NEUROSCIENCE, like many other sciences, has a bottomless appetite for data. Flashy enterprises such as the BRAIN Initiative, announced by Barack Obama in 2013, or the Human Brain Project, approved by the European Union in the same year, aim to analyse the way that thousands or even millions of nerve cells interact in a real brain. The hope is that the torrents of data these schemes generate will contain some crucial nuggets that let neuroscientists get closer to understanding how exactly the brain does what it does.
But a paper just published in PLOS Computational Biology questions whether more information is the same thing as more understanding. It does so by way of neuroscience’s favourite analogy: comparing the brain to a computer. Like brains, computers process information by shuffling electricity around complicated circuits. Unlike the workings of brains, though, those of computers are understood on every level.
Eric Jonas of the University of California, Berkeley, and Konrad Kording of Northwestern University, in Chicago, who both have backgrounds in neuroscience and electronic engineering, reasoned that a computer was therefore a good way to test the analytical toolkit used by modern neuroscience. Their idea was to see whether applying those techniques to a microprocessor produced information that matched what they already knew to be true about how the chip works.
So far, humans have relied on the written word to record what we know as history. When artificial intelligence researchers ran billions of those words from decades of news coverage through an automated analysis, however, even more patterns and insights were revealed.
A team from the University of Bristol ran 35 million articles from 100 local British newspapers spanning 150 years through both a simple content analysis and more sophisticated machine learning processes. By having machines “read” the nearly 30 billion words, the simple analysis allowed researchers to easily and accurately identify big events like wars and epidemics.
PLOS response to NIH RFI on Strategies for NIH Data Management, Sharing, and Citation.
This Request for Information (RFI) seeks public comments on data management and sharing strategies and priorities in order to consider: (1) how digital scientific data generated from NIH-funded research should be managed, and to the fullest extent possible, made publicly available; and, (2) how to set standards for citing shared data and software.
Lexalytics®, the leader in cloud and on-prem text analytics solutions, announced today that it is unveiling the Magic Machines™ AI Labs in Amherst, MA, to speed innovation in artificial intelligence (AI). In stealth mode for the past year, Magic Machines has been focusing on “force-multiplying” AI technologies. These breakthrough technologies researched by Magic Machines are already accelerating Lexalytics’ product development cycle. Over the coming months, Lexalytics will announce products that enable data scientists to handle more projects and empower business intelligence users to shape AI.
Lexalytics has also partnered with two leading academic institutions dedicated to the advancement of data science and marketing technology: University of Massachusetts Amherst’s Center for Data Science and Northwestern University’s Medill School of Journalism, Media and Integrated Marketing Communications. Through its affiliation with UMass, Lexalytics will work with faculty and students on the underlying challenges of analyzing, visualizing and drawing insights from massive volumes and varieties of data. At the other end of the problem, Northwestern and Lexalytics will partner to ensure the usability and applicability of Magic Machines AI technologies to a broad set of business users.
In case you missed the presentation from Brendan Frey from Deep Genomics, we are sharing with you the full recording of the video below!
Deep Genomics bring together machine learning and experimental biology. Their systems “predict the molecular effect of genetic variation, opening a new and exciting path to discovery for disease diagnostics and therapies.” Brenden talks about the recently developed gene editing systems that has made it possible to edit our genomes. His talk focused around why machine learning and AI will play a central role in this transformation of medicine and of humanity.
Big data continues to dominate the U.S. job market with data scientists again heading a list of the top 50 jobs in American during 2017.
The employment site Glassdoor reported that data scientists are again the most sought after candidates followed by DevOps engineers and data technicians. The annual rankings weigh factors such as earning potential, job satisfaction and the number of job openings. The survey released on Tuesday (Jan. 24) also highlights the continuing skills gap as companies scramble to find qualified data science candidates and university graduate programs in data science ramp up.
HHMI strongly encourages all HHMI laboratory heads to publish their original, peer-reviewed research in
journals that make publications freely available and downloadable on-line immediately after publication (i.e.
open access journals). If a laboratory head chooses to publish an original, peer-reviewed research publication
on which he or she is a major author in a journal that is not open access, the laboratory head is responsible
for ensuring that the publication is freely available and downloadable on-line as soon as reasonably possible
after publication, and in any event within twelve months of publication. “Major author” normally includes
both the first and last authors; if a middle author is designated in the paper as the corresponding author, then
that author is also considered to be a major author.
In addition, if an HHMI laboratory head’s original, peer-reviewed research publication is in a journal in the
biological or biomedical sciences, the publication must be available through PubMed Central as soon as
reasonably possible, and in any event within twelve months of publication. [pdf]
Cut to Pittsburgh, where four leading professional poker players are pitting their wits against an AI program created by two Carnegie Mellon university researchers. They’re playing a particular kind of high-stakes poker called heads-up no-limit Texas hold’em. The program is called Libratus, which is Latin for “balanced”. There is, however, nothing balanced about its performance. Just over halfway through the 20-day game, Libratus was $800,000 up on the humans, some of whom were beginning to feel depressed. “I didn’t realise how good it was until today,” said one of them, Dong Kim. “I felt like I was playing against someone who was cheating, like it could see my cards. I’m not accusing it of cheating. It was just that good.”
The match doesn’t end until tomorrow (you can follow it via the hashtag #brainsvsai) and so we don’t know yet if the computer will win. But even if it doesn’t, another milestone has been passed on the road to artificial intelligence, because the Pittsburgh contest has shown that a machine can play a pretty mean game of poker against real human experts. This is a big deal, because poker requires reasoning and intelligence that have up until now eluded machines
‘Exascale’ refers to high-performance computing systems capable of at least a billion billion calculations per second, or a factor of 50 times faster than the nation’s most powerful supercomputers in use today. Computational scientists aim to use these systems to generate new insights and accelerate discoveries in materials science, precision medicine, national security and numerous other fields.
As collaborators in four co-design centers created by the U.S. Department of Energy’s (DOE) Exascale Computing Project (ECP), researchers at the DOE’s Argonne National Laboratory are helping to solve some of these complex challenges and pave the way for the creation of exascale supercomputers.
The major barrier to meeting the world’s clean energy needs isn’t scientific or technical; instead it’s financial, said MIT Sloan Professor Andrew W. Lo.
Fusion, the type of nuclear energy that powers the sun, offers the promise of clean, safe, and plentiful carbon-free energy—but it has not yet proven to be practicable.
“What will it take to get to 50 million degrees and three atmospheres of pressure in seven seconds?” Lo asked, naming the technical requirements experts have outlined as necessary to fusion’s success. “I know one thing it’s going to take, and that’s financing.” Specifically, $10 billion.
While the term biohacking wasn’t tossed around until roughly the early 2000s, the inclusion of outsiders in knowledge production—and for that matter, the occurrence of scientific discovery outside of academia—is an old one. Rob Carlson wrote in Wired in 2005, “The era of garage biology is upon us,” but in 2009, Stanford University bioengineering professor Drew Endy suggested to the San Francisco Chronicle that, in fact, it may have been upon us a long time ago: “Darwin may have been the original do-it-yourself biologist, as he didn’t originally work for any institution,” he said.
It’s not often that a scientific discipline gains a 23-satellite constellation overnight. But today, space weather scientists are reaping such a windfall, as the Los Alamos National Laboratory in New Mexico has released 16 years of radiation measurements recorded by GPS satellites.
Although billions of people globally use data from GPS satellites, they remain U.S. military assets. Scientists have long sought the data generated by sensors used to monitor the status of the satellites, which operate in the heavy radiation of medium-Earth orbit and can be vulnerable to solar storms. But few have been allowed to tap this resource. “There’s a general hesitancy to broadcast even fairly innocuous things out to the broad community,” says Marc Kippen, a program manager at Los Alamos, which developed the radiation-measuring instruments.
A team of physicists has used data from GPS satellites to hunt for dark matter, the mysterious stuff whose gravity appears to hold galaxies together. They found no signs of a hypothetical type of dark matter, which consists of flaws in the fabric of space called topological defects, the researchers reported here on Saturday at a meeting of the American Physical Society. But the physicists say they have vastly narrowed the characteristics for how the defects—if they exist—would interact with ordinary matter. Their findings show how surprisingly innovative—and, in this case, cheap—methods might be used to test new ideas of what dark matter might be.
“It is so interesting and refreshing and exciting, and the cost is basically zero,” says Dmitry Budker, an experimental physicist at the Johannes Gutenberg University of Mainz in Germany, who was not involved in the work. “It’s basically the cost of the students analyzing the data.”
To ensure that NIPS remains an open forum for research, we ask the organisers to move the primary venue to a location outside of the US, where people from all nationalities can attend. We also ask NIPS to make provisions for those trapped in the US to participate as fully as possible (e.g. through live streams).
“Deep learning has been at the root of significant progress in many application areas, such as computer perception and natural language processing. But almost all of these systems currently use supervised learning with human-curated labels. The challenge of the next several years is to let machines learn from raw, unlabeled data, such as images, videos and text.” [video, 54:37]
We recently introduced our report on probabilistic programming. The accompanying prototype allows you to explore the past and future of the New York residential real estate market.
This post gives a feel for the content in our report by introducing the algorithms and technology that make probabilistic programming possible.