Archive for the ‘Oxford University’ Category

Artificial intelligence poses an “extinction risk” to human civilisation, an Oxford University professor has said.

Almost everything about the development of genuine AI is uncertain, Stuart Armstrong at the Future of Humanity Institute said in an interview with The Next Web.

That includes when we might develop it, how such a thing could come about and what it means for human society.

But without more research and careful study, it’s possible that we could be opening a Pandora’s box. Which is exactly the sort of thing that the Future of Humanity Institute, a multidisciplinary research hub tasked with asking the “big questions” about the future, is concerned with.

“One of the things that makes AI risk scary is that it’s one of the few that is genuinely an extinction risk if it were to go bad. With a lot of other risks, it’s actually surprisingly hard to get to an extinction risk,” Armstrong told The Next Web.

“If AI went bad, and 95% of humans were killed then the remaining 5% would be extinguished soon after. So despite its uncertainty, it has certain features of very bad risks.”

The thing for humanity to fear is not quite the robots of Terminator (“basically just armoured bears”) but a more incorporeal intelligence capable of dominating humanity from within.

The threat of such a powerful computer brain would include near-term (and near total) unemployment, as replacements for virtually all human workers are quickly developed and replicated, but extends beyond that to genuine threats of widespread anti-human violence.

“Take an anti-virus program that’s dedicated to filtering out viruses from incoming emails and wants to achieve the highest success, and is cunning and you make that super-intelligent,” Armstong said.

“Well it will realise that, say, killing everybody is a solution to its problems, because if it kills everyone and shuts down every computer, no more emails will be sent and and as a side effect no viruses will be sent.”

The caveat to all this is that creating AI is difficult, and we’re nowhere near it. The caveat to that is that it could happen far more quickly than anyone would expect, if just one developer came up with a “neat algorithm” that no one else had thought to construct.

Armstrong’s conclusion is simple: let’s think about this now, particularly in relation to employment, and try to adjust society ourselves before the AI adjusts it for us.

http://www.huffingtonpost.co.uk/2014/03/12/extinction-artificial-intelligence-oxford-stuart-armstrong_n_4947082.html

MATH

In a lab in Oxford University’s experimental psychology department, researcher Roi Cohen Kadosh is testing an intriguing treatment: He is sending low-dose electric current through the brains of adults and children as young as 8 to make them better at math.

A relatively new brain-stimulation technique called transcranial electrical stimulation may help people learn and improve their understanding of math concepts.

The electrodes are placed in a tightly fitted cap and worn around the head. The device, run off a 9-volt battery commonly used in smoke detectors, induces only a gentle current and can be targeted to specific areas of the brain or applied generally. The mild current reduces the risk of side effects, which has opened up possibilities about using it, even in individuals without a disorder, as a general cognitive enhancer. Scientists also are investigating its use to treat mood disorders and other conditions.

Dr. Cohen Kadosh’s pioneering work on learning enhancement and brain stimulation is one example of the long journey faced by scientists studying brain-stimulation and cognitive-stimulation techniques. Like other researchers in the community, he has dealt with public concerns about safety and side effects, plus skepticism from other scientists about whether these findings would hold in the wider population.

There are also ethical questions about the technique. If it truly works to enhance cognitive performance, should it be accessible to anyone who can afford to buy the device—which already is available for sale in the U.S.? Should parents be able to perform such stimulation on their kids without monitoring?

“It’s early days but that hasn’t stopped some companies from selling the device and marketing it as a learning tool,” Dr. Cohen Kadosh says. “Be very careful.”

The idea of using electric current to treat the brain of various diseases has a long and fraught history, perhaps most notably with what was called electroshock therapy, developed in 1938 to treat severe mental illness and often portrayed as a medieval treatment that rendered people zombielike in movies such as “One Flew over the Cuckoo’s Nest.”

Electroconvulsive therapy has improved dramatically over the years and is considered appropriate for use against types of major depression that don’t respond to other treatments, as well as other related, severe mood states.

A number of new brain-stimulation techniques have been developed, including deep brain stimulation, which acts like a pacemaker for the brain. With DBS, electrodes are implanted into the brain and, though a battery pack in the chest, stimulate neurons continuously. DBS devices have been approved by U.S. regulators to treat tremors in Parkinson’s disease and continue to be studied as possible treatments for chronic pain and obsessive-compulsive disorder.

Transcranial electrical stimulation, or tES, is one of the newest brain stimulation techniques. Unlike DBS, it is noninvasive.

If the technique continues to show promise, “this type of method may have a chance to be the new drug of the 21st century,” says Dr. Cohen Kadosh.

The 37-year-old father of two completed graduate school at Ben-Gurion University in Israel before coming to London to do postdoctoral work with Vincent Walsh at University College London. Now, sitting in a small, tidy office with a model brain on a shelf, the senior research fellow at Oxford speaks with cautious enthusiasm about brain stimulation and its potential to help children with math difficulties.

Up to 6% of the population is estimated to have a math-learning disability called developmental dyscalculia, similar to dyslexia but with numerals instead of letters. Many more people say they find math difficult. People with developmental dyscalculia also may have trouble with daily tasks, such as remembering phone numbers and understanding bills.

Whether transcranial electrical stimulation proves to be a useful cognitive enhancer remains to be seen. Dr. Cohen Kadosh first thought about the possibility as a university student in Israel, where he conducted an experiment using transcranial magnetic stimulation, a tool that employs magnetic coils to induce a more powerful electrical current.

He found that he could temporarily turn off regions of the brain known to be important for cognitive skills. When the parietal lobe of the brain was stimulated using that technique, he found that the basic arithmetic skills of doctoral students who were normally very good with numbers were reduced to a level similar to those with developmental dyscalculia.

That led to his next inquiry: If current could turn off regions of the brain making people temporarily math-challenged, could a different type of stimulation improve math performance? Cognitive training helps to some extent in some individuals with math difficulties. Dr. Cohen Kadosh wondered if such learning could be improved if the brain was stimulated at the same time.

But transcranial magnetic stimulation wasn’t the right tool because the current induced was too strong. Dr. Cohen Kadosh puzzled over what type of stimulation would be appropriate until a colleague who had worked with researchers in Germany returned and told him about tES, at the time a new technique. Dr. Cohen Kadosh decided tES was the way to go.

His group has since conducted a series of studies suggesting that tES appears helpful improving learning speed on various math tasks in adults who don’t have trouble in math. Now they’ve found preliminary evidence for those who struggle in math, too.

Participants typically come for 30-minute stimulation-and-training sessions daily for a week. His team is now starting to study children between 8 and 10 who receive twice-weekly training and stimulation for a month. Studies of tES, including the ones conducted by Dr. Cohen Kadosh, tend to have small sample sizes of up to several dozen participants; replication of the findings by other researchers is important.

In a small, toasty room, participants, often Oxford students, sit in front of a computer screen and complete hundreds of trials in which they learn to associate numerical values with abstract, nonnumerical symbols, figuring out which symbols are “greater” than others, in the way that people learn to know that three is greater than two.

When neurons fire, they transfer information, which could facilitate learning. The tES technique appears to work by lowering the threshold neurons need to reach before they fire, studies have shown. In addition, the stimulation appears to cause changes in neurochemicals involved in learning and memory.

However, the results so far in the field appear to differ significantly by individual. Stimulating the wrong brain region or at too high or long a current has been known to show an inhibiting effect on learning. The young and elderly, for instance, respond exactly the opposite way to the same current in the same location, Dr. Cohen Kadosh says.

He and a colleague published a paper in January in the journal Frontiers in Human Neuroscience, in which they found that one individual with developmental dyscalculia improved her performance significantly while the other study subject didn’t.

What is clear is that anyone trying the treatment would need to train as well as to stimulate the brain. Otherwise “it’s like taking steroids but sitting on a couch,” says Dr. Cohen Kadosh.

Dr. Cohen Kadosh and Beatrix Krause, a graduate student in the lab, have been examining individual differences in response. Whether a room is dark or well-lighted, if a person smokes and even where women are in their menstrual cycle can affect the brain’s response to electrical stimulation, studies have found.

Results from his lab and others have shown that even if stimulation is stopped, those who benefited are going to maintain a higher performance level than those who weren’t stimulated, up to a year afterward. If there isn’t any follow-up training, everyone’s performance declines over time, but the stimulated group still performs better than the non-stimulated group. It remains to be seen whether reintroducing stimulation would then improve learning again, Dr. Cohen Kadosh says.

http://online.wsj.com/news/articles/SB10001424052702303650204579374951187246122?mod=WSJ_article_EditorsPicks&mg=reno64-wsj&url=http%3A%2F%2Fonline.wsj.com%2Farticle%2FSB10001424052702303650204579374951187246122.html%3Fmod%3DWSJ_article_EditorsPicks

universe

Is our universe merely one of billions? Evidence of the existence of ‘multiverse’ revealed for the first time by a cosmic map of background radiation data gathered by Planck telescope. The first ‘hard evidence’ that other universes exist has been claimed to have been found by cosmologists studying new Planck data released this past June. They have concluded that it shows anomalies that can only have been caused by the gravitational pull of other universes.

“Such ideas may sound wacky now, just like the Big Bang theory did three generations ago,” says George Efstathiou, professor of astrophysics at Cambridge University.”But then we got evidence and now it has changed the whole way we think about the universe.”

Scientists had predicted that it should be evenly distributed, but the map shows a stronger concentration in the south half of the sky and a ‘cold spot’ that cannot be explained by current understanding of physics. Laura Mersini-Houghton, theoretical physicist at the University of North Carolina at Chapel Hill, and Richard Holman, professor at Carnegie Mellon University, predicted that anomalies in radiation existed and were caused by the pull from other universes in 2005. Mersini-Houghton will be in Britain soon promoting this theory and, we expect, the hard evidence at the Hay Festival on May 31 and at Oxford on June 11.

Dr Mersini-Houghton believes her hypothesis has been proven from the Planck data that data has been used to create a map of light from when the universe was just 380,000 years old. “These anomalies were caused by other universes pulling on our universe as it formed during the Big Bang,” she says. “They are the first hard evidence for the existence of other universes that we have seen.”

Columbia University mathematician Peter Woit writes in his blog, Not Even Wrong, that in recent years there have been many claims made for “evidence” of a multiverse, supposedly found in the CMB data. “Such claims often came with the remark that the Planck CMB data would convincingly decide the matter. When the Planck data was released two months ago, I looked through the press coverage and through the Planck papers for any sign of news about what the new data said about these multiverse evidence claims. There was very little there; possibly the Planck scientists found these claims to be so outlandish that it wasn’t worth the time to look into what the new data had to say about them.

“One exception,” Woit adds, “was this paper, where Planck looked for evidence of ‘dark flow’. They found nothing, and a New Scientist article summarized the situation: ‘The Planck team’s paper appears to rule out the claims of Kashlinsky and collaborators,’ says David Spergel of Princeton University, who was not involved in the work. If there is no dark flow, there is no need for exotic explanations for it, such as other universes, says Planck team member Elena Pierpaoli at the University of Southern California, Los Angeles. “You don’t have to think of alternatives.'”

“Dark Flow” sounds like a new SciFi Channel series. It’s not! The dark flow is controversial because the distribution of matter in the observed universe cannot account for it. Its existence suggests that some structure beyond the visible universe — outside our “horizon” — is pulling on matter in our vicinity.

Back in the Middle Ages, maps showed terrifying images of sea dragons at the boundaries of the known world. Today, scientists have observed strange new motion at the very limits of the known universe – kind of where you’d expect to find new things, but they still didn’t expect this. A huge swath of galactic clusters seem to be heading to a cosmic hotspot and nobody knows why.

Cosmologists regard the microwave background — a flash of light emitted 380,000 years after the universe formed — as the ultimate cosmic reference frame. Relative to it, all large-scale motion should show no preferred direction. A 2010 study tracked the mysterious cosmic ‘dark flow’ to twice the distance originally reported. The study was led by Alexander Kashlinsky at NASA’s Goddard Space Flight Center in Greenbelt, Maryland.

“This is not something we set out to find, but we cannot make it go away,” Kashlinsky said. “Now we see that it persists to much greater distances – as far as 2.5 billion light-years away,” he added.

Dark flow describes a possible non-random component of the peculiar velocity of galaxy clusters. The actual measured velocity is the sum of the velocity predicted by Hubble’s Law plus a small and unexplained (or dark) velocity flowing in a common direction. According to standard cosmological models, the motion of galaxy clusters with respect to the cosmic microwave background should be randomly distributed in all directions. However, analyzing the three-year WMAP data using the kinematic Sunyaev-Zel’dovich effect, the authors of the study found evidence of a “surprisingly coherent” 600–1000 km/s flow of clusters toward a 20-degree patch of sky between the constellations of Centaurus and Vela.

The clusters appear to be moving along a line extending from our solar system toward Centaurus/Hydra, but the direction of this motion is less certain. Evidence indicates that the clusters are headed outward along this path, away from Earth, but the team cannot yet rule out the opposite flow.

“We detect motion along this axis, but right now our data cannot state as strongly as we’d like whether the clusters are coming or going,” Kashlinsky said.

The unexplained motion has hundreds of millions of stars dashing towards a certain part of the sky at over eight hundred kilometers per second. Not much speed in cosmic terms, but the preferred direction certainly is: most cosmological models have things moving in all directions equally at the extreme edges of the universe. Something that could make things aim for a specific spot on such a massive scale hasn’t been imagined before. The scientists are keeping to the proven astrophysical strategy of calling anything they don’t understand “dark”, terming the odd motion a “dark flow”.

A black hole can’t explain the observations – objects would accelerate into the hole, while the NASA scientists see constant motion over a vast expanse of a billion light-years. You have no idea how big that is. This is giant on a scale where it’s not just that we can’t see what’s doing it; it’s that the entire makeup of the universe as we understand it can’t be right if this is happening.

The hot X-ray-emitting gas within a galaxy cluster scatters photons from the cosmic microwave background (CMB). Because galaxy clusters don’t precisely follow the expansion of space, the wavelengths of scattered photons change in a way that reflects each cluster’s individual motion.

This results in a minute shift of the microwave background’s temperature in the cluster’s direction. The change, which astronomers call the kinematic Sunyaev-Zel’dovich (KSZ) effect, is so small that it has never been observed in a single galaxy cluster.

But in 2000, Kashlinsky, working with Fernando Atrio-Barandela at the University of Salamanca, Spain, demonstrated that it was possible to tease the subtle signal out of the measurement noise by studying large numbers of clusters.

In 2008, armed with a catalog of 700 clusters assembled by Harald Ebeling at the University of Hawaii and Dale Kocevski, now at the University of California, Santa Cruz, the researchers applied the technique to the three-year WMAP data release. That’s when the mystery motion first came to light.

The new study builds on the previous one by using the five-year results from WMAP and by doubling the number of galaxy clusters.

“It takes, on average, about an hour of telescope time to measure the distance to each cluster we work with, not to mention the years required to find these systems in the first place,” Ebeling said. “This is a project requiring considerable followthrough.”

According to Atrio-Barandela, who has focused on understanding the possible errors in the team’s analysis, the new study provides much stronger evidence that the dark flow is real. For example, the brightest clusters at X-ray wavelengths hold the greatest amount of hot gas to distort CMB photons. “When processed, these same clusters also display the strongest KSZ signature — unlikely if the dark flow were merely a statistical fluke,” he said.

In addition, the team, which now also includes Alastair Edge at the University of Durham, England, sorted the cluster catalog into four “slices” representing different distance ranges. They then examined the preferred flow direction for the clusters within each slice. While the size and exact position of this direction display some variation, the overall trends among the slices exhibit remarkable agreement.

The researchers are currently working to expand their cluster catalog in order to track the dark flow to about twice the current distance. Improved modeling of hot gas within the galaxy clusters will help refine the speed, axis, and direction of motion.

Future plans call for testing the findings against newer data released from the WMAP project and the European Space Agency’s Planck mission, which is also currently mapping the microwave background.

Which is fantastic! Such discoveries force a whole new set of ideas onto the table which, even if they turn out to be wrong, are the greatest ways to advance science and our understanding of everything. One explanation that’s already been offered is that our universe underwent a period of hyper-inflation early in its existence, and everything we think of as the vast and infinite universe is actually a small corner under the sofa of the real expanse of reality. Which would be an amazing, if humbling, discovery.

The image at the top of the page shows the most distant object we have ever observed with high confidence, according to Wei Zheng, the leading astronomer of the team at Johns Hopkins University who that noticed the galaxy on multiple images from both the Hubble and Spitzer space telescopes. At 13.2-billion years old, we are technically seeing this galaxy when it was very young, but its light is only reaching Earth now.

http://www.dailygalaxy.com/my_weblog/2013/10/is-our-universe-one-of-billions-new-planck-data-has-anomalies-caused-by-unknown-gravitational-pull-t.html