This is a repository for all cool scientific discussion and fascination. Scientific facts, theories, and overall cool scientific stuff that you'd like to share with others. Stuff that makes you smile and wonder at the amazing shit going on around us, that most people don't notice.
Post pictures, vidoes, stories, or links. Ask questions. Share science.
In a feat reminiscent of the controversial victory by supercomputer ‘Deep Blue’ over world chess champion Garry Kasparov, a computer program has managed to beat a string of professional poker players at the game.
DeepStack, as it was called, defeated 10 out of 11 players who took part in a total of 3,000 games as part of a scientific study into artificial intelligence.
The 11th player also lost, but by a margin that the researchers decided was not large enough to be statistically significant.
This is not the first time a computer has won at poker. Libratus, a program developed by Carnegie Mellon University academics, won $1.76m (£1.4m) from professionals in January, for example.
But the researchers said DeepStack's performance represented a "paradigm shift" in AI that could have implications for the defence industry and medicine.
One player who took on the algorithm, Irish professional Dara O’Kearney, said it felt like he had been playing a human who was “a bit better than me, but not massively better”.
He warned there was already “a lot of evidence” that bots were winning money from human players in online poker games.
Writing in the journal Science, the researchers, from Alberta University in Canada, said: “Artificial intelligence has seen several breakthroughs in recent years, with games often serving as milestones.
“A common feature of these games is that players have perfect information. Poker is the quintessential game of imperfect information, and a longstanding challenge problem in artificial intelligence.
“In a study involving 44,000 hands of poker, DeepStack defeated with statistical significance professional poker players in heads-up, no-limit Texas hold’em.”
This type of poker involves just two players, the computer and the human in this case.
The researchers said DeepStack had been able to win despite being given no training from expert human games.
“The implications go beyond being a milestone for artificial intelligence,” the Science paper said.
“DeepStack represents a paradigm shift in approximating solutions to large, sequential imperfect information games.
“With many real world problems involving information asymmetry, DeepStack also has implications for seeing powerful AI applied more in settings that do not fit the perfect information assumption.
“The abstraction paradigm for handling imperfect information has shown promise in applications like defending strategic resources and robust decision making as needed for medical treatment recommendations.
“DeepStack’s continual re-solving paradigm will hopefully open up many more possibilities.”
Dara O'Kearney, an Irish poker professional who completed 456 hands, told The Independent that DeepStack played in a style similar to one used by some human players, based on game theory.
“I would say there wasn’t a massive difference. If I hadn’t been told it was a computer, there was nothing it was doing that would have tipped me off that it was a computer,” he said.
“I felt I did pretty much okay, but … I did feel the computer was a bit better than me, but not massively better.
“Heads up, no limits poker is not my speciality. It’s possible a human who specialises in that might do better.” [Reply]
New light is shed on the famous paradox of Einstein, Podolsky and Rosen after 80 years. A group of researchers from the Faculty of Physics at the University of Warsaw has created a multidimensional entangled state of a single photon and a trillion of hot rubidium atoms. This hybrid entanglement has been stored in the laboratory for several microseconds. The research has been published in the prestigious Optica journal.
In their famous Physical Review article published in 1935, A. Einstein, B. Podolsky and N. Rosen have considered a decay of a particle into two products. In their thought-experiment, two products of decay were projected in exactly opposite directions, or more scientifically speaking their momenta were anti-correlated. It would not be a mystery within the framework of classical physics, however when applying the rules of the Quantum theory, the three researchers quickly arrived at a paradox. The Heisenberg uncertainty principle, dictating that position and momentum of a particle cannot be measured at the same time within arbitrary precision, lies at the center of this paradox. In Einstein's thought-experiment one can measure momentum of one particle and immediately know momentum of the other without measurement, as it is exactly opposite. Then, one only needs to measure position of this second particle and the Heisenberg uncertainty principle seems to be violated, which seriously baffled the three physicists.
Only today we know that this experiment is not, in fact, a paradox. The mistake of Einstein and co-workers was to use one-particle uncertainty principle to a system of two particles. If we treat these two particles as described by a single quantum state, we learn that the original uncertainty principle ceases to apply, especially if these particles are entangled.
In the Quantum Memories Laboratory at the University of Warsaw, the group of three physicists was first to create such an entangled state consisting of a macroscopic object - a group of about one trillion atoms, and a single photon - a particle of light. "Single photons, scattered during the interaction of a laser beam with atoms, are registered on a sensitive camera. A single registered photon carries information about the quantum state of the entire group of atoms. The atoms may be stored, and their state may be retrieved on demand." - says Michal Dabrowski, PhD student and co-author of the article.
The results of the experiment confirm that the atoms and the single photon are in a joint, entangled state. By measuring position and momentum of the photon, we gain all information about the state of atoms. To confirm this, polish scientists convert the atomic state into another photon, which again is measured using the state-of-the-art camera developed in the Quantum Memories Laboratory. "We demonstrate the Einstein-Podolsky-Rosen apparent paradox in a very similar version as originally proposed in 1935, however we extend the experiment by adding storage of light within the large group of atoms. Atoms store the photon in a form of a wave made of atomic spins, containing one trillion atoms. Such a state is very robust against loss of a single atoms, as information is spread across so many particles." - says Michal Parniak, PhD student taking part in the study.
The experiment performed by the group from the University of Warsaw is unique in one other way as well. The quantum memory storing the entangled state, created thanks to "PRELUDIUM" grant from the Poland's National Science Centre and "Diamentowy Grant" from the Polish Ministry of Science and Higher Education, allows for storage of up to 12 photons at once. This enhanced capacity is promising in terms of applications in quantum information processing. "The multidimensional entanglement is stored in our device for several microseconds, which is roughly a thousand times longer than in any previous experiments, and at the same time long enough to perform subtle quantum operations on the atomic state during storage" - explains Dr. Wojciech Wasilewski, group leader of the Quantum Memories Laboratory team.
The entanglement in the real and momentum space, described in the Optica article, can be used jointly with other well-known degrees of freedom such as polarization, allowing generation of so-called hyper-entanglement. Such elaborate ideas constitute new and original test of the fundamentals of quantum mechanics - a theory that is unceasingly mysterious yet brings immense technological progress. [Reply]
Artificial human life could soon be grown from scratch in the lab, after scientists successfully created a mammal embryo using only stem cells.
Cambridge University mixed two kinds of mouse stem cells and placed them on a 3D scaffold. After four days of growth in a tank of chemicals designed to mimic conditions inside the womb, the cells formed the structure of a living mouse embryo.
The breakthrough has been described as a ‘masterpiece’ in bioengineering, which could eventually allow scientists to grow artificial human embryos in the lab without the need for a sperm or an egg.
Growing embryos would help researchers to study the very early stages of human life so they could understand why so many pregnancies fail, but is likely to prove controversial and raise ethical questions about what constitutes human life.
Currently scientists can carry out experiments on leftover embryos from IVF treatments, but they are in short supply and must be destroyed after 14 days. Scientists say that being able to create unlimited numbers of artificial embryos in the lab could speed up research while potentially removing some of the ethical boundaries.
“We think that it will be possible to mimic a lot of the developmental events occurring before 14 days using human stem cells using a similar approach to our technique using mouse stem cells," said Professor Magdalena Zernicka-Goetz from the Department of Physiology, Development and Neuroscience at Cambridge, who led the research.
"We are very optimistic that this will allow us to study key events of this critical stage of human development without actually having to work on (IVF) embryos. Knowing how development normally occurs will allow us to understand why it so often goes wrong."
The embryos were created using genetically engineered stem cells coupled with extra-embryonic trophoblast stem cells (TSCs) which form the placenta in a normal pregnancy.
Previous attempts to grow embryos using only one kind of stem cell proved unsuccessful because the cells would not assemble into their correct positions.
But scientists discovered that when they added the second ‘placental’ stem cells, they two types began to talk to each other, effectively telling each other where to go.
Together they eventually melded together to form an embryonic structure, with two distinct clusters of cells at each end, and a cavity in the middle in which the embryo would continue to develop. The embryo would not grow into a mouse because it lacks the stem cells which make a yolk sack.
Britain is currently leading the world in fertility research, and last year a group at the Francis Crick Institute was granted permission to genetically modify human embryos, the first time in the world such a procedure had been approved by regulators.
However such work raises important ethical questions about the sanctity of human life and whether it should be manipulated or created in the lab at all.
Critics warn that allowing embryos to be grown for science opens the door to designer babies and genetically modified humans.
Dr David King, director of the watchdog group, Human Genetics Alert, said: "What concerns me about the possibility of artificial embryos is that this may become a route to creating GM or even cloned babies.
"Until there is an enforceable global ban on those possibilities, as we saw with mitochondrial transfer, this kind of research risks doing the scientific groundwork for entrepreneurs, who will use the technologies in countries with no regulation."
The scientists would need to seek permission from the Human Fertility and Embryology Authority (HFEA), before attempting to create human embryos using the technique, and experts called for 'international dialogue' before going ahead.
Prof James Adjaye, Chair of Stem Cell Research and Regenerative Medicine, Heinrich Heine University, in Germany, said: “A regulatory body will ultimately decide on whether human stem cell embryos can be generated and for how long they can be left in the petri dish to develop further.
“Of course, there should be an international dialogue on the regulation of such experiments.”
But the study was welcomed by the scientific community who said it was a significant breakthrough.
Dr Dusko Ilic, Reader in Stem Cell Science, King’s College London, said the research was ‘masterpiece’ in creating the earliest steps of life in a lab.
“This report is significant. The group from Cambridge is actually making the embryos de novo, using two different cell types, mixing them in a specific ratio and letting them to assemble together the embryo. This is science at its best.”
The research was published in the journal Science and was funded by the Wellcome Trust and the European Research Council. [Reply]
Originally Posted by Nickhead:
not the most glamorous of science, but what the fuck is going on here?
Turkeys are really dumb. I agree with the explanation below. They were likely just following each other single file, and ended up in a circle of stupid after the leader checked out the cat and then started following the turkey in front of it.
Originally Posted by :
Boston Magazine called "Massachusetts' foremost turkey expert," who said it was possible that the birds were trailing behind one animal who set out to investigate the corpse, and then wound up — follow-the-leader style — in a perfect circle.
Biologist Alan Krakauer, an expert on the behavioral ecology of birds, offers an explanation that ascribes a little more intention to the birds. He tells NPR it's "most likely ... predator inspection behavior."
He explains that turkeys, when faced with an animal that might eat them, sometimes follow the risky route of actually approaching the predator. It can help signal to other turkeys in the area that there's a threat, or help the turkeys evaluate how big of a danger the predator is.
"Or it could let the predator know the prey are aware of its presence, which might encourage the predator to move elsewhere to seek an easier meal," he says — "a lot like a group of small songbirds mobbing a hawk or owl they have discovered."
Originally Posted by Fish:
Turkeys are really dumb. I agree with the explanation below. They were likely just following each other single file, and ended up in a circle of stupid after the leader checked out the cat and then started following the turkey in front of it.
It's funny you mention that. They have an animal expert from the Miami Zoo on the Dan Le Batard Show every week and he takes callers. Someone called about this, and he talked about how intelligent turkeys were, and said this was likely predator inspection, similar to what antelopes/gazelle will do on the African plains. [Reply]
Originally Posted by 'Hamas' Jenkins:
It's funny you mention that. They have an animal expert from the Miami Zoo on the Dan Le Batard Show every week and he takes callers. Someone called about this, and he talked about how intelligent turkeys were, and said this was likely predator inspection, similar to what antelopes/gazelle will do on the African plains.
I don't have a dog in the fight, but turkeys are stupid. Grandpa always told me he had some turkeys when he was younger and they always had to lock them in the coup if it was going to rain because they'd stare up at the lightning and drown...
Mom and Dad had some turkeys move in at the farm after I went to college and they were pretty stupid. There was nothing coordinated about predator identification or evasion. It was pretty much "Panic and run" (Sorry. My daughter watches Lion Guard.)
I really enjoy the Ron McGill segment, and he may very well be correct, but in my experience, turkeys are epic dumb. [Reply]
Originally Posted by Fish: Turkeys are really dumb. I agree with the explanation below. They were likely just following each other single file, and ended up in a circle of stupid after the leader checked out the cat and then started following the turkey in front of it.
Originally Posted by 'Hamas' Jenkins: It's funny you mention that. They have an animal expert from the Miami Zoo on the Dan Le Batard Show every week and he takes callers. Someone called about this, and he talked about how intelligent turkeys were, and said this was likely predator inspection, similar to what antelopes/gazelle will do on the African plains.
Ya think?
Originally Posted by Later in Fish's post:
Biologist Alan Krakauer, an expert on the behavioral ecology of birds, offers an explanation that ascribes a little more intention to the birds. He tells NPR it's "most likely ... predator inspection behavior."
A microprocessor expanded and annotated visually for a museum so visitors can observe each and every discrete operation, and even slow it down or stop it for more detailed inspection.
A team in China has corrected genetic mutations in at least some of the cells in three normal human embryos using the CRISPR genome editing technique. The latest study is the first to describe the results of using CRISPR in viable human embryos, New Scientist can reveal.
While this study – which attempted to repair the DNA of six embryos in total – was very small, the results suggest CRISPR works much better in normal embryos than it did in previous tests on abnormal embryos that could not develop into children.
“It is encouraging,” says Robin Lovell-Badge of the Francis Crick Institute in London, who has contributed to several major reports on human genome editing. The numbers are far too low to make strong conclusions though, he cautions.
The CRISPR gene editing technique is a very efficient way of disabling genes, by introducing small mutations that disrupt the code of a DNA sequence. CRISPR can also be used to repair genes, but this is much more difficult.
Until now, results have only been published from experiments in which the CRISPR technique was used in abnormal embryos, made when two sperm fertilise the same egg. The idea behind this work was that it was more ethical to test the technique on embryos that could never fully develop.
In the first attempt to fix genes in human embryos, fewer than 1 in 10 cells were successfully repaired – an efficiency rate that is too low to make the method practical. A second study published in 2016 also had a low rate of efficiency. However, because these embryos were very genetically abnormal, these experiments may not have given an accurate indication of how well the technique would work in healthier embryos.
The Chinese team behind the latest study, at the Third Affiliated Hospital of Guangzhou Medical University, first carried out experiments with abnormal embryos, and found the repair rate was very low. But they had more success when they tried to repair mutations in normal embryos derived from immature eggs donated by people undergoing IVF.
Genetic disease
Immature eggs like these are usually discarded by IVF clinics, as the success rate is much lower than with mature eggs. However, children have been born from such immature eggs.
Jianqiao Liu and his team matured donated immature eggs, and fertilised each by injecting sperm from one of two men with a hereditary disease. They then injected the CRISPR machinery into these single-cell embryos before they started dividing.
The first sperm donor had a mutation called G1376T in the gene for the G6PD enzyme. This is a common cause of favism in China, a disorder in which eating certain foods such as fava beans can trigger the destruction of red blood cells.
In two of the resulting embryos, the G1376T mutation was corrected. But in one of the embryos, not all the cells were corrected. CRISPR turned off the G6PD gene in some of its cells rather than fixing it – making it what is known as a “mosaic”.
The second sperm donor had a mutation called beta41-42, which is one of the causes of the blood disease beta-thalassemia. Four of the resulting embryos carried the mutation. In one, CRISPR induced another mutation rather than fixing the beta41-42. In another, the mutation was successfully repaired in only some of the cells, creating another mosaic embryo. It did not work at all in the other two embryos.
In total, the mutation in one embryo was corrected in every cell, and two were corrected in some of the cells.
While firm conclusions cannot be drawn based on just six embryos, these results are encouraging as they suggest CRISPR gene repair is more efficient in normal cells. “It does look more promising than previous papers,” says Fredrik Lanner of the Karolinska Institute in Sweden, whose team has begun using CRISPR to disable genes in human embryos to study embryonic development.
Preventing mosaics
Several other groups have begun editing the genomes of normal human embryos or plan to start soon. There are rumours that another three or four studies on the use of CRISPR in human embryos have been completed but not yet published. It isn’t clear why this is the case, but the controversy surrounding the area may have made both researchers and journals wary.
The results so far, however, show the technology is far from the point where it could be safely used for editing embryos.
To make it safer to use gene editing to prevent children inheriting disease-causing mutations, researchers will need to find a way to prevent mosaicism. Edited embryos would always be tested before being implanted in a woman, but if they are mosaics such tests cannot guarantee the resulting child will be disease-free.
“This would need to be solved before the methods could be used clinically to correct a disease,” says Lovell-Badge. Progress is already being made: at least two teams have already found ways of reducing the risk of mosaicism in animals.
Mosaicism could also be avoided by editing the genomes of sperm and eggs prior to IVF, rather than embryos. This is expected to become possible in people in the next few years.
There are also a few diseases where mosaicism might not matter, Lovell-Badge points out, such as metabolic liver diseases where only 20 per cent function is enough to keep people healthy.
However, a major report on gene editing by the US National Academy of Sciences recently concluded that trials of germline gene editing should be allowed only if they meet a number of criteria – the first being “the absence of reasonable alternatives”.
Yet almost all inherited diseases can already be prevented by existing forms of screening, such as testing IVF embryos and implanting only disease-free ones. There are only a small number of cases where this method – called preimplantation genetic diagnosis – will not work because none of a couple’s embryos will be disease-free. [Reply]
Life expectancy will soon exceed 90 years for the first time, scientists have predicted, overturning all the assumptions about human longevity that prevailed at the beginning of the 20th century.
Women born in South Korea in 2030 are forecast to have a life expectancy of 90, a study has found. But other developed countries are not far behind, raising serious questions about the health and social care that will be needed by large numbers of the population living through their 80s.
The findings are from an international team of scientists funded by the UK Medical Research Council and the US Environmental Protection Agency, and come with caveats. It is impossible to accurately forecast the natural disasters, disease outbreaks or climate changes that may take a toll of lives around the world.
But the study in the Lancet medical journal shows a significant rise in life expectancy in most of the 35 developed countries studied. A notable exception is the US, where a combination of obesity, deaths of mothers and babies at birth, homicides and lack of equal access to healthcare is predicted to cause life expectancy to rise more slowly than in most comparable countries.
Among developed nations, South Korea is likely to see the largest increase in life expectancy, with women born in 2030 averaging 90.8 years, 6.6 years longer than those born in 2010
Boys born in 2030 in the US may expect to have similar lifespans to those in the Czech Republic, the study suggests, and girls will have life expectancy similar to those in Croatia and Mexico. Life expectancy for babies born in the US in 2030 is predicted to be 83.3 in 2030 for women and 79.5 for men, a small rise from the 2010 figures of 81.2 and 76.5 respectively.
The authors point out that the US is the only country in the Organisation for Economic Cooperation and Development without universal healthcare coverage. “Not only does the US have high and rising health inequalities, but also life expectancy has stagnated or even declined in some population subgroups,” write the authors.
