces of technological innovations inevitably bring about our nostalgic longings for a simpler time. It was in the midst of this very shift—the Industrial Revolution—that Mary Shelley published her iconic novel of horror in 1818, a work that explored the dehumanizing forces of an increasingly technological society through the lens of literature’s first modern scientist: Victor Frankenstein.

• • •

Mary Shelley was only eighteen when she wrote Frankenstein. In one of the coldest and dreariest summers ever on record—a huge volcano erupted in Indonesia in 1815, sending a vast cloud of ashes across the Northern Hemisphere the following year—Shelley was holed up in an elegant Swiss villa with a bohemian coterie that included Lord Byron, her stepsister Claire Clairmont, and her romantic partner, Percy Shelley. After countless nights of rain and thunder, Lord Byron suggested they should all write ghost stories. Shelley had only recently left the home of her father, the famous progressive William Godwin, where she was exposed to the work of the great scientists of the day, taking in a wide range of ideas in subjects ranging from anatomy to galvanism—the study of muscle contractions as a result of electrical currents. At Byron’s suggestion, Shelley’s influences and inspirations came together to create her characterization of Victor Frankenstein.

In twentieth-century films and stage plays, Frankenstein’s monster was rendered in more melodramatic fashion with pulsing electrodes and nodules on his head. In Shelley’s original version, however, scientist Victor Frankenstein uses elemental principles of life to imbue vitality into inanimate matter. She even makes a distinct point of stating that Frankenstein draws upon the scientific method for his work. The moment she describes her vision of the monster, however, is a passage that harkens back to writing in the ancient Greek myth of Prometheus or from Genesis in the Old Testament:

I saw the pale student of unhallowed arts kneeling beside the thing he had put together. I saw the hideous phantasm of a man stretched out, and then, on the working of some powerful engine, show signs of life and stir with an uneasy, half-vital motion. Frightful must it be, for supremely frightful would be the effect of any human endeavour to mock the stupendous mechanism of the Creator of the world.

In this passage, we can see the crux of a key philosophical quandary, the cause of our unease: Who can be a creator? And when does our creation truly acquire agency?

Frankenstein posed these questions to an increasingly modern society, but for as long as humanity has been creating, we have also been in conflict with our role as creators. Many of the world’s greatest religions have decrees and practices around aniconism, banning material representations of living creatures and the divine. There are countless passages in the Hebrew Bible against graven imagery, beginning with the Second Commandment:

Thou shalt have no other gods before me. Thou shalt not make unto thee any graven image, or any likeness of any thing that is in heaven above, or that is in the earth beneath, or that is in the water under the earth: Thou shalt not bow down thyself to them, nor serve them: for I the Lord thy God am a jealous God, visiting the iniquity of the fathers upon the children unto the third and fourth generation of them that hate me.

And yet Hellenic stories abound with tales of automated representational machinery. In ancient Greek myths, inventions by characters like Hephaestus, the blacksmith of Olympus, are celebrated as wondrous innovations. He even designed what is surely the world’s first notion of a fully automated waitstaff with twenty wheeled devices propelling themselves in and out of the halls of Olympus, serving the gods while they indulged in drink and wine.

Historian and writer Pamela McCorduck, who has spent much of her career documenting the history of artificial intelligence, characterizes these two perspectives toward our creation of lifelike machinery—wondrous and useful versus arrogant and dangerous—as the polarities that inform our conflict even to this day. Do we dare usurp that power from a god or the natural world? And if we do, are we prepared to face the unintended consequences?

The Genesis story in the Quran, for example, quite similar to that in the Old Testament, narrates that Adam was cast from clay when God breathed his spirit into the clay form. God then taught Adam the nature of all things. With knowledge and the power to think thus imparted, God deemed Adam the most superior among all creation. According to this particular creation story, God ordered the angels and the fire-being Iblis, or Lucifer, to bow before this new, autonomous creation. When the angels innocently asked why, God explained that it was because Adam could acquire knowledge. His thought was not static, and that allowed him to learn, grow, and change. And so, in this seminal narrative of creation, the first form of autonomous intelligence distinct from God—human intelligence—was unleashed upon the universe.

Today, of course, we deal not with myth and allegory but with the real-world implications of an explosion of artificial intelligence that is outpacing our ability to understand its consequences. Even Silicon Valley’s most powerful players seem divided about which side of the technological divide they are on. In 2014, Tesla and SpaceX founder Elon Musk spoke at Massachusetts Institute of Technology and called AI humanity’s “biggest existential threat.” His talk went on to accuse artificial intelligence of “summoning the demon.”

“You know all those stories where there’s the guy with the pentagram and the holy water and he’s like, yeah, he’s sure he can control the demon?” he asked the crowd before saying, “Doesn’t work out.”

Elon Musk is not alone in sounding out the alarms. He is joined by intellectual heavyweights like Stephen Hawking, Bill Gates, Oxford professor and existential philosopher Nick Bostrom, and Henry Kissinger. In 2014, Hawking told the BBC: “The development of full artificial intelligence could spell the end of the human race,” and in a 2015 interview session on the popular Silicon Valley website Reddit, Bill Gates added, “I am in the camp that is concerned about super intelligence. First the machines will do a lot of jobs for us and not be super-intelligent. That should be positive if we manage it well. A few decades after that, though, the intelligence is strong enough to be a concern. I agree with Elon Musk and some others on this and don’t understand why some people are not concerned.”

In late 2015, Musk announced his intent to provide $1 billion in funding to a new nonprofit dedicated to safe artificial intelligence research. The organization, called OpenAI, brings together talented practitioners in the field and proposes to make their designs and code publicly available. Soon after, Musk, Hawking, and a group of a thousand other tech titans and power players signed a letter calling for a ban on autonomous weapons. In the fall of 2015, Silicon Valley’s most prominent technology companies came together to create the Partnership on Artificial Intelligence, exploring all of the issues, dangers, and ethical concerns that surround this coming age of intelligent machinery.

In our journey together, we will explore how anti-AI movements could be a threat to developing much-needed technology to solve this century’s most complex problems. In Part Two, I will explain why trying to suppress such work, or subject it to draconian regulation, will be incredibly harmful to us as a civilization. Before we approach these coming challenges, however, we need to situate ourselves in the culture of fear that surrounds our notions of artificial intelligence. And we need to fully understand how algorithms excel at tasks that human brains find so daunting. What is machine intelligence exactly?

• • •

Too often, we frame our discussions of AI around its anthropocentric characteristics: How much does it resemble us? Can it “pass” as human? This kind of thinking can be traced back to the 1950s when the preeminent mathematician Alan Turing, memorialized in the recent film The Imitation Game, published his paper “Computing Machinery and Intelligence” and described what we now call the Turing Test. In Turing’s test, a human interrogator interacts with both a human and a machine using a text-only channel, such as a screen. The computer succeeds—according to Turing’s paper—when it can fool the human interrogator with its answers, making the interrogator believe that it is human. Versions of the Turing Test as a barometer of success still abound all around us. In voice-recognition software programs, it’s considered a major coup when humans speak into the phone, imagining that they are speaking to a friendly sales representative in Nebraska, not a computer algorithm.

This is a curious form of narcissism. Do we really imagine that human intelligence is the only kind of intelligence worth imitating? Is mimicry really the ultimate goal? Machines have much to teach us about “thoughts” that have nothing to do with human thought.

The Turing Test and other such “mimicry” metrics for machine le...