The Future of Artificial Intelligence and Cybernetics

Science fiction has, for many years, looked to a future in which robots are intelligent and cyborgs are commonplace. The Terminator, The Matrix, Blade Runner and I, Robot are all good examples of this vision.

But until the last decade, consideration of what this might actually mean in the future was unnecessary because it was all science fiction, not scientific reality. Now, however, science has not only done some catching up; it’s also introduced practicalities that the original story lines didn’t appear to include (and, in some cases, still don’t include).

What we consider here are several different experiments linking biology and technology together in a cybernetic way—essentially ultimately combining humans and machines in a relatively permanent merger.

When we typically first think of a robot, we regard it simply as a machine. We tend to think that it might be operated remotely by a human, or that it may be controlled by a simple computer program.

But what if the robot has a biological brain made up of brain cells, possibly even human neurons? Neurons grown under laboratory conditions on an array of non-invasive electrodes provide an attractive alternative with which to realize a new form of robot controller. In the near future, we will see thinking robots with brains not very dissimilar to those of humans.

That development will raise many social and ethical questions. For example, if the robot brain has roughly the same number of human neurons as a typical human brain, then could it, or should it, have rights similar to those of a person? Also, if such robots have far more human neurons than in a typical human brain—for example, a million times more neurons—would they, rather than humans, make all future decisions?

Many human brain–computer interfaces are used for therapeutic purposes to overcome medical or neurological problems, with one example being the deep brain stimulation (DBS) electrodes used to relieve the symptoms of Parkinson’s disease. However, even here it’s possible to consider using such technology in ways that would give people abilities that humans don’t normally possess—in other words, human enhancement. In some cases, those who have undergone amputations or suffered spinal injuries due to accidents may be able to regain control of devices via their still-functioning neural signals.

Meanwhile, stroke patients can be given limited control of their surroundings, as indeed can those who have motor neurone disease. With those cases, the situation isn’t straightforward, as patients receive abilities that normal humans don’t have—for example, the ability to move a cursor on a computer screen using nothing but neural signals.

It’s clear that connecting a human brain with a computer network via an implant could, in the long term, open up the distinct advantages of machine intelligence, communication, and sensing abilities to the individual receiving the implant. Currently, obtaining the go-ahead for each implantation requires ethical approval from the local authority governing the hospital where the procedure is performed. But looking ahead, it’s quite possible that commercial influences, coupled with societal wishes to communicate more effectively and perceive the world in a richer form, will drive market desire.

For some, brain–computer interfaces are perhaps a step too far just now—particularly if the approach means tampering directly with the brain. As a result, the most studied brain–computer interface to date is that involving electroencephalography (EEG). While EEG experimentation is relatively cheap, portable, and easy to set up, it’s still difficult to see its widespread future use. It certainly has a role to play in externally assessing some aspects of brain functioning for medical purposes. However, the idea of people driving around while wearing skullcap of electrodes, with no need for a steering wheel, doesn’t seem realistic. Completely autonomous vehicles are much more likely.

Such experimental cases indicate how humans—and animals, for that matter—can merge with technology. That, in turn, generates a plethora of social and ethical considerations as well as technical issues. That’s why it’s vital to include a sense of reflection so that the additional experimentation we’ll now witness will be guided by the informed feedback that results.

Toy-inspired experiment on behavior of quantum systems

With its suspended metallic spheres that clack back and forth, Newton's cradle is more than a popular desktop plaything. It has taught a generation of students about conservation of momentum and energy. It is also the inspiration for an experiment Benjamin Lev, associate professor of physics and of applied physics at Stanford University, has created to study quantum systems.

Lev and his group built their own quantum version of Newton's cradle in order to answer questions about how the chaotic motion of quantum particles eventually leads to thermal equilibrium in a process called thermalization. Answering how this occurs in quantum systems could help in developing quantum computers, sensors and devices, which Lev characterizes as a "quantum engineering revolution."

"If we want to be able to create devices that are robust and useful, we need to understand how quantum systems behave out of equilibrium -- when they're kicked, like the Newton's cradle -- at a level as fundamental as we understand that for classical systems," Lev said.

With the cradle, the researchers observed for the first time how, after inducing small amounts of chaotic motion, a quantum system reaches thermal equilibrium. They published their findings May 2 in Physical Review X.

The results of these experiments, which did not fit previous predictions, have led to a theory of how this process works in quantum systems.

Extremely cold, strongly magnetic

The turbulent swirl of milk as it's added to coffee is a familiar example of chaos in the non-quantum world. Over time, the coffee mixture becomes homogenous and, therefore, reaches equilibrium. What the Lev lab wanted to know is how this evolution occurs in quantum systems after they induce just a touch of chaos. Through experiments with their cradle, the researchers were the first to observe this process as it happened.

The Lev lab's quantum Newton's cradle is different from anything you've seen in your co-worker's cubicle. The researchers shine laser beams through an airtight chamber to cool a gas of atoms down to nearly absolute zero -- one of the coldest known gases in the universe -- and then they load those atoms into an array of laser tubes that act as the structure for the Newton's cradle. Each of the 700 parallel cradles contains around 50 atoms in a row. Then, another laser kicks the atoms, starting the movement of the cradle.

Unlike a previous quantum Newton's cradle developed by David Weiss at Penn State, where weakly magnetic atoms took the place of the cradle's metal spheres, the Lev lab's cradle includes strongly magnetic atoms.

This work builds on the lab's previous achievement of making the first quantum gas of the highly magnetic element dysprosium -- tied with terbium as the most magnetic of all elements. President Obama gave Lev a Presidential Early Career Award for Scientists and Engineers for this milestone in 2011. It was atoms of dysprosium the researchers loaded into the airtight chamber.

The researchers can tune how these atoms affect their neighbors. They can make the cradle act as though the atoms are not magnetic so that it will produce the periodic motion typical of Newton's cradle. Or they can produce chaotic motion by turning up the magnetism -- like a Newton's cradle with magnets strapped to the spheres.

Until now, physicists haven't had a theory of how thermalization arises in subtly chaotic quantum systems. Previous research with computational simulations has resulted in varying conclusions. Now, through their experiments, the researchers directly showed that the cradles' oscillation reached equilibrium in a sequence of two exponential steps, which was an unexpected result.

They also confirmed their experimental results in an extensive computer simulation. Based on these experiments and simulations, the group developed a theory that explains their findings.

"It means we can have a very general, simple theory for how complicated quantum systems like this one thermalize," Lev said. "That's beautiful because it allows you to translate that to other systems."

Atom by atom

Already, the researchers have several experiments planned for the magnetic quantum Newton's cradle and they anticipate many more opportunities for building upon this work as the quantum revolution evolves.

"Very sophisticated laser technologies can manipulate systems atom by atom," said Yijun Tang, a recently graduated doctoral student in the Lev lab and lead author of the paper. "So, maybe what we can do will go beyond fundamental science questions. Maybe, at some point, we can turn these technologies into something more practical as well."

In experiments to come, the researchers may add disorder to the cradle's tubes, in the form of speckled laser light, to see if they can create a sort of quantum glass that evades thermalization. The experiments that contributed to this paper were all done with one version of dysprosium isotopes, called bosons, so the group also plans to repeat its work with the alternative version, fermions. They aren't sure whether the change to fermions will make a difference to thermalization, Lev said, and they would welcome another surprise.

Additional Stanford co-authors are Wil Kao, Kuan-Yu Li and Sangwon Seo. Krishnanand Mallayya and Marcos Rigol of Penn State and Sarang Gopalakrishnan of the City University of New York are also co-authors. Lev is a faculty member in the School of Humanities and Sciences at Stanford.

This research was funded by the National Science Foundation and the Air Force Office of Scientific Research.

Engineers upgrade ancient, sun-powered tech to purify water with near-perfect efficiency

The idea of using energy from the sun to evaporate and purify water is ancient. The Greek philosopher Aristotle reportedly described such a process more than 2,000 years ago.

Now, researchers are bringing this technology into the modern age, using it to sanitize water at what they report to be record-breaking rates.

By draping black, carbon-dipped paper in a triangular shape and using it to both absorb and vaporize water, they have developed a method for using sunlight to generate clean water with near-perfect efficiency.

"Our technique is able to produce drinking water at a faster pace than is theoretically calculated under natural sunlight," says lead researcher Qiaoqiang Gan, PhD, associate professor of electrical engineering in the University at Buffalo School of Engineering and Applied Sciences.

As Gan explains, "Usually, when solar energy is used to evaporate water, some of the energy is wasted as heat is lost to the surrounding environment. This makes the process less than 100 percent efficient. Our system has a way of drawing heat in from the surrounding environment, allowing us to achieve near-perfect efficiency."

The low-cost technology could provide drinking water in regions where resources are scarce, or where natural disasters have struck. The advancements are described in a study published on May 3 in the journal Advanced Science.

The project, funded by the National Science Foundation (NSF), was a collaboration between UB, Fudan University in China and the University of Wisconsin-Madison. UB electrical engineering PhD graduate Haomin Song and PhD candidate Youhai Liu were the study's first authors.

Gan, Song and other colleagues have launched a startup, Sunny Clean Water, to bring the invention to people who need it. With support from the NSF Small Business Innovation Research program, the company is integrating the new evaporation system into a prototype of a solar still, a sun-powered water purifier.

"When you talk to government officials or nonprofits working in disaster zones, they want to know: 'How much water can you generate every day?' We have a strategy to boost daily performance," Song says. "With a solar still the size of a mini fridge, we estimate that we can generate 10 to 20 liters of clean water every single day."

Modernizing an age-old technology

Solar stills have been around for a long time. These devices use the sun's heat to evaporate water, leaving salt, bacteria and dirt behind. Then, the water vapor cools and returns to a liquid state, at which point it's collected in a clean container.

The technique has many advantages. It's simple, and the power source -- the sun -- is available just about everywhere. But unfortunately, even the latest solar still models are somewhat inefficient at vaporizing water.

Gan's team addressed this challenge through a neat, counterintuitive trick: They increased the efficiency of their evaporation system by cooling it down.

A central component of their technology is a sheet of carbon-dipped paper that is folded into an upside-down "V" shape, like the roof of a birdhouse. The bottom edges of the paper hang in a pool of water, soaking up the fluid like a napkin. At the same time, the carbon coating absorbs solar energy and transforms it into heat for evaporation.

As Gan explains, the paper's sloped geometry keeps it cool by weakening the intensity of the sunlight illuminating it. (A flat surface would be hit directly by the sun's rays.) Because most of the carbon-coated paper stays under room temperature, it can draw in heat from the surrounding area, compensating for the regular loss of solar energy that occurs during the vaporization process.

Using this set-up, researchers evaporated the equivalent of 2.2 liters of water per hour for every square meter of area illuminated by the regular sun, higher than the theoretical upper limit of 1.68 liters, according to the new study. The team conducted its tests in the lab, using a solar simulator to generate light at the intensity of one regular sun.

"Most groups working on solar evaporation technologies are trying to develop advanced materials, such as metallic plasmonic and carbon-based nanomaterials," Gan says. "We focused on using extremely low-cost materials and were still able to realize record-breaking performance.

"Importantly, this is the only example I know of where the thermal efficiency of the solar evaporation process is 100 percent when you consider solar energy input. By developing a technique where the vapor is below ambient temperature, we create new research possibilities for exploring alternatives to high-temperature steam generation."

Should ethics or human intuition drive the moral judgments of driverless cars?

When faced with driving dilemmas, people show a high willingness to sacrifice themselves for others, make decisions based on the victim's age and swerve onto sidewalks to minimize the number of lives lost, reveals new research published in open-access journal Frontiers in Behavioral Neuroscience. This is at odds with ethical guidelines in these circumstances, which often dictate that no life should be valued over another. This research hopes to initiate discussions about the way self-driving vehicles should be programmed to deal with situations that endanger human life.

"The technological advancement and adoption of autonomous vehicles is moving quickly but the social and ethical discussions about their behavior is lagging behind," says lead author Lasse T. Bergmann, who completed this study with a team at the Institute of Cognitive Science, University of Osnabrück, Germany. "The behavior that will be considered as right in such situations depends on which factors are considered to be both morally relevant and socially acceptable."

Traffic accidents are a major source of death and injury in the world. As technology improves, automated vehicles will outperform their human counterparts, saving lives by eliminating accidents caused by human error. Despite this, there will still be circumstances where self-driving vehicles will need to make decisions in a morally challenging situation. For example, a car can swerve to avoid hitting a child that has run into the road but in doing so endangers other lives. How should it be programmed to behave?

An ethics commission initiated by the German Ministry for Transportation has created a set of guidelines, representing its members' best judgement on a variety of issues concerning self-driving cars. These expert judgments may, however, not reflect human intuition.

Bergmann and colleagues designed a virtual reality experiment to examine human intuition in a variety of possible driving scenarios. Different sets of tests were created to highlight different factors that may or may not be perceived as morally relevant.

Based on a traditional ethical thought experiment, the trolley problem, test subjects could choose between two lanes on which their vehicle drove at constant speed. They were presented with a morally challenging driving dilemma, such as an option to move lanes to minimize lives lost, a choice between victims of different age, or a possibility for self-sacrifice to save others.

It revealed that human intuition was often at odds with ethical guidelines.

Bergmann explains, "The German ethics commission proposes that a passenger in the vehicle may not be sacrificed to save more people; an intuition not generally shared by subjects in our experiment. We also find that people chose to save more lives, even if this involves swerving onto the sidewalk -- endangering people uninvolved in the traffic incident. Furthermore, subjects considered the factor of age, for example, choosing to save children over the elderly."

He continues, "If autonomous vehicles abide with guidelines dictated by the ethics commission, our experimental evidence suggests that people would not be happy with the decisions their cars make for them."

Professor Gordon Pipa, co-author, also based at the University of Osnabrück continues, "It is urgent that we start engaging into a societal discussion to define the goals and constraints of future rules that apply to self-drive vehicles. This needs to happen before they become an integral part of our daily lives."

Bergmann explains that further research is needed. "While 'dilemma' situations deserve more study, other questions should also be discussed. Driving requires an intricate weighing of risks versus rewards, for example speed versus the danger of a critical situation unfolding. Decision making-processes that precede or avoid a critical situation should also be investigated."

Taming the multiverse: Stephen Hawking's final theory about the big bang

Professor Stephen Hawking's final theory on the origin of the universe, which he worked on in collaboration with Professor Thomas Hertog from KU Leuven, has been published today in the Journal of High Energy Physics.

The theory, which was submitted for publication before Hawking's death earlier this year, is based on string theory and predicts the universe is finite and far simpler than many current theories about the big bang say.

Professor Hertog, whose work has been supported by the European Research Council, first announced the new theory at a conference at the University of Cambridge in July of last year, organised on the occasion of Professor Hawking's 75th birthday.

Modern theories of the big bang predict that our local universe came into existence with a brief burst of inflation -- in other words, a tiny fraction of a second after the big bang itself, the universe expanded at an exponential rate. It is widely believed, however, that once inflation starts, there are regions where it never stops. It is thought that quantum effects can keep inflation going forever in some regions of the universe so that globally, inflation is eternal. The observable part of our universe would then be just a hospitable pocket universe, a region in which inflation has ended and stars and galaxies formed.

"The usual theory of eternal inflation predicts that globally our universe is like an infinite fractal, with a mosaic of different pocket universes, separated by an inflating ocean," said Hawking in an interview last autumn. "The local laws of physics and chemistry can differ from one pocket universe to another, which together would form a multiverse. But I have never been a fan of the multiverse. If the scale of different universes in the multiverse is large or infinite the theory can't be tested. "

In their new paper, Hawking and Hertog say this account of eternal inflation as a theory of the big bang is wrong. "The problem with the usual account of eternal inflation is that it assumes an existing background universe that evolves according to Einstein's theory of general relativity and treats the quantum effects as small fluctuations around this," said Hertog. "However, the dynamics of eternal inflation wipes out the separation between classical and quantum physics. As a consequence, Einstein's theory breaks down in eternal inflation."

"We predict that our universe, on the largest scales, is reasonably smooth and globally finite. So it is not a fractal structure," said Hawking.

The theory of eternal inflation that Hawking and Hertog put forward is based on string theory: a branch of theoretical physics that attempts to reconcile gravity and general relativity with quantum physics, in part by describing the fundamental constituents of the universe as tiny vibrating strings. Their approach uses the string theory concept of holography, which postulates that the universe is a large and complex hologram: physical reality in certain 3D spaces can be mathematically reduced to 2D projections on a surface.

Hawking and Hertog developed a variation of this concept of holography to project out the time dimension in eternal inflation. This enabled them to describe eternal inflation without having to rely on Einstein' theory. In the new theory, eternal inflation is reduced to a timeless state defined on a spatial surface at the beginning of time.

"When we trace the evolution of our universe backwards in time, at some point we arrive at the threshold of eternal inflation, where our familiar notion of time ceases to have any meaning," said Hertog.

Hawking's earlier 'no boundary theory' predicted that if you go back in time to the beginning of the universe, the universe shrinks and closes off like a sphere, but this new theory represents a step away from the earlier work. "Now we're saying that there is a boundary in our past," said Hertog.

Hertog and Hawking used their new theory to derive more reliable predictions about the global structure of the universe. They predicted the universe that emerges from eternal inflation on the past boundary is finite and far simpler than the infinite fractal structure predicted by the old theory of eternal inflation.

Their results, if confirmed by further work, would have far-reaching implications for the multiverse paradigm. "We are not down to a single, unique universe, but our findings imply a significant reduction of the multiverse, to a much smaller range of possible universes," said Hawking.

This makes the theory more predictive and testable.

Hertog now plans to study the implications of the new theory on smaller scales that are within reach of our space telescopes. He believes that primordial gravitational waves -- ripples in spacetime -- generated at the exit from eternal inflation constitute the most promising "smoking gun" to test the model. The expansion of our universe since the beginning means such gravitational waves would have very long wavelengths, outside the range of the current LIGO detectors. But they might be heard by the planned European space-based gravitational wave observatory, LISA, or seen in future experiments measuring the cosmic microwave background.

Atoms may hum a tune from grand cosmic symphony

Researchers playing with a cloud of ultracold atoms uncovered behavior that bears a striking resemblance to the universe in microcosm. Their work, which forges new connections between atomic physics and the sudden expansion of the early universe, will be published in Physical Review X and highlighted by Physics.

"From the atomic physics perspective, the experiment is beautifully described by existing theory," says Stephen Eckel, an atomic physicist at the National Institute of Standards and Technology (NIST) and the lead author of the new paper. "But even more striking is how that theory connects with cosmology."

In several sets of experiments, Eckel and his colleagues rapidly expanded the size of a doughnut-shaped cloud of atoms, taking snapshots during the process. The growth happens so fast that the cloud is left humming, and a related hum may have appeared on cosmic scales during the rapid expansion of the early universe -- an epoch that cosmologists refer to as the period of inflation.

The work brought together experts in atomic physics and gravity, and the authors say it is a testament to the versatility of the Bose-Einstein condensate (BEC) -- an ultracold cloud of atoms that can be described as a single quantum object -- as a platform for testing ideas from other areas of physics.

"Maybe this will one day inform future models of cosmology," Eckel says. "Or vice versa. Maybe there will be a model of cosmology that's difficult to solve but that you could simulate using a cold atomic gas."

It's not the first time that researchers have connected BECs and cosmology. Prior studies mimicked black holes and searched for analogs of the radiation predicted to pour forth from their shadowy boundaries. The new experiments focus instead on the BEC's response to a rapid expansion, a process that suggests several analogies to what may have happened during the period of inflation.

The first and most direct analogy involves the way that waves travel through an expanding medium. Such a situation doesn't arise often in physics, but it happened during inflation on a grand scale. During that expansion, space itself stretched any waves to much larger sizes and stole energy from them through a process known as Hubble friction.

In one set of experiments, researchers spotted analogous features in their cloud of atoms. They imprinted a sound wave onto their cloud -- alternating regions of more atoms and fewer atoms around the ring, like a wave in the early universe -- and watched it disperse during expansion. Unsurprisingly, the sound wave stretched out, but its amplitude also decreased. The math revealed that this damping looked just like Hubble friction, and the behavior was captured well by calculations and numerical simulations.

"It's like we're hitting the BEC with a hammer," says Gretchen Campbell, the NIST co-director of the Joint Quantum Institute (JQI) and a coauthor of the paper, "and it's sort of shocking to me that these simulations so nicely replicate what's going on."

In a second set of experiments, the team uncovered another, more speculative analogy. For these tests they left the BEC free of any sound waves but provoked the same expansion, watching the BEC slosh back and forth until it relaxed.

In a way, that relaxation also resembled inflation. Some of the energy that drove the expansion of the universe ultimately ended up creating all of the matter and light around us. And although there are many theories for how this happened, cosmologists aren't exactly sure how that leftover energy got converted into all the stuff we see today.

In the BEC, the energy of the expansion was quickly transferred to things like sound waves traveling around the ring. Some early guesses for why this was happening looked promising, but they fell short of predicting the energy transfer accurately. So the team turned to numerical simulations that could capture a more complete picture of the physics.

What emerged was a complicated account of the energy conversion: After the expansion stopped, atoms at the outer edge of the ring hit their new, expanded boundary and got reflected back toward the center of the cloud. There, they interfered with atoms still traveling outward, creating a zone in the middle where almost no atoms could live. Atoms on either side of this inhospitable area had mismatched quantum properties, like two neighboring clocks that are out of sync.

The situation was highly unstable and eventually collapsed, leading to the creation of vortices throughout the cloud. These vortices, or little quantum whirlpools, would break apart and generate sound waves that ran around the ring, like the particles and radiation left over after inflation. Some vortices even escaped from the edge of the BEC, creating an imbalance that left the cloud rotating.

Unlike the analogy to Hubble friction, the complicated story of how sloshing atoms can create dozens of quantum whirlpools may bear no resemblance to what goes on during and after inflation. But Ted Jacobson, a coauthor of the new paper and a physics professor at the University of Maryland specializing in black holes, says that his interaction with atomic physicists yielded benefits outside these technical results.

"What I learned from them, and from thinking so much about an experiment like that, are new ways to think about the cosmology problem," Jacobson says. "And they learned to think about aspects of the BEC that they would never have thought about before. Whether those are useful or important remains to be seen, but it was certainly stimulating."

Eckel echoes the same thought. "Ted got me to think about the processes in BECs differently," he says, "and any time you approach a problem and you can see it from a different perspective, it gives you a better chance of actually solving that problem."

Future experiments may study the complicated transfer of energy during expansion more closely, or even search for further cosmological analogies. "The nice thing is that from these results, we now know how to design experiments in the future to target the different effects that we hope to see," Campbell says. "And as theorists come up with models, it does give us a testbed where we could actually study those models and see what happens."

The new paper included contributions from two coauthors not mentioned in the text: Avinash Kumar, a graduate student at JQI; and Ian Spielman, a JQI Fellow and NIST physicist.

An expanding cloud of atoms could offer insight into unanswered cosmological questions