Tag Archives: talking

#434837 In Defense of Black Box AI

Deep learning is powering some amazing new capabilities, but we find it hard to scrutinize the workings of these algorithms. Lack of interpretability in AI is a common concern and many are trying to fix it, but is it really always necessary to know what’s going on inside these “black boxes”?

In a recent perspective piece for Science, Elizabeth Holm, a professor of materials science and engineering at Carnegie Mellon University, argued in defense of the black box algorithm. I caught up with her last week to find out more.

Edd Gent: What’s your experience with black box algorithms?

Elizabeth Holm: I got a dual PhD in materials science and engineering and scientific computing. I came to academia about six years ago and part of what I wanted to do in making this career change was to refresh and revitalize my computer science side.

I realized that computer science had changed completely. It used to be about algorithms and making codes run fast, but now it’s about data and artificial intelligence. There are the interpretable methods like random forest algorithms, where we can tell how the machine is making its decisions. And then there are the black box methods, like convolutional neural networks.

Once in a while we can find some information about their inner workings, but most of the time we have to accept their answers and kind of probe around the edges to figure out the space in which we can use them and how reliable and accurate they are.

EG: What made you feel like you had to mount a defense of these black box algorithms?

EH: When I started talking with my colleagues, I found that the black box nature of many of these algorithms was a real problem for them. I could understand that because we’re scientists, we always want to know why and how.

It got me thinking as a bit of a contrarian, “Are black boxes all bad? Must we reject them?” Surely not, because human thought processes are fairly black box. We often rely on human thought processes that the thinker can’t necessarily explain.

It’s looking like we’re going to be stuck with these methods for a while, because they’re really helpful. They do amazing things. And so there’s a very pragmatic realization that these are the best methods we’ve got to do some really important problems, and we’re not right now seeing alternatives that are interpretable. We’re going to have to use them, so we better figure out how.

EG: In what situations do you think we should be using black box algorithms?

EH: I came up with three rules. The simplest rule is: when the cost of a bad decision is small and the value of a good decision is high, it’s worth it. The example I gave in the paper is targeted advertising. If you send an ad no one wants it doesn’t cost a lot. If you’re the receiver it doesn’t cost a lot to get rid of it.

There are cases where the cost is high, and that’s then we choose the black box if it’s the best option to do the job. Things get a little trickier here because we have to ask “what are the costs of bad decisions, and do we really have them fully characterized?” We also have to be very careful knowing that our systems may have biases, they may have limitations in where you can apply them, they may be breakable.

But at the same time, there are certainly domains where we’re going to test these systems so extensively that we know their performance in virtually every situation. And if their performance is better than the other methods, we need to do it. Self driving vehicles are a significant example—it’s almost certain they’re going to have to use black box methods, and that they’re going to end up being better drivers than humans.

The third rule is the more fun one for me as a scientist, and that’s the case where the black box really enlightens us as to a new way to look at something. We have trained a black box to recognize the fracture energy of breaking a piece of metal from a picture of the broken surface. It did a really good job, and humans can’t do this and we don’t know why.

What the computer seems to be seeing is noise. There’s a signal in that noise, and finding it is very difficult, but if we do we may find something significant to the fracture process, and that would be an awesome scientific discovery.

EG: Do you think there’s been too much emphasis on interpretability?

EH: I think the interpretability problem is a fundamental, fascinating computer science grand challenge and there are significant issues where we need to have an interpretable model. But how I would frame it is not that there’s too much emphasis on interpretability, but rather that there’s too much dismissiveness of uninterpretable models.

I think that some of the current social and political issues surrounding some very bad black box outcomes have convinced people that all machine learning and AI should be interpretable because that will somehow solve those problems.

Asking humans to explain their rationale has not eliminated bias, or stereotyping, or bad decision-making in humans. Relying too much on interpreted ability perhaps puts the responsibility in the wrong place for getting better results. I can make a better black box without knowing exactly in what way the first one was bad.

EG: Looking further into the future, do you think there will be situations where humans will have to rely on black box algorithms to solve problems we can’t get our heads around?

EH: I do think so, and it’s not as much of a stretch as we think it is. For example, humans don’t design the circuit map of computer chips anymore. We haven’t for years. It’s not a black box algorithm that designs those circuit boards, but we’ve long since given up trying to understand a particular computer chip’s design.

With the billions of circuits in every computer chip, the human mind can’t encompass it, either in scope or just the pure time that it would take to trace every circuit. There are going to be cases where we want a system so complex that only the patience that computers have and their ability to work in very high-dimensional spaces is going to be able to do it.

So we can continue to argue about interpretability, but we need to acknowledge that we’re going to need to use black boxes. And this is our opportunity to do our due diligence to understand how to use them responsibly, ethically, and with benefits rather than harm. And that’s going to be a social conversation as well as as a scientific one.

*Responses have been edited for length and style

Image Credit: Chingraph / Shutterstock.com Continue reading

Posted in Human Robots

#434759 To Be Ethical, AI Must Become ...

As over-hyped as artificial intelligence is—everyone’s talking about it, few fully understand it, it might leave us all unemployed but also solve all the world’s problems—its list of accomplishments is growing. AI can now write realistic-sounding text, give a debating champ a run for his money, diagnose illnesses, and generate fake human faces—among much more.

After training these systems on massive datasets, their creators essentially just let them do their thing to arrive at certain conclusions or outcomes. The problem is that more often than not, even the creators don’t know exactly why they’ve arrived at those conclusions or outcomes. There’s no easy way to trace a machine learning system’s rationale, so to speak. The further we let AI go down this opaque path, the more likely we are to end up somewhere we don’t want to be—and may not be able to come back from.

In a panel at the South by Southwest interactive festival last week titled “Ethics and AI: How to plan for the unpredictable,” experts in the field shared their thoughts on building more transparent, explainable, and accountable AI systems.

Not New, but Different
Ryan Welsh, founder and director of explainable AI startup Kyndi, pointed out that having knowledge-based systems perform advanced tasks isn’t new; he cited logistical, scheduling, and tax software as examples. What’s new is the learning component, our inability to trace how that learning occurs, and the ethical implications that could result.

“Now we have these systems that are learning from data, and we’re trying to understand why they’re arriving at certain outcomes,” Welsh said. “We’ve never actually had this broad society discussion about ethics in those scenarios.”

Rather than continuing to build AIs with opaque inner workings, engineers must start focusing on explainability, which Welsh broke down into three subcategories. Transparency and interpretability come first, and refer to being able to find the units of high influence in a machine learning network, as well as the weights of those units and how they map to specific data and outputs.

Then there’s provenance: knowing where something comes from. In an ideal scenario, for example, Open AI’s new text generator would be able to generate citations in its text that reference academic (and human-created) papers or studies.

Explainability itself is the highest and final bar and refers to a system’s ability to explain itself in natural language to the average user by being able to say, “I generated this output because x, y, z.”

“Humans are unique in our ability and our desire to ask why,” said Josh Marcuse, executive director of the Defense Innovation Board, which advises Department of Defense senior leaders on innovation. “The reason we want explanations from people is so we can understand their belief system and see if we agree with it and want to continue to work with them.”

Similarly, we need to have the ability to interrogate AIs.

Two Types of Thinking
Welsh explained that one big barrier standing in the way of explainability is the tension between the deep learning community and the symbolic AI community, which see themselves as two different paradigms and historically haven’t collaborated much.

Symbolic or classical AI focuses on concepts and rules, while deep learning is centered around perceptions. In human thought this is the difference between, for example, deciding to pass a soccer ball to a teammate who is open (you make the decision because conceptually you know that only open players can receive passes), and registering that the ball is at your feet when someone else passes it to you (you’re taking in information without making a decision about it).

“Symbolic AI has abstractions and representation based on logic that’s more humanly comprehensible,” Welsh said. To truly mimic human thinking, AI needs to be able to both perceive information and conceptualize it. An example of perception (deep learning) in an AI is recognizing numbers within an image, while conceptualization (symbolic learning) would give those numbers a hierarchical order and extract rules from the hierachy (4 is greater than 3, and 5 is greater than 4, therefore 5 is also greater than 3).

Explainability comes in when the system can say, “I saw a, b, and c, and based on that decided x, y, or z.” DeepMind and others have recently published papers emphasizing the need to fuse the two paradigms together.

Implications Across Industries
One of the most prominent fields where AI ethics will come into play, and where the transparency and accountability of AI systems will be crucial, is defense. Marcuse said, “We’re accountable beings, and we’re responsible for the choices we make. Bringing in tech or AI to a battlefield doesn’t strip away that meaning and accountability.”

In fact, he added, rather than worrying about how AI might degrade human values, people should be asking how the tech could be used to help us make better moral choices.

It’s also important not to conflate AI with autonomy—a worst-case scenario that springs to mind is an intelligent destructive machine on a rampage. But in fact, Marcuse said, in the defense space, “We have autonomous systems today that don’t rely on AI, and most of the AI systems we’re contemplating won’t be autonomous.”

The US Department of Defense released its 2018 artificial intelligence strategy last month. It includes developing a robust and transparent set of principles for defense AI, investing in research and development for AI that’s reliable and secure, continuing to fund research in explainability, advocating for a global set of military AI guidelines, and finding ways to use AI to reduce the risk of civilian casualties and other collateral damage.

Though these were designed with defense-specific aims in mind, Marcuse said, their implications extend across industries. “The defense community thinks of their problems as being unique, that no one deals with the stakes and complexity we deal with. That’s just wrong,” he said. Making high-stakes decisions with technology is widespread; safety-critical systems are key to aviation, medicine, and self-driving cars, to name a few.

Marcuse believes the Department of Defense can invest in AI safety in a way that has far-reaching benefits. “We all depend on technology to keep us alive and safe, and no one wants machines to harm us,” he said.

A Creation Superior to Its Creator
That said, we’ve come to expect technology to meet our needs in just the way we want, all the time—servers must never be down, GPS had better not take us on a longer route, Google must always produce the answer we’re looking for.

With AI, though, our expectations of perfection may be less reasonable.

“Right now we’re holding machines to superhuman standards,” Marcuse said. “We expect them to be perfect and infallible.” Take self-driving cars. They’re conceived of, built by, and programmed by people, and people as a whole generally aren’t great drivers—just look at traffic accident death rates to confirm that. But the few times self-driving cars have had fatal accidents, there’s been an ensuing uproar and backlash against the industry, as well as talk of implementing more restrictive regulations.

This can be extrapolated to ethics more generally. We as humans have the ability to explain our decisions, but many of us aren’t very good at doing so. As Marcuse put it, “People are emotional, they confabulate, they lie, they’re full of unconscious motivations. They don’t pass the explainability test.”

Why, then, should explainability be the standard for AI?

Even if humans aren’t good at explaining our choices, at least we can try, and we can answer questions that probe at our decision-making process. A deep learning system can’t do this yet, so working towards being able to identify which input data the systems are triggering on to make decisions—even if the decisions and the process aren’t perfect—is the direction we need to head.

Image Credit: a-image / Shutterstock.com Continue reading

Posted in Human Robots

#434336 These Smart Seafaring Robots Have a ...

Drones. Self-driving cars. Flying robo taxis. If the headlines of the last few years are to be believed, terrestrial transportation in the future will someday be filled with robotic conveyances and contraptions that will require little input from a human other than to download an app.

But what about the other 70 percent of the planet’s surface—the part that’s made up of water?

Sure, there are underwater drones that can capture 4K video for the next BBC documentary. Remotely operated vehicles (ROVs) are capable of diving down thousands of meters to investigate ocean vents or repair industrial infrastructure.

Yet most of the robots on or below the water today still lean heavily on the human element to operate. That’s not surprising given the unstructured environment of the seas and the poor communication capabilities for anything moving below the waves. Autonomous underwater vehicles (AUVs) are probably the closest thing today to smart cars in the ocean, but they generally follow pre-programmed instructions.

A new generation of seafaring robots—leveraging artificial intelligence, machine vision, and advanced sensors, among other technologies—are beginning to plunge into the ocean depths. Here are some of the latest and most exciting ones.

The Transformer of the Sea
Nic Radford, chief technology officer of Houston Mechatronics Inc. (HMI), is hesitant about throwing around the word “autonomy” when talking about his startup’s star creation, Aquanaut. He prefers the term “shared control.”

Whatever you want to call it, Aquanaut seems like something out of the script of a Transformers movie. The underwater robot begins each mission in a submarine-like shape, capable of autonomously traveling up to 200 kilometers on battery power, depending on the assignment.

When Aquanaut reaches its destination—oil and gas is the primary industry HMI hopes to disrupt to start—its four specially-designed and built linear actuators go to work. Aquanaut then unfolds into a robot with a head, upper torso, and two manipulator arms, all while maintaining proper buoyancy to get its job done.

The lightbulb moment of how to engineer this transformation from submarine to robot came one day while Aquanaut’s engineers were watching the office’s stand-up desks bob up and down. The answer to the engineering challenge of the hull suddenly seemed obvious.

“We’re just gonna build a big, gigantic, underwater stand-up desk,” Radford told Singularity Hub.

Hardware wasn’t the only problem the team, comprised of veteran NASA roboticists like Radford, had to solve. In order to ditch the expensive support vessels and large teams of humans required to operate traditional ROVs, Aquanaut would have to be able to sense its environment in great detail and relay that information back to headquarters using an underwater acoustics communications system that harkens back to the days of dial-up internet connections.

To tackle that problem of low bandwidth, HMI equipped Aquanaut with a machine vision system comprised of acoustic, optical, and laser-based sensors. All of that dense data is compressed using in-house designed technology and transmitted to a single human operator who controls Aquanaut with a few clicks of a mouse. In other words, no joystick required.

“I don’t know of anyone trying to do this level of autonomy as it relates to interacting with the environment,” Radford said.

HMI got $20 million earlier this year in Series B funding co-led by Transocean, one of the world’s largest offshore drilling contractors. That should be enough money to finish the Aquanaut prototype, which Radford said is about 99.8 percent complete. Some “high-profile” demonstrations are planned for early next year, with commercial deployments as early as 2020.

“What just gives us an incredible advantage here is that we have been born and bred on doing robotic systems for remote locations,” Radford noted. “This is my life, and I’ve bet the farm on it, and it takes this kind of fortitude and passion to see these things through, because these are not easy problems to solve.”

On Cruise Control
Meanwhile, a Boston-based startup is trying to solve the problem of making ships at sea autonomous. Sea Machines is backed by about $12.5 million in capital venture funding, with Toyota AI joining the list of investors in a $10 million Series A earlier this month.

Sea Machines is looking to the self-driving industry for inspiration, developing what it calls “vessel intelligence” systems that can be retrofitted on existing commercial vessels or installed on newly-built working ships.

For instance, the startup announced a deal earlier this year with Maersk, the world’s largest container shipping company, to deploy a system of artificial intelligence, computer vision, and LiDAR on the Danish company’s new ice-class container ship. The technology works similar to advanced driver-assistance systems found in automobiles to avoid hazards. The proof of concept will lay the foundation for a future autonomous collision avoidance system.

It’s not just startups making a splash in autonomous shipping. Radford noted that Rolls Royce—yes, that Rolls Royce—is leading the way in the development of autonomous ships. Its Intelligence Awareness system pulls in nearly every type of hyped technology on the market today: neural networks, augmented reality, virtual reality, and LiDAR.

In augmented reality mode, for example, a live feed video from the ship’s sensors can detect both static and moving objects, overlaying the scene with details about the types of vessels in the area, as well as their distance, heading, and other pertinent data.

While safety is a primary motivation for vessel automation—more than 1,100 ships have been lost over the past decade—these new technologies could make ships more efficient and less expensive to operate, according to a story in Wired about the Rolls Royce Intelligence Awareness system.

Sea Hunt Meets Science
As Singularity Hub noted in a previous article, ocean robots can also play a critical role in saving the seas from environmental threats. One poster child that has emerged—or, invaded—is the spindly lionfish.

A venomous critter endemic to the Indo-Pacific region, the lionfish is now found up and down the east coast of North America and beyond. And it is voracious, eating up to 30 times its own stomach volume and reducing juvenile reef fish populations by nearly 90 percent in as little as five weeks, according to the Ocean Support Foundation.

That has made the colorful but deadly fish Public Enemy No. 1 for many marine conservationists. Both researchers and startups are developing autonomous robots to hunt down the invasive predator.

At the Worcester Polytechnic Institute, for example, students are building a spear-carrying robot that uses machine learning and computer vision to distinguish lionfish from other aquatic species. The students trained the algorithms on thousands of different images of lionfish. The result: a lionfish-killing machine that boasts an accuracy of greater than 95 percent.

Meanwhile, a small startup called the American Marine Research Corporation out of Pensacola, Florida is applying similar technology to seek and destroy lionfish. Rather than spearfishing, the AMRC drone would stun and capture the lionfish, turning a profit by selling the creatures to local seafood restaurants.

Lionfish: It’s what’s for dinner.

Water Bots
A new wave of smart, independent robots are diving, swimming, and cruising across the ocean and its deepest depths. These autonomous systems aren’t necessarily designed to replace humans, but to venture where we can’t go or to improve safety at sea. And, perhaps, these latest innovations may inspire the robots that will someday plumb the depths of watery planets far from Earth.

Image Credit: Houston Mechatronics, Inc. Continue reading

Posted in Human Robots

#434297 How Can Leaders Ensure Humanity in a ...

It’s hard to avoid the prominence of AI in our lives, and there is a plethora of predictions about how it will influence our future. In their new book Solomon’s Code: Humanity in a World of Thinking Machines, co-authors Olaf Groth, Professor of Strategy, Innovation and Economics at HULT International Business School and CEO of advisory network Cambrian.ai, and Mark Nitzberg, Executive Director of UC Berkeley’s Center for Human-Compatible AI, believe that the shift in balance of power between intelligent machines and humans is already here.

I caught up with the authors about how the continued integration between technology and humans, and their call for a “Digital Magna Carta,” a broadly-accepted charter developed by a multi-stakeholder congress that would help guide the development of advanced technologies to harness their power for the benefit of all humanity.

Lisa Kay Solomon: Your new book, Solomon’s Code, explores artificial intelligence and its broader human, ethical, and societal implications that all leaders need to consider. AI is a technology that’s been in development for decades. Why is it so urgent to focus on these topics now?

Olaf Groth and Mark Nitzberg: Popular perception always thinks of AI in terms of game-changing narratives—for instance, Deep Blue beating Gary Kasparov at chess. But it’s the way these AI applications are “getting into our heads” and making decisions for us that really influences our lives. That’s not to say the big, headline-grabbing breakthroughs aren’t important; they are.

But it’s the proliferation of prosaic apps and bots that changes our lives the most, by either empowering or counteracting who we are and what we do. Today, we turn a rapidly growing number of our decisions over to these machines, often without knowing it—and even more often without understanding the second- and third-order effects of both the technologies and our decisions to rely on them.

There is genuine power in what we call a “symbio-intelligent” partnership between human, machine, and natural intelligences. These relationships can optimize not just economic interests, but help improve human well-being, create a more purposeful workplace, and bring more fulfillment to our lives.

However, mitigating the risks while taking advantage of the opportunities will require a serious, multidisciplinary consideration of how AI influences human values, trust, and power relationships. Whether or not we acknowledge their existence in our everyday life, these questions are no longer just thought exercises or fodder for science fiction.

In many ways, these technologies can challenge what it means to be human, and their ramifications already affect us in real and often subtle ways. We need to understand how

LKS: There is a lot of hype and misconceptions about AI. In your book, you provide a useful distinction between the cognitive capability that we often associate with AI processes, and the more human elements of consciousness and conscience. Why are these distinctions so important to understand?

OG & MN: Could machines take over consciousness some day as they become more powerful and complex? It’s hard to say. But there’s little doubt that, as machines become more capable, humans will start to think of them as something conscious—if for no other reason than our natural inclination to anthropomorphize.

Machines are already learning to recognize our emotional states and our physical health. Once they start talking that back to us and adjusting their behavior accordingly, we will be tempted to develop a certain rapport with them, potentially more trusting or more intimate because the machine recognizes us in our various states.

Consciousness is hard to define and may well be an emergent property, rather than something you can easily create or—in turn—deduce to its parts. So, could it happen as we put more and more elements together, from the realms of AI, quantum computing, or brain-computer interfaces? We can’t exclude that possibility.

Either way, we need to make sure we’re charting out a clear path and guardrails for this development through the Three Cs in machines: cognition (where AI is today); consciousness (where AI could go); and conscience (what we need to instill in AI before we get there). The real concern is that we reach machine consciousness—or what humans decide to grant as consciousness—without a conscience. If that happens, we will have created an artificial sociopath.

LKS: We have been seeing major developments in how AI is influencing product development and industry shifts. How is the rise of AI changing power at the global level?

OG & MN: Both in the public and private sectors, the data holder has the power. We’ve already seen the ascendance of about 10 “digital barons” in the US and China who sit on huge troves of data, massive computing power, and the resources and money to attract the world’s top AI talent. With these gaps already open between the haves and the have-nots on the technological and corporate side, we’re becoming increasingly aware that similar inequalities are forming at a societal level as well.

Economic power flows with data, leaving few options for socio-economically underprivileged populations and their corrupt, biased, or sparse digital footprints. By concentrating power and overlooking values, we fracture trust.

We can already see this tension emerging between the two dominant geopolitical models of AI. China and the US have emerged as the most powerful in both technological and economic terms, and both remain eager to drive that influence around the world. The EU countries are more contained on these economic and geopolitical measures, but they’ve leaped ahead on privacy and social concerns.

The problem is, no one has yet combined leadership on all three critical elements of values, trust, and power. The nations and organizations that foster all three of these elements in their AI systems and strategies will lead the future. Some are starting to recognize the need for the combination, but we found just 13 countries that have created significant AI strategies. Countries that wait too long to join them risk subjecting themselves to a new “data colonialism” that could change their economies and societies from the outside.

LKS: Solomon’s Code looks at AI from a variety of perspectives, considering both positive and potentially dangerous effects. You caution against the rising global threat and weaponization of AI and data, suggesting that “biased or dirty data is more threatening than nuclear arms or a pandemic.” For global leaders, entrepreneurs, technologists, policy makers and social change agents reading this, what specific strategies do you recommend to ensure ethical development and application of AI?

OG & MN: We’ve surrendered many of our most critical decisions to the Cult of Data. In most cases, that’s a great thing, as we rely more on scientific evidence to understand our world and our way through it. But we swing too far in other instances, assuming that datasets and algorithms produce a complete story that’s unsullied by human biases or intellectual shortcomings. We might choose to ignore it, but no one is blind to the dangers of nuclear war or pandemic disease. Yet, we willfully blind ourselves to the threat of dirty data, instead believing it to be pristine.

So, what do we do about it? On an individual level, it’s a matter of awareness, knowing who controls your data and how outsourcing of decisions to thinking machines can present opportunities and threats alike.

For business, government, and political leaders, we need to see a much broader expansion of ethics committees with transparent criteria with which to evaluate new products and services. We might consider something akin to clinical trials for pharmaceuticals—a sort of testing scheme that can transparently and independently measure the effects on humans of algorithms, bots, and the like. All of this needs to be multidisciplinary, bringing in expertise from across technology, social systems, ethics, anthropology, psychology, and so on.

Finally, on a global level, we need a new charter of rights—a Digital Magna Carta—that formalizes these protections and guides the development of new AI technologies toward all of humanity’s benefit. We’ve suggested the creation of a multi-stakeholder Cambrian Congress (harkening back to the explosion of life during the Cambrian period) that can not only begin to frame benefits for humanity, but build the global consensus around principles for a basic code-of-conduct, and ideas for evaluation and enforcement mechanisms, so we can get there without any large-scale failures or backlash in society. So, it’s not one or the other—it’s both.

Image Credit: whiteMocca / Shutterstock.com Continue reading

Posted in Human Robots

#433799 The First Novel Written by AI Is ...

Last year, a novelist went on a road trip across the USA. The trip was an attempt to emulate Jack Kerouac—to go out on the road and find something essential to write about in the experience. There is, however, a key difference between this writer and anyone else talking your ear off in the bar. This writer is just a microphone, a GPS, and a camera hooked up to a laptop and a whole bunch of linear algebra.

People who are optimistic that artificial intelligence and machine learning won’t put us all out of a job say that human ingenuity and creativity will be difficult to imitate. The classic argument is that, just as machines freed us from repetitive manual tasks, machine learning will free us from repetitive intellectual tasks.

This leaves us free to spend more time on the rewarding aspects of our work, pursuing creative hobbies, spending time with loved ones, and generally being human.

In this worldview, creative works like a great novel or symphony, and the emotions they evoke, cannot be reduced to lines of code. Humans retain a dimension of superiority over algorithms.

But is creativity a fundamentally human phenomenon? Or can it be learned by machines?

And if they learn to understand us better than we understand ourselves, could the great AI novel—tailored, of course, to your own predispositions in fiction—be the best you’ll ever read?

Maybe Not a Beach Read
This is the futurist’s view, of course. The reality, as the jury-rigged contraption in Ross Goodwin’s Cadillac for that road trip can attest, is some way off.

“This is very much an imperfect document, a rapid prototyping project. The output isn’t perfect. I don’t think it’s a human novel, or anywhere near it,” Goodwin said of the novel that his machine created. 1 The Road is currently marketed as the first novel written by AI.

Once the neural network has been trained, it can generate any length of text that the author desires, either at random or working from a specific seed word or phrase. Goodwin used the sights and sounds of the road trip to provide these seeds: the novel is written one sentence at a time, based on images, locations, dialogue from the microphone, and even the computer’s own internal clock.

The results are… mixed.

The novel begins suitably enough, quoting the time: “It was nine seventeen in the morning, and the house was heavy.” Descriptions of locations begin according to the Foursquare dataset fed into the algorithm, but rapidly veer off into the weeds, becoming surreal. While experimentation in literature is a wonderful thing, repeatedly quoting longitude and latitude coordinates verbatim is unlikely to win anyone the Booker Prize.

Data In, Art Out?
Neural networks as creative agents have some advantages. They excel at being trained on large datasets, identifying the patterns in those datasets, and producing output that follows those same rules. Music inspired by or written by AI has become a growing subgenre—there’s even a pop album by human-machine collaborators called the Songularity.

A neural network can “listen to” all of Bach and Mozart in hours, and train itself on the works of Shakespeare to produce passable pseudo-Bard. The idea of artificial creativity has become so widespread that there’s even a meme format about forcibly training neural network ‘bots’ on human writing samples, with hilarious consequences—although the best joke was undoubtedly human in origin.

The AI that roamed from New York to New Orleans was an LSTM (long short-term memory) neural net. By default, information contained in individual neurons is preserved, and only small parts can be “forgotten” or “learned” in an individual timestep, rather than neurons being entirely overwritten.

The LSTM architecture performs better than previous recurrent neural networks at tasks such as handwriting and speech recognition. The neural net—and its programmer—looked further in search of literary influences, ingesting 60 million words (360 MB) of raw literature according to Goodwin’s recipe: one third poetry, one third science fiction, and one third “bleak” literature.

In this way, Goodwin has some creative control over the project; the source material influences the machine’s vocabulary and sentence structuring, and hence the tone of the piece.

The Thoughts Beneath the Words
The problem with artificially intelligent novelists is the same problem with conversational artificial intelligence that computer scientists have been trying to solve from Turing’s day. The machines can understand and reproduce complex patterns increasingly better than humans can, but they have no understanding of what these patterns mean.

Goodwin’s neural network spits out sentences one letter at a time, on a tiny printer hooked up to the laptop. Statistical associations such as those tracked by neural nets can form words from letters, and sentences from words, but they know nothing of character or plot.

When talking to a chatbot, the code has no real understanding of what’s been said before, and there is no dataset large enough to train it through all of the billions of possible conversations.

Unless restricted to a predetermined set of options, it loses the thread of the conversation after a reply or two. In a similar way, the creative neural nets have no real grasp of what they’re writing, and no way to produce anything with any overarching coherence or narrative.

Goodwin’s experiment is an attempt to add some coherent backbone to the AI “novel” by repeatedly grounding it with stimuli from the cameras or microphones—the thematic links and narrative provided by the American landscape the neural network drives through.

Goodwin feels that this approach (the car itself moving through the landscape, as if a character) borrows some continuity and coherence from the journey itself. “Coherent prose is the holy grail of natural-language generation—feeling that I had somehow solved a small part of the problem was exhilarating. And I do think it makes a point about language in time that’s unexpected and interesting.”

AI Is Still No Kerouac
A coherent tone and semantic “style” might be enough to produce some vaguely-convincing teenage poetry, as Google did, and experimental fiction that uses neural networks can have intriguing results. But wading through the surreal AI prose of this era, searching for some meaning or motif beyond novelty value, can be a frustrating experience.

Maybe machines can learn the complexities of the human heart and brain, or how to write evocative or entertaining prose. But they’re a long way off, and somehow “more layers!” or a bigger corpus of data doesn’t feel like enough to bridge that gulf.

Real attempts by machines to write fiction have so far been broadly incoherent, but with flashes of poetry—dreamlike, hallucinatory ramblings.

Neural networks might not be capable of writing intricately-plotted works with charm and wit, like Dickens or Dostoevsky, but there’s still an eeriness to trying to decipher the surreal, Finnegans’ Wake mish-mash.

You might see, in the odd line, the flickering ghost of something like consciousness, a deeper understanding. Or you might just see fragments of meaning thrown into a neural network blender, full of hype and fury, obeying rules in an occasionally striking way, but ultimately signifying nothing. In that sense, at least, the RNN’s grappling with metaphor feels like a metaphor for the hype surrounding the latest AI summer as a whole.

Or, as the human author of On The Road put it: “You guys are going somewhere or just going?”

Image Credit: eurobanks / Shutterstock.com Continue reading

Posted in Human Robots