Are we in an AI summer or AI winter?
Neither. We are entering an AI autumn.
Adam Frank is a professor of astrophysics at the University of Rochester and a leading expert on the final stages of evolution for stars like the sun. Frank's computational research group at the University of Rochester has developed advanced supercomputer tools for studying how stars form and how they die. A self-described “evangelist of science," he is the author of four books and the co-founder of 13.8, where he explores the beauty and power of science in culture with physicist Marcelo Gleiser.
- The history of AI shows boom periods (AI summers) followed by busts (AI winters).
- The cyclical nature of AI funding is due to hype and promises not fulfilling expectations.
- This time, we might enter something resembling an AI autumn rather than an AI winter, but fundamental questions remain if true AI is even possible.
The dream of building a machine that can think like a human stretches back to the origins of electronic computers. But ever since research into artificial intelligence (AI) began in earnest after World War II, the field has gone through a series of boom and bust cycles called "AI summers" and "AI winters."
Each cycle begins with optimistic claims that a fully, generally intelligent machine is just a decade or so away. Funding pours in and progress seems swift. Then, a decade or so later, progress stalls and funding dries up. Over the last ten years, we've clearly been in an AI summer as vast improvements in computing power and new techniques like deep learning have led to remarkable advances. But now, as we enter the third decade of the 21st century, some who follow AI feel the cold winds at their back leading them to ask, "Is Winter Coming?" If so, what went wrong this time?
How to build an A.I. brain that can conceive of itself | Joscha Bach | Big Think www.youtube.com
A brief history of AI
To see if the winds of winter are really coming for AI, it is useful to look at the field's history. The first real summer can be pegged to 1956 and the famous Dartmouth University Workshop where one of the field's pioneers, John McCarthy, coined the term "artificial intelligence." The conference was attended by scientists like Marvin Minsky and H. A. Simon, whose names would go on to become synonymous with the field. For those researchers, the task ahead was clear: capture the processes of human reasoning through the manipulation of symbolic systems (i.e., computer programs).
Unless we are talking about very specific tasks, any 6-year-old is infinitely more flexible and general in his or her intelligence than the "smartest" Amazon robot.
Throughout the 1960s, progress seemed to come swiftly as researchers developed computer systems that could play chess, deduce mathematical theorems, and even engage in simple discussions with a person. Government funding flowed generously. Optimism was so high that, in 1970, Minsky famously proclaimed, "In three to eight years we will have a machine with the general intelligence of a human being."
By the mid 1970s, however, it was clear that Minsky's optimism was unwarranted. Progress stalled as many of the innovations of the previous decade proved too narrow in their applicability, seeming more like toys than steps toward a general version of artificial intelligence. Funding dried up so completely that researchers soon took pains not to refer to their work as AI, as the term carried a stink that killed proposals.
The cycle repeated itself in the 1980s with the rise of expert systems and the renewed interest in what we now call neural networks (i.e., programs based on connectivity architectures that mimic neurons in the brain). Once again, there was wild optimism and big increases in funding. What was novel in this cycle was the addition of significant private funding as more companies began to rely on computers as essential components of their business. But, once again, the big promises were never realized, and funding dried up again.
AI: Hype vs. reality
The AI summer we're currently experiencing began sometime in the first decade of the new millennium. Vast increases in both computing speed and storage ushered in the era of deep learning and big data. Deep learning methods use stacked layers of neural networks that pass information to each other to solve complex problems like facial recognition. Big data provides these systems with vast oceans of examples (like images of faces) to train on. The applications of this progress are all around us: Google Maps give you near-perfect directions; you can talk with Siri anytime you want; IBM's Deep Think computer beat Jeopardy's greatest human champions.
In response, the hype rose again. True AI, we were told, must be just around the corner. In 2015, for example, The Guardian reported that self-driving cars, the killer app of modern AI, was close at hand. Readers were told, "By 2020 you will become a permanent backseat driver." And just two years ago, Elon Musk claimed that by 2020 "we'd have over a million cars with full self-driving software."
The general intelligence — i.e., the understanding — we humans exhibit may be inseparable from our experiencing. If that's true, then our physical embodiment, enmeshed in a context-rich world, may be difficult if not impossible to capture in symbolic processing systems.
By now, it's obvious that a world of fully self-driving cars is still years away. Likewise, in spite of the remarkable progress we've made in machine learning, we're still far from creating systems that possess general intelligence. The emphasis is on the term general because that's what AI really has been promising all these years: a machine that's flexible in dealing with any situation as it comes up. Instead, what researchers have found is that, despite all their remarkable progress, the systems they've built remain brittle, which is a technical term meaning "they do very wrong things when given unexpected inputs." Try asking Siri to find "restaurants that aren't McDonald's." You won't like the results.
Unless we are talking about very specific tasks, any 6-year-old is infinitely more flexible and general in his or her intelligence than the "smartest" Amazon robot.
Even more important is the sense that, as remarkable as they are, none of the systems we've built understand anything about what they are doing. As philosopher Alva Noe said of Deep Think's famous Jeopardy! victory, "Watson answered no questions. It participated in no competition. It didn't do anything. All the doing was on our side. We played Jeapordy! with Watson." Considering this fact, some researchers claim that the general intelligence — i.e., the understanding — we humans exhibit may be inseparable from our experiencing. If that's true, then our physical embodiment, enmeshed in a context-rich world, may be difficult if not impossible to capture in symbolic processing systems.
Not the (AI) winter of our discontent
Thus, talk a of a new AI winter is popping up again. Given the importance of deep learning and big data in technology, it's hard to imagine funding for these domains drying up any time soon. What we may be seeing, however, is a kind of AI autumn when researchers wisely recalibrate their expectations and perhaps rethink their perspectives.
Cross-disciplinary cooperation is needed to save civilization.
- There is a great disconnect between the sciences and the humanities.
- Solutions to most of our real-world problems need both ways of knowing.
- Moving beyond the two-culture divide is an essential step to ensure our project of civilization.
For the past five years, I ran the Institute for Cross-Disciplinary Engagement at Dartmouth, an initiative sponsored by the John Templeton Foundation. Our mission has been to find ways to bring scientists and humanists together, often in public venues or — after Covid-19 — online, to discuss questions that transcend the narrow confines of a single discipline.
It turns out that these questions are at the very center of the much needed and urgent conversation about our collective future. While the complexity of the problems we face asks for a multi-cultural integration of different ways of knowing, the tools at hand are scarce and mostly ineffective. We need to rethink and learn how to collaborate productively across disciplinary cultures.
The danger of hyper-specialization
The explosive expansion of knowledge that started in the mid 1800s led to hyper-specialization inside and outside academia. Even within a single discipline, say philosophy or physics, professionals often don't understand one another. As I wrote here before, "This fragmentation of knowledge inside and outside of academia is the hallmark of our times, an amplification of the clash of the Two Cultures that physicist and novelist C.P. Snow admonished his Cambridge colleagues in 1959." The loss is palpable, intellectually and socially. Knowledge is not adept to reductionism. Sure, a specialist will make progress in her chosen field, but the tunnel vision of hyper-specialization creates a loss of context: you do the work not knowing how it fits into the bigger picture or, more alarmingly, how it may impact society.
Many of the existential risks we face today — AI and its impact on the workforce, the dangerous loss of privacy due to data mining and sharing, the threat of cyberwarfare, the threat of biowarfare, the threat of global warming, the threat of nuclear terrorism, the threat to our humanity by the development of genetic engineering — are consequences of the growing ease of access to cutting-edge technologies and the irreversible dependence we all have on our gadgets. Technological innovation is seductive: we want to have the latest "smart" phone, 5k TV, and VR goggles because they are objects of desire and social placement.
Are we ready for the genetic revolution?
When the time comes, and experts believe it is coming sooner than we expect or are prepared for, genetic meddling with the human genome may drive social inequality to an unprecedented level with not just differences in wealth distribution but in what kind of being you become and who retains power. This is the kind of nightmare that Nobel Prize-winning geneticist Jennifer Doudna talked about in a recent Big Think video.
CRISPR 101: Curing Sickle Cell, Growing Organs, Mosquito Makeovers | Jennifer Doudna | Big Think www.youtube.com
At the heart of these advances is the dual-use nature of science, its light and shadow selves. Most technological developments are perceived and sold as spectacular advances that will either alleviate human suffering or bring increasing levels of comfort and accessibility to a growing number of people. Curing diseases is what motivated Doudna and other scientists involved with CRISPR research. But with that also came the potential for altering the genetic makeup of humanity in ways that, again, can be used for good or evil purposes.
This is not a sci-fi movie plot. The main difference between biohacking and nuclear hacking is one of scale. Nuclear technologies require industrial-level infrastructure, which is very costly and demanding. This is why nuclear research and its technological implementation have been mostly relegated to governments. Biohacking can be done in someone's backyard garage with equipment that is not very costly. The Netflix documentary series Unnatural Selection brings this point home in terrifying ways. The essential problem is this: once the genie is out of the bottle, it is virtually impossible to enforce any kind of control. The genie will not be pushed back in.
Cross-disciplinary cooperation is needed to save civilization
What, then, can be done? Such technological challenges go beyond the reach of a single discipline. CRISPR, for example, may be an invention within genetics, but its impact is vast, asking for oversight and ethical safeguards that are far from our current reality. The same with global warming, rampant environmental destruction, and growing levels of air pollution/greenhouse gas emissions that are fast emerging as we crawl into a post-pandemic era. Instead of learning the lessons from our 18 months of seclusion — that we are fragile to nature's powers, that we are co-dependent and globally linked in irreversible ways, that our individual choices affect many more than ourselves — we seem to be bent on decompressing our accumulated urges with impunity.
The experience from our experiment with the Institute for Cross-Disciplinary Engagement has taught us a few lessons that we hope can be extrapolated to the rest of society: (1) that there is huge public interest in this kind of cross-disciplinary conversation between the sciences and the humanities; (2) that there is growing consensus in academia that this conversation is needed and urgent, as similar institutes emerge in other schools; (3) that in order for an open cross-disciplinary exchange to be successful, a common language needs to be established with people talking to each other and not past each other; (4) that university and high school curricula should strive to create more courses where this sort of cross-disciplinary exchange is the norm and not the exception; (5) that this conversation needs to be taken to all sectors of society and not kept within isolated silos of intellectualism.
Moving beyond the two-culture divide is not simply an interesting intellectual exercise; it is, as humanity wrestles with its own indecisions and uncertainties, an essential step to ensure our project of civilization.
New study analyzes gravitational waves to confirm the late Stephen Hawking's black hole area theorem.
- A new paper confirms Stephen Hawking's black hole area theorem.
- The researchers used gravitational wave data to prove the theorem.
- The data came from Caltech and MIT's Advanced Laser Interferometer Gravitational-Wave Observatory.
The late Stephen Hawking's black hole area theorem is correct, a new study shows. Scientists used gravitational waves to prove the famous British physicist's idea, which may lead to uncovering more underlying laws of the universe.
The theorem, elaborated by Hawking in 1971, uses Einstein's theory of general relativity as a springboard to conclude that it is not possible for the surface area of a black hole to become smaller over time. The theorem parallels the second law of thermodynamics that says the entropy (disorder) of a closed system can't decrease over time. Since the entropy of a black hole is proportional to its surface area, both must continue to increase.
As a black hole gobbles up more matter, its mass and surface area grow. But as it grows, it also spins faster, which decreases its surface area. Hawking's theorem maintains that the increase in surface area that comes from the added mass would always be larger than the decrease in surface area because of the added spin.
Will Farr, one of the co-authors of the study that was published in Physical Review Letters, said their finding demonstrates that "black hole areas are something fundamental and important." His colleague Maximiliano Isi agreed in an interview with Live Science: "Black holes have an entropy, and it's proportional to their area. It's not just a funny coincidence, it's a deep fact about the world that they reveal."
What are gravitational waves?
Gravitational waves are "ripples" in spacetime, predicted by Albert Einstein in 1916, that are created by very violent processes happening in space. Einstein showed that very massive, accelerating space objects like neutron stars or black holes that orbit each other could cause disturbances in spacetime. Like the ripples produced by tossing a rock into a lake, they would bring about "waves" of spacetime that would spread in all directions.
As LIGO shared, "These cosmic ripples would travel at the speed of light, carrying with them information about their origins, as well as clues to the nature of gravity itself."
The gravitational waves discovered by LIGO's 3,000-kilometer-long laser beam, which can detect the smallest distortions in spacetime, were generated 1.3 billion years ago by two giant black holes that were quickly spiraling toward each other.
What Stephen Hawking would have discovered if he lived longer | NASA's Michelle Thaller | Big Think www.youtube.com
Confirming Hawking's black hole area theorem
The researchers separated the signal into two parts, depending on whether it was from before or after the black holes merged. This allowed them to figure out the mass and spin of the original black holes as well as the mass and spin of the merged black hole. With this information, they calculated the surface areas of the black holes before and after the merger.
"As they spin around each other faster and faster, the gravitational waves increase in amplitude more and more until they eventually plunge into each other — making this big burst of waves," Isi elaborated. "What you're left with is a new black hole that's in this excited state, which you can then study by analyzing how it's vibrating. It's like if you ping a bell, the specific pitches and durations it rings with will tell you the structure of that bell, and also what it's made out of."
The surface area of the resulting black holes was larger than the combined area of the original black holes. This conformed to Hawking's area law.
As a form of civil disobedience, hacking can help make the world a better place.
- Hackers' motivations range from altruistic to nihilistic.
- Altruistic hackers expose injustices, while nihilistic ones make society more dangerous.
- The line between ethical and unethical hacking is not always clear.
The following is an excerpt from Coding Democracy by Maureen Webb, which is publishing in paperback on July 21. Reprinted with Permission from The MIT PRESS. Copyright 2020.
As people begin to hack more concertedly at the structures of the status quo, the reactions of those who benefit from things as they are will become more fierce and more punitive, at least until the "hackers" succeed in shifting the relevant power relationships. We know this from the history of social movements. At the dawning of the digital age, farmers who hack tractors will be ruthlessly punished.
Somewhere on the continuum of altruism and transgression is the kind of hacking that might lead the world toward more accountable government and informed citizenries.
Of course, it must be acknowledged that hackers are engaged in a whole range of acts, from the altruistic to the plainly nihilistic and dangerous. On the altruistic side of the continuum, they are creating free software (GNU/Linux and other software under GPL licenses), Creative Commons (Creative Commons licensing), and Open Access (designing digital interfaces to make public records and publicly funded research accessible). They are hacking surveillance and monopoly power (creating privacy tools, alternative services, cooperative platforms, and a new decentralized internet) and electoral politics and decision making (Cinque Stelle, En Comú, Ethelo, Liquid Democracy, and PartidoX). They have engaged in stunts to expose the technical flaws in voting, communications, and security systems widely used by, or imposed on, the public (by playing chess with Germany's election voting machines, hacking the German Bildschirmtext system, and stealing ministers' biometric identifiers). They have punished shady contractors like HackingTeam, HBGary, and Stratfor, spilling their corporate dealings and personal information across the internet. They have exposed the corruption of oligarchs, politicians, and hegemons (through the Panama Papers, WikiLeaks, and Xnet).
More notoriously, they have coordinated distributed denial of service (DDoS) attacks to retaliate against corporate and government conduct (such as the Anonymous DDoS that protested PayPal's boycott of WikiLeaks; the ingenious use of the Internet of Things to DDoS Amazon; and the shutdown of US and Canadian government IT systems). They have hacked into databases (Manning and Snowden), leaked state secrets (Manning, Snowden, and WikiLeaks), and, in doing so, betrayed their own governments (Manning betrayed US war secrets, and Snowden betrayed US security secrets). They have interfered with elections (such as the hack and leak of the Democratic National Committee in the middle of the 2016 US election) and sown disinformation (the Russian hacking of US social media). They have interfered with property rights in order to assert user ownership, self-determination, and free software's four freedoms (farmers have hacked DRM code to repair their tractors, and Geohot unlocked the iPhone and hacked the Samsung phone to allow users administrator-level access to their devices) and to assert open access to publicly funded research. They have created black markets to evade state justice systems (such as Silk Road on the dark web) and cryptocurrencies that could undermine state-regulated monetary systems. They have meddled in geopolitics as free agents (Anonymous and the Arab Spring, and Julian Assange and his conduct with the Trump campaign). They have mucked around in and could potentially impair or shut down critical infrastructure. (The notorious "WANK worm" attack on NASA is an early, notorious, example, but hackers could potentially target banking systems, stock exchanges, electrical grids, telecommunications systems, air traffic control, chemical plants, nuclear plants, and even military "doomsday machines.")
It is impossible to calculate where these acts nudge us as a species. Some uses of hacking — such as the malicious, nihilistic hacking that harms critical infrastructure and threatens lives, and the hacking in cyberwarfare that injures the critical interests of other countries and undermines their democratic processes — are abhorrent and cannot be defended. The unfolding digital era looks very grim when one considers the threat this kind of hacking poses to peace and democracy combined with the dystopian direction states and corporations are going with digital tech.
But somewhere on the continuum of altruism and transgression is the kind of hacking that might lead the world toward more accountable government and informed citizenries, less corrupt and unfair economic systems, wiser public uses of digital tech, more self-determination for the ordinary user, fairer commercial contracts, better conditions for innovation and creativity, more decentralized and robust infrastructure systems, and an abolition of doomsday machines. In short, some hacking might move us toward a digital world in which there are more rather than fewer democratic, humanist outcomes.
It is not clear where the line between "good" and "bad" hacking should be drawn or how to regulate it wisely in every instance. Citizens should inform themselves and begin to consider this line-drawing seriously, however, since we will be grappling intensely with it for the next century or more. My personal view is that digital tech should not be used for everything. I think we should go back to simpler ways of running electrical grids and elections, for example. Systems are more resilient when they are not wholly digital and when they are smaller, more local, and modular. Consumers should have analogue options for things like fridges and cars, and design priorities for household goods should be durability and clean energy use, not interconnectedness.
In setting legal standards, prohibiting something and enforcing the prohibition are two different things. Sometimes a desired social norm can be struck by prohibiting a thing and not enforcing it strenuously. And the law can also recognize the constructive role that civil disobedience plays in the evolution of social norms, through prosecutorial discretion and judicial discretion in sentencing.
Wau Holland told the young hackers at the Paradiso that the Chaos Computer Club was "not just a bunch of techno freaks: we've been thinking about the social consequences of technology from the very beginning." Societies themselves, however, are generally just beginning to grapple with the social consequences of digital technology and with how to characterize the various acts performed by hackers, morally and legally. Each act raises a set of complex questions. Societies' responses will be part of the dialectic that determines where we end up. Should these various hacker acts be treated as incidents of public service, free speech, free association, legitimate protest, civil disobedience, and harmless pranksterism? Or should they be treated as trespass, tortious interference, intellectual property infringement, theft, fraud, conspiracy, extortion, espionage, terrorism, and treason? I invite you to think about this as you consider how hacking has been treated by societies to date.
SMARTER FASTER trademarks owned by Freethink Media, Inc. All rights reserved.