Tag Archives: believe
There is a saying that has emerged among the tech set in recent years: AI is the new electricity. The platitude refers to the disruptive power of artificial intelligence for driving advances in everything from transportation to predicting the weather.
Of course, the computers and data centers that support AI’s complex algorithms are very much dependent on electricity. While that may seem pretty obvious, it may be surprising to learn that AI can be extremely power-hungry, especially when it comes to training the models that enable machines to recognize your face in a photo or for Alexa to understand a voice command.
The scale of the problem is difficult to measure, but there have been some attempts to put hard numbers on the environmental cost.
For instance, one paper published on the open-access repository arXiv claimed that the carbon emissions for training a basic natural language processing (NLP) model—algorithms that process and understand language-based data—are equal to the CO2 produced by the average American lifestyle over two years. A more robust model required the equivalent of about 17 years’ worth of emissions.
The authors noted that about a decade ago, NLP models could do the job on a regular commercial laptop. Today, much more sophisticated AI models use specialized hardware like graphics processing units, or GPUs, a chip technology popularized by Nvidia for gaming that also proved capable of supporting computing tasks for AI.
OpenAI, a nonprofit research organization co-founded by tech prophet and profiteer Elon Musk, said that the computing power “used in the largest AI training runs has been increasing exponentially with a 3.4-month doubling time” since 2012. That’s about the time that GPUs started making their way into AI computing systems.
Getting Smarter About AI Chip Design
While GPUs from Nvidia remain the gold standard in AI hardware today, a number of startups have emerged to challenge the company’s industry dominance. Many are building chipsets designed to work more like the human brain, an area that’s been dubbed neuromorphic computing.
One of the leading companies in this arena is Graphcore, a UK startup that has raised more than $450 million and boasts a valuation of $1.95 billion. The company’s version of the GPU is an IPU, which stands for intelligence processing unit.
To build a computer brain more akin to a human one, the big brains at Graphcore are bypassing the precise but time-consuming number-crunching typical of a conventional microprocessor with one that’s content to get by on less precise arithmetic.
The results are essentially the same, but IPUs get the job done much quicker. Graphcore claimed it was able to train the popular BERT NLP model in just 56 hours, while tripling throughput and reducing latency by 20 percent.
An article in Bloomberg compared the approach to the “human brain shifting from calculating the exact GPS coordinates of a restaurant to just remembering its name and neighborhood.”
Graphcore’s hardware architecture also features more built-in memory processing, boosting efficiency because there’s less need to send as much data back and forth between chips. That’s similar to an approach adopted by a team of researchers in Italy that recently published a paper about a new computing circuit.
The novel circuit uses a device called a memristor that can execute a mathematical function known as a regression in just one operation. The approach attempts to mimic the human brain by processing data directly within the memory.
Daniele Ielmini at Politecnico di Milano, co-author of the Science Advances paper, told Singularity Hub that the main advantage of in-memory computing is the lack of any data movement, which is the main bottleneck of conventional digital computers, as well as the parallel processing of data that enables the intimate interactions among various currents and voltages within the memory array.
Ielmini explained that in-memory computing can have a “tremendous impact on energy efficiency of AI, as it can accelerate very advanced tasks by physical computation within the memory circuit.” He added that such “radical ideas” in hardware design will be needed in order to make a quantum leap in energy efficiency and time.
It’s Not Just a Hardware Problem
The emphasis on designing more efficient chip architecture might suggest that AI’s power hunger is essentially a hardware problem. That’s not the case, Ielmini noted.
“We believe that significant progress could be made by similar breakthroughs at the algorithm and dataset levels,” he said.
He’s not the only one.
One of the key research areas at Qualcomm’s AI research lab is energy efficiency. Max Welling, vice president of Qualcomm Technology R&D division, has written about the need for more power-efficient algorithms. He has gone so far as to suggest that AI algorithms will be measured by the amount of intelligence they provide per joule.
One emerging area being studied, Welling wrote, is the use of Bayesian deep learning for deep neural networks.
It’s all pretty heady stuff and easily the subject of a PhD thesis. The main thing to understand in this context is that Bayesian deep learning is another attempt to mimic how the brain processes information by introducing random values into the neural network. A benefit of Bayesian deep learning is that it compresses and quantifies data in order to reduce the complexity of a neural network. In turn, that reduces the number of “steps” required to recognize a dog as a dog—and the energy required to get the right result.
A team at Oak Ridge National Laboratory has previously demonstrated another way to improve AI energy efficiency by converting deep learning neural networks into what’s called a spiking neural network. The researchers spiked their deep spiking neural network (DSNN) by introducing a stochastic process that adds random values like Bayesian deep learning.
The DSNN actually imitates the way neurons interact with synapses, which send signals between brain cells. Individual “spikes” in the network indicate where to perform computations, lowering energy consumption because it disregards unnecessary computations.
The system is being used by cancer researchers to scan millions of clinical reports to unearth insights on causes and treatments of the disease.
Helping battle cancer is only one of many rewards we may reap from artificial intelligence in the future, as long as the benefits of those algorithms outweigh the costs of using them.
“Making AI more energy-efficient is an overarching objective that spans the fields of algorithms, systems, architecture, circuits, and devices,” Ielmini said.
Image Credit: analogicus from Pixabay Continue reading
Assuming that the emergence of consciousness in artificial minds is possible, those minds will feel the urge to create art. But will we be able to understand it? To answer this question, we need to consider two subquestions: when does the machine become an author of an artwork? And how can we form an understanding of the art that it makes?
Empathy, we argue, is the force behind our capacity to understand works of art. Think of what happens when you are confronted with an artwork. We maintain that, to understand the piece, you use your own conscious experience to ask what could possibly motivate you to make such an artwork yourself—and then you use that first-person perspective to try to come to a plausible explanation that allows you to relate to the artwork. Your interpretation of the work will be personal and could differ significantly from the artist’s own reasons, but if we share sufficient experiences and cultural references, it might be a plausible one, even for the artist. This is why we can relate so differently to a work of art after learning that it is a forgery or imitation: the artist’s intent to deceive or imitate is very different from the attempt to express something original. Gathering contextual information before jumping to conclusions about other people’s actions—in art, as in life—can enable us to relate better to their intentions.
But the artist and you share something far more important than cultural references: you share a similar kind of body and, with it, a similar kind of embodied perspective. Our subjective human experience stems, among many other things, from being born and slowly educated within a society of fellow humans, from fighting the inevitability of our own death, from cherishing memories, from the lonely curiosity of our own mind, from the omnipresence of the needs and quirks of our biological body, and from the way it dictates the space- and time-scales we can grasp. All conscious machines will have embodied experiences of their own, but in bodies that will be entirely alien to us.
We are able to empathize with nonhuman characters or intelligent machines in human-made fiction because they have been conceived by other human beings from the only subjective perspective accessible to us: “What would it be like for a human to behave as x?” In order to understand machinic art as such—and assuming that we stand a chance of even recognizing it in the first place—we would need a way to conceive a first-person experience of what it is like to be that machine. That is something we cannot do even for beings that are much closer to us. It might very well happen that we understand some actions or artifacts created by machines of their own volition as art, but in doing so we will inevitably anthropomorphize the machine’s intentions. Art made by a machine can be meaningfully interpreted in a way that is plausible only from the perspective of that machine, and any coherent anthropomorphized interpretation will be implausibly alien from the machine perspective. As such, it will be a misinterpretation of the artwork.
But what if we grant the machine privileged access to our ways of reasoning, to the peculiarities of our perception apparatus, to endless examples of human culture? Wouldn’t that enable the machine to make art that a human could understand? Our answer is yes, but this would also make the artworks human—not authentically machinic. All examples so far of “art made by machines” are actually just straightforward examples of human art made with computers, with the artists being the computer programmers. It might seem like a strange claim: how can the programmers be the authors of the artwork if, most of the time, they can’t control—or even anticipate—the actual materializations of the artwork? It turns out that this is a long-standing artistic practice.
Suppose that your local orchestra is playing Beethoven’s Symphony No 7 (1812). Even though Beethoven will not be directly responsible for any of the sounds produced there, you would still say that you are listening to Beethoven. Your experience might depend considerably on the interpretation of the performers, the acoustics of the room, the behavior of fellow audience members or your state of mind. Those and other aspects are the result of choices made by specific individuals or of accidents happening to them. But the author of the music? Ludwig van Beethoven. Let’s say that, as a somewhat odd choice for the program, John Cage’s Imaginary Landscape No 4 (March No 2) (1951) is also played, with 24 performers controlling 12 radios according to a musical score. In this case, the responsibility for the sounds being heard should be attributed to unsuspecting radio hosts, or even to electromagnetic fields. Yet, the shaping of sounds over time—the composition—should be credited to Cage. Each performance of this piece will vary immensely in its sonic materialization, but it will always be a performance of Imaginary Landscape No 4.
Why should we change these principles when artists use computers if, in these respects at least, computer art does not bring anything new to the table? The (human) artists might not be in direct control of the final materializations, or even be able to predict them but, despite that, they are the authors of the work. Various materializations of the same idea—in this case formalized as an algorithm—are instantiations of the same work manifesting different contextual conditions. In fact, a common use of computation in the arts is the production of variations of a process, and artists make extensive use of systems that are sensitive to initial conditions, external inputs, or pseudo-randomness to deliberately avoid repetition of outputs. Having a computer executing a procedure to build an artwork, even if using pseudo-random processes or machine-learning algorithms, is no different than throwing dice to arrange a piece of music, or to pursuing innumerable variations of the same formula. After all, the idea of machines that make art has an artistic tradition that long predates the current trend of artworks made by artificial intelligence.
Machinic art is a term that we believe should be reserved for art made by an artificial mind’s own volition, not for that based on (or directed towards) an anthropocentric view of art. From a human point of view, machinic artworks will still be procedural, algorithmic, and computational. They will be generative, because they will be autonomous from a human artist. And they might be interactive, with humans or other systems. But they will not be the result of a human deferring decisions to a machine, because the first of those—the decision to make art—needs to be the result of a machine’s volition, intentions, and decisions. Only then will we no longer have human art made with computers, but proper machinic art.
The problem is not whether machines will or will not develop a sense of self that leads to an eagerness to create art. The problem is that if—or when—they do, they will have such a different Umwelt that we will be completely unable to relate to it from our own subjective, embodied perspective. Machinic art will always lie beyond our ability to understand it because the boundaries of our comprehension—in art, as in life—are those of the human experience.
This article was originally published at Aeon and has been republished under Creative Commons.
Image Credit: Rene Böhmer / Unsplash Continue reading
This week at MIT, academics and industry officials compared notes, studies, and predictions about AI and the future of work. During the discussions, an insurance company executive shared details about one AI program that rolled out at his firm earlier this year. A chatbot the company introduced, the executive said, now handles 150,000 calls per month.
Later in the day, a panelist—David Fanning, founder of PBS’s Frontline—remarked that this statistic is emblematic of broader fears he saw when reporting a new Frontline documentary about AI. “People are scared,” Fanning said of the public’s AI anxiety.
Fanning was part of a daylong symposium about AI’s economic consequences—good, bad, and otherwise—convened by MIT’s Task Force on the Work of the Future.
“Dig into every industry, and you’ll find AI changing the nature of work,” said Daniela Rus, director of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). She cited recent McKinsey research that found 45 percent of the work people are paid to do today can be automated with currently available technologies. Those activities, McKinsey found, represent some US $2 trillion in wages.
However, the threat of automation—whether by AI or other technologies—isn’t as new as technologists on America’s coasts seem to believe, said panelist Fred Goff, CEO of Jobcase, Inc.
“If you live in Detroit or Toledo, where I come from, technology has been displacing jobs for the last half-century,” Goff said. “I don’t think that most people in this country have the increased anxiety that the coasts do, because they’ve been living this.”
Goff added that the challenge AI poses for the workforce is not, as he put it, “getting coal miners to code.” Rather, he said, as AI automates some jobs, it will also open opportunities for “reskilling” that may have nothing to do with AI or automation. He touted trade schools—teaching skills like welding, plumbing, and electrical work—and certification programs for sales industry software packages like Salesforce.
On the other hand, a documentarian who reported another recent program on AI—Krishna Andavolu, senior correspondent for Vice Media—said “reskilling” may not be an easy answer.
“People in rooms like this … don’t realize that a lot of people don’t want to work that much,” Andavolu said. “They’re not driven by passion for their career, they’re driven by passion for life. We’re telling a lot of these workers that they need to reskill. But to a lot of people that sounds like, ‘I’ve got to work twice as hard for what I have now.’ That sounds scary. We underestimate that at our peril.”
Part of the problem with “reskilling,” Andavolu said, is that some high-growth industries involve caregiving for seniors and in medical facilities—roles which are traditionally considered “feminized” careers. Destigmatizing these jobs, and increasing the pay to match the salaries of displaced jobs like long-haul truck drivers, is another challenge.
Daron Acemoglu, MIT Institute Professor of Economics, faulted the comparably slim funding of academic research into AI.
“There is nothing preordained about the progress of technology,” he said. Computers, the Internet, antibiotics, and sensors all grew out of government and academic research programs. What he called the “blue-sky thinking” of non-corporate AI research can also develop applications that are not purely focused on maximizing profits.
American companies, Acemoglu said, get tax breaks for capital R&D—but not for developing new technologies for their employees. “We turn around and [tell companies], ‘Use your technologies to empower workers,’” he said. “But why should they do that? Hiring workers is expensive in many ways. And we’re subsidizing capital.”
Said Sarita Gupta, director of the Ford Foundation’s Future of Work(ers) Program, “Low and middle income workers have for over 30 years been experiencing stagnant and declining pay, shrinking benefits, and less power on the job. Now technology is brilliant at enabling scale. But the question we sit with is—how do we make sure that we’re not scaling these longstanding problems?”
Andrew McAfee, co-director of MIT’s Initiative on the Digital Economy, said AI may not reduce the number of jobs available in the workplace today. But the quality of those jobs is another story. He cited the Dutch economist Jan Tinbergen who decades ago said that “Inequality is a race between technology and education.”
McAfee said, ultimately, the time to solve the economic problems AI poses for workers in the United States is when the U.S. economy is doing well—like right now.
“We do have the wind at our backs,” said Elisabeth Reynolds, executive director of MIT’s Task Force on the Work of the Future.
“We have some breathing room right now,” McAfee agreed. “Economic growth has been pretty good. Unemployment is pretty low. Interest rates are very, very low. We might not have that war chest in the future.” Continue reading