An anonymous reader Phys.Org: The Korea Institute of Science and Technology (KIST) announced that the collaborative research team led by Dr. Kwan Hyi Lee from the Biomaterials Research Center and Professor In Gab Jeong from Asan Medical Center developed a technique for diagnosing prostate cancer from urine within only 20 minutes with almost 100% accuracy. The research team developed this technique by introducing a smart AI analysis method to an electrical-signal-based ultrasensitive biosensor. As a noninvasive method, a diagnostic test using urine is convenient for patients and does not need invasive biopsy, thereby diagnosing cancer without side effects. However, as the concentration of cancer factors is low in urine, urine-based biosensors are only used for classifying risk groups rather than for precise diagnosis thus far. Dr. Lee's team at the KIST has been working toward developing a technique for diagnosing disease from urine with an electrical-signal-based ultrasensitive biosensor. An approach using a single cancer factor associated with a cancer diagnosis was limited in increasing the diagnostic accuracy to over 90%. However, to overcome this limitation, the team simultaneously used different kinds of cancer factors instead of using only one to enhance the diagnostic accuracy innovatively. The team developed an ultrasensitive semiconductor sensor system capable of simultaneously measuring trace amounts of four selected cancer factors in urine for diagnosing prostate cancer. They trained AI by using the correlation between the four cancer factors, which were obtained from the developed sensor. The trained AI algorithm was then used to identify those with prostate cancer by analyzing complex patterns of the detected signals. The diagnosis of prostate cancer by utilizing the AI analysis successfully detected 76 urinary samples with almost 100 percent accuracy. The results of the study have been published in the journal ACS Nano.

Read more of this story at Slashdot.

An anonymous reader writes: China produces as many artificial intelligence researchers as the US, but it lags in fundamental research. The government hopes to make up ground with a new AI lab in Beijing that brings together top researchers from AI and industry to focus on things like the mathematical foundations of machine learning and neuroscience-inspired AI. But as WIRED reports, it also suggests that even the Chinese government has concerns about the ethical challenges raised by AI. Among the first projects that the government is funding: a Chinese version of GPT-3 for government use. From the article: Noam Yuchtman, a professor at the London School of Economics, has published work that uses evidence from China to suggest that AI benefits uniquely from state intervention, because algorithms are so hungry for data and computer power that governments have access to. But he adds that such a fast-moving and unpredictable technology may also pose problems for governments. "Innovation by its very nature is sort of uncertain, and perhaps nowhere more so than in AI," he says.

Read more of this story at Slashdot.

Google artificial intelligence researcher Margaret Mitchell has been locked out of corporate systems, making her the second outspoken critic at the company to be sidelined after colleague Timnit Gebru departed in acrimonious circumstances last month. From a report: The Alphabet unit has an Ethical AI team, led by Mitchell, and a set of principles for developing the technology in a socially responsible manner. Gebru tweeted on Tuesday that Mitchell's "corp access is now locked" and that the researcher had been told she would remain locked out "for at least a few days."

Read more of this story at Slashdot.

Dallas, TX-based cloud security firm Trend Micro interviewed 500 IT directors and managers, CIOs and CTOs — and discovered that over two-fifths of them believe they'll be replaced by AI by 2030. ZDNet reports: Only 9% of respondents were confident that AI would definitely not replace their job within the next decade. In fact, nearly a third (32%) said they thought the technology would eventually work to completely automate all cybersecurity, with little need for human intervention. Almost one in five (19%) believe that attackers using AI to enhance their arsenal will be commonplace by 2025. Around a quarter (24%) of IT leaders polled also claimed that by 2030, data access will be tied to biometric or DNA data, making unauthorised access impossible. In the shorter term, respondents also predicted the following outcomes would happen by 2025. They predict that most organisations will have significantly reduced investment in property as remote working becomes the norm (22%) Nationwide 5G will have entirely transformed network and security infrastructure (21%), and security will be self-managing and automated using AI (15%). However, attackers using AI to enhance their arsenal will be commonplace (19%)

Read more of this story at Slashdot.

"Behold the cakie: It has the crispiness of a cookie and the, well, 'cakiness' of a cake." So says a triumphant blog post by Google Cloud's developer advocate and an applied AI Engineer for Google's Cloud AI. "We also made breakies, which were more like fluffy cookies, almost the consistency of a muffin" (or bread). Food and Wine explains the project (in an article shared by Slashdot reader John Trumpian): Inspired by the pandemic-spawned spike in searches for baking, the team at Google Cloud "decided to dive a little deeper into the trend and try to understand the science behind what makes cookies crunchy, cake spongy and bread fluffy," according to a post on their blog. Then, once armed with that machine learning knowledge, they attempted to mix these attributes into what they bill as "two completely new baking recipes...." [T]hese Google Cloud employees organized about 700 recipes covering cookies, cakes, and breads — standardizing measurements, isolating the key ingredients, and re-categorizing things like banana breads that aren't really "breads." Then, they fed them into a tool called "AutoML Tables" to create a machine learning model that was able to predict whether a recipe was a cookie, cake, or bread based on its ingredient amounts. ["If you've never tried AutoML Tables, it's a code-free way to build models from the type of data you'd find in a spreadsheet like numbers and categories — no data science background required," explains the blog post.] Of course, recipes don't necessarily fit perfectly into one category. As Sara Robinson, who led the project, explained, a recipe might come back as 97 percent bread, 2 percent cake, and 1 percent cookie. So what if she asked the model to create its own recipe: something that's 50 percent cookie and 50 percent cake? That's how the Cakie was born. And she was happily surprised by the results. "It is yummy," Robinson said. "And it strangely tastes like what I'd imagine would happen if I told a machine to make a cake cookie hybrid." Based on that success, she and colleague Dale Markowitz continued to tweak their model — which resulted in the Breakie. "We should caveat that while our model gave us ingredients, it didn't spit out any baking directions, so we had to improvise those ourselves," the blog post explains. "And, we added chocolate chips and cinnamon for good measure." Robinson also built a prediction-making web app to help quickly experiment with different ingredient ratios. They ultimately identified which ingredients were the biggest "signal" of cake-ness, cookie-ness, and bread-ness, concluding that "In our case, the amount of butter, sugar, yeast and egg in a recipe all seemed to be important indicators..."

Read more of this story at Slashdot.

schwit1 shares a report from ScienceAlert: [S]cientists have just delivered their verdict on whether we'd be able to control a high-level computer super-intelligence. The answer? Almost definitely not. The catch is that controlling a super-intelligence far beyond human comprehension would require a simulation of that super-intelligence which we can analyze. But if we're unable to comprehend it, it's impossible to create such a simulation. Rules such as "cause no harm to humans" can't be set if we don't understand the kind of scenarios that an AI is going to come up with, suggest the authors of the new paper. Once a computer system is working on a level above the scope of our programmers, we can no longer set limits. Part of the team's reasoning comes from the halting problem put forward by Alan Turing in 1936. The problem centers on knowing whether or not a computer program will reach a conclusion and answer (so it halts), or simply loop forever trying to find one. As Turing proved through some smart math, while we can know that for some specific programs, it's logically impossible to find a way that will allow us to know that for every potential program that could ever be written. That brings us back to AI, which in a super-intelligent state could feasibly hold every possible computer program in its memory at once. Any program written to stop AI harming humans and destroying the world, for example, may reach a conclusion (and halt) or not -- it's mathematically impossible for us to be absolutely sure either way, which means it's not containable. The alternative to teaching AI some ethics and telling it not to destroy the world -- something which no algorithm can be absolutely certain of doing, the researchers say -- is to limit the capabilities of the super-intelligence. It could be cut off from parts of the internet or from certain networks, for example. The new study rejects this idea too, suggesting that it would limit the reach of the artificial intelligence -- the argument goes that if we're not going to use it to solve problems beyond the scope of humans, then why create it at all? If we are going to push ahead with artificial intelligence, we might not even know when a super-intelligence beyond our control arrives, such is its incomprehensibility. That means we need to start asking some serious questions about the directions we're going in.

Read more of this story at Slashdot.

Amazon.com is offering other companies the ability to use the building blocks of the Alexa digital assistant for their own automated versions, the latest effort to embed the company's voice software into other devices. From a report: Fiat Chrysler Automobiles NV will be the first to use Alexa Custom Assistant, relying on Amazon-built speech recognition and other software to power the automaker's in-car tools, Amazon said Friday in a statement. The retail and technology giant also invited other companies to customize the underlying Alexa system with their own wake word, voice and unique capabilities. Alexa is most closely associated with Echo smart speakers, but Amazon has been working to extend the software's reach, and fend off rivals like Apple and Alphabet's Google, by adding utility for tasks like home automation and the potentially lucrative and fiercely contested market for in-car software. Amazon, which lacks the massive base of captive smartphone users of its main rivals, has suggested voice assistants should be able to talk to one another. The company, like competitors, already offers for rent elements of the technology that powers its digital assistant, but Alexa Custom Assistant represents a more complete set of tools, Amazon said.

Read more of this story at Slashdot.

Researchers have created a machine learning system that they claim can determine a person's political party, with reasonable accuracy, based only on their face. TechCrunch reports: The study, which appeared this week in the Nature journal Scientific Reports, was conducted by Stanford University's Michal Kosinski. Kosinski made headlines in 2017 with work that found that a person's sexual preference could be predicted from facial data. [...] The algorithm itself is not some hyper-advanced technology. Kosinski's paper describes a fairly ordinary process of feeding a machine learning system images of more than a million faces, collected from dating sites in the U.S., Canada and the U.K., as well as American Facebook users. The people whose faces were used identified as politically conservative or liberal as part of the site's questionnaire. The algorithm was based on open-source facial recognition software, and after basic processing to crop to just the face (that way no background items creep in as factors), the faces are reduced to 2,048 scores representing various features -- as with other face recognition algorithms, these aren't necessary intuitive things like "eyebrow color" and "nose type" but more computer-native concepts. The system was given political affiliation data sourced from the people themselves, and with this it diligently began to study the differences between the facial stats of people identifying as conservatives and those identifying as liberal. Because it turns out, there are differences. Of course it's not as simple as "conservatives have bushier eyebrows" or "liberals frown more." Nor does it come down to demographics, which would make things too easy and simple. After all, if political party identification correlates with both age and skin color, that makes for a simple prediction algorithm right there. But although the software mechanisms used by Kosinski are quite standard, he was careful to cover his bases in order that this study, like the last one, can't be dismissed as pseudoscience. The most obvious way of addressing this is by having the system make guesses as to the political party of people of the same age, gender and ethnicity. The test involved being presented with two faces, one of each party, and guessing which was which. Obviously chance accuracy is 50%. Humans aren't very good at this task, performing only slightly above chance, about 55% accurate. The algorithm managed to reach as high as 71% accurate when predicting political party between two like individuals, and 73% presented with two individuals of any age, ethnicity or gender (but still guaranteed to be one conservative, one liberal).

Read more of this story at Slashdot.

An anonymous reader quotes a report from VentureBeat: Google researchers developed and benchmarked techniques they claim enabled them to train a language model containing more than a trillion parameters. They say their 1.6-trillion-parameter model, which appears to be the largest of its size to date, achieved an up to 4 times speedup over the previously largest Google-developed language model (T5-XXL). As the researchers note in a paper detailing their work, large-scale training is an effective path toward powerful models. Simple architectures, backed by large datasets and parameter counts, surpass far more complicated algorithms. But effective, large-scale training is extremely computationally intensive. That's why the researchers pursued what they call the Switch Transformer, a "sparsely activated" technique that uses only a subset of a model's weights, or the parameters that transform input data within the model. In an experiment, the researchers pretrained several different Switch Transformer models using 32 TPU cores on the Colossal Clean Crawled Corpus, a 750GB-sized dataset of text scraped from Reddit, Wikipedia, and other web sources. They tasked the models with predicting missing words in passages where 15% of the words had been masked out, as well as other challenges, like retrieving text to answer a list of increasingly difficult questions. The researchers claim their 1.6-trillion-parameter model with 2,048 experts (Switch-C) exhibited "no training instability at all," in contrast to a smaller model (Switch-XXL) containing 395 billion parameters and 64 experts. However, on one benchmark -- the Sanford Question Answering Dataset (SQuAD) -- Switch-C scored lower (87.7) versus Switch-XXL (89.6), which the researchers attribute to the opaque relationship between fine-tuning quality, computational requirements, and the number of parameters. This being the case, the Switch Transformer led to gains in a number of downstream tasks. For example, it enabled an over 7 times pretraining speedup while using the same amount of computational resources, according to the researchers, who demonstrated that the large sparse models could be used to create smaller, dense models fine-tuned on tasks with 30% of the quality gains of the larger model. In one test where a Switch Transformer model was trained to translate between over 100 different languages, the researchers observed "a universal improvement" across 101 languages, with 91% of the languages benefitting from an over 4 times speedup compared with a baseline model. "Though this work has focused on extremely large models, we also find that models with as few as two experts improve performance while easily fitting within memory constraints of commonly available GPUs or TPUs," the researchers wrote in the paper. "We cannot fully preserve the model quality, but compression rates of 10 to 100 times are achievable by distilling our sparse models into dense models while achieving ~30% of the quality gain of the expert model."

Read more of this story at Slashdot.

The maker of a defunct cloud photo storage app that pivoted to selling facial recognition services has been ordered to delete user data and any algorithms trained on it, under the terms of an FTC settlement. TechCrunch reports: The regulator investigated complaints the Ever app -- which gained earlier notoriety for using dark patterns to spam users' contacts -- had applied facial recognition to users' photographs without properly informing them what it was doing with their selfies. Under the proposed settlement, Ever must delete photos and videos of users who deactivated their accounts and also delete all face embeddings (i.e. data related to facial features which can be used for facial recognition purposes) that it derived from photos of users who did not give express consent to such a use. Moreover, it must delete any facial recognition models or algorithms developed with users' photos or videos. This full suite of deletion requirements -- not just data but anything derived from it and trained off of it -- is causing great excitement in legal and tech policy circles, with experts suggesting it could have implications for other facial recognition software trained on data that wasn't lawfully processed. Or, to put it another way, tech giants that surreptitiously harvest data to train AIs could find their algorithms in hot water with the US regulator.

Read more of this story at Slashdot.

Amazon, Walmart, and other companies are using artificial intelligence "to decide whether it makes economic sense to process a return," reports the Wall Street Journal: For inexpensive items or large ones that would incur hefty shipping fees, it is often cheaper to refund the purchase price and let customers keep the products. The relatively new approach, popularized by Amazon and a few other chains, is being adopted more broadly during the Covid-19 pandemic, as a surge in online shopping forces companies to rethink how they handle returns. "We are getting so many inquiries about this that you will see it take off in coming months," said Amit Sharma, chief executive of Narvar Inc., which processes returns for retailers... A Target Corp. spokeswoman said the retailer gives customers refunds and encourages them to donate or keep the item in a small number of cases in which the company deems that option is easier than returning the purchase. A Walmart spokeswoman said the "keep it" option is designed for merchandise it doesn't plan to resell and is determined by customers' purchase history, the value of the products and the cost of processing the returns... Processing online returns can cost $10 to $20, excluding freight, depending on the item, said Rick Faulk, chief executive of Locus Robotics, which uses robots to help automate returns.

Read more of this story at Slashdot.

The author of the book Online Afterlives describes the unusual projects of people like Eugenia Kuyda, co-founder of Luka, an AI-powered chat simulator that books restaurant reservations and makes recommendations. Kuyda worked with computer scientists to convert several thousand text messages between deceased tech entrepreneur Roman Mazurenko and his friends and relatives into a chatbot simulation: "How are you there?" asks a friend. "I'm OK. A little down. I hope you aren't doing anything interesting without me," Roman responds. His friend replies that they all miss him. Another acquaintance asks him if God and the soul exist. Having probably indicated his atheism in chats while he was alive, he says no. "Only sadness." Not content with Luka, Eugenia also designed a chatbot called Replika. A cross between a diary and a personal assistant, Replika asks users a series of questions, eventually learning to mimic their personalities. The goal is to get closer to creating a digital avatar that would be able to reproduce us and replace us once we're dead, but also one that is able to create "friendships" with humans. Since the second half of 2017, over two million people have downloaded Replika onto their mobile devices... Luka and Replika are not the only inventions designed to give a voice to the digital ghosts of the deceased. A few years ago, James Vlahos, an American journalist who has been an AI enthusiast since childhood, created what he calls a "Dadbot." It all started on April 24, 2016, when his father John was diagnosed with lung cancer. Upon learning of his father's illness, James began recording all of their conversations with the idea of writing a commemorative book after his father's death. After 12 sessions, each an hour and a half, he found himself with 91,970 words. The printed transcripts filled around 203 pages... He decided to use the recordings of his father to create something other than a commemorative book. He remembered writing an article that discussed PullString (previously known as ToyTalk), a program designed to create conversations with fictional characters... James used PullString to reorganize the MP3 recordings of his father. He also used it to create his Dadbot, software that works on his smartphone and simulates a written conversation with John, based on the processing of almost 100,000 recorded words... The tone of the conversations reflects the personality of the deceased: "Where are you now?" asks James. "As a bot I suppose I exist somewhere on a computer server in San Francisco. "And also, I suppose, in the minds of people who chat with me."

Read more of this story at Slashdot.

Slashdot reader the_newsbeagle shares an article from IEEE Spectrum: Many associate XPrize with a $10-million award offered in 1996 to motivate a breakthrough in private space flight. But the organization has since held other competitions related to exploration, ecology, and education. And in November, they launched the Pandemic Response Challenge, which will culminate in a $500,000 award to be split between two teams that not only best predict the continuing global spread of COVID-19, but also prescribe policies to curtail it... For Phase 1, teams had to submit prediction models by 22 December... Up to 50 teams will make it to Phase 2, where they must submit a prescription model... The top two teams will split half a million dollars. The competition may not end there. Amir Banifatemi, XPrize's chief innovation and growth officer, says a third phase might test models on vaccine deployment prescriptions. And beyond the contest, some cities or countries might put some of the Phase 2 or 3 models into practice, if Banifatemi can find adventurous takers. The organizers expect a wide variety of solutions. Banifatemi says the field includes teams from AI strongholds such as Stanford, Microsoft, MIT, Oxford, and Quebec's Mila, but one team consists of three women in Tunisia. In all, 104 teams from 28 countries have registered. "We're hoping that this competition can be a springboard for developing solutions for other really big problems as well," Miikkulainen says. Those problems include pandemics, global warming, and challenges in business, education, and healthcare. In this scenario, "humans are still in charge," he emphasizes. "They still decide what they want, and AI gives them the best alternatives from which the decision-makers choose." But Miikkulainen hopes that data science can help humanity find its way. "Maybe in the future, it's considered irresponsible not to use AI for making these policies," he says. For the Covid-19 competition, Banifatemi emphasized that one goal was "to make the resulting insights available freely to everyone, in an open-source manner — especially for all those communities that may not have access to data and epidemiology divisions, statisticians, or data scientists."

Read more of this story at Slashdot.

The machine learning company OpenAI is developing models that improve computer vision and can produce original images from a text prompt. From a report: The new models are the latest steps in ongoing efforts to create machine learning systems that exhibit elements of general intelligence, while performing tasks that are actually useful in the real world -- without breaking the bank on computing power. OpenAI this week is announcing two new systems that attempt to do for images what its landmark GPT-3 model did last year for text generation. DALL-E is a neural network that can "take any text and make an image out of it," says Ilya Sutskever, OpenAI co-founder and chief scientist. That includes concepts it would never have encountered in training, like the drawing of an anthropomorphic daikon radish walking a dog. DALL-E operates somewhat similarly to GPT-3, the huge transformer model that can generate original passages of text based on a short prompt. CLIP, the other new neural network, "can take any set of visual categories and instantly create very strong and reliable visually classifiable text descriptions," says Sutskever, improving on existing computer vision techniques with less training and expensive computational power.

Read more of this story at Slashdot.

Farming equipment like combine harvesters "beam back reams of data to its manufacturer," reports Forbes: GPS records the combine's precise path through the field as it moves. Sensors tally the number of crops gathered per acre and the spacing between them. On a sister machine called a planter, algorithms adjust the distribution of seeds based on which parts of the soil have in past years performed best. Another machine, a sprayer, uses algorithms to scan for weeds and zap them with pesticides. Meanwhile sensors record the wear and tear on the machines, so that when the farmer who operates them heads to the local distributor to look for a replacement part, it has already been ordered and is waiting for them. Farming may be an earthy industry, but much of it now takes place in the cloud. Leading farm machine makers like Chicago-based John Deere or Georgia's AGCO collect data from all around the world thanks to the ability of their bulky machines to extract a huge variety of metrics from farmers' fields and store it online... The amassing of all that data in the hands of the few major companies that sell farm equipment across the country or worldwide has opened up big opportunities for the "smart farming" industry, even as many in the farming community are reluctant to part with information about the fields they plow.... Equipment makers with sufficient sales of machines around the country may in theory actually be able to predict, at least to some small but meaningful extent, the prices of various crops by analyzing the data its machines are sending in — such as "yields" of crops per acre, the amount of fertilizer used, or the average number of seeds of a given crop planted in various regions. Were the company then to sell that data to a commodities trader, say, it could likely reap a windfall: normally, the markets must wait for highly-anticipated government surveys to run their course.

Read more of this story at Slashdot.

Rohit Prasad, Vice President and Head Scientist of Alexa at Amazon, writes: While Turing's original vision continues to be inspiring, interpreting his test as the ultimate mark of AI's progress is limited by the era when it was introduced. For one, the Turing Test all but discounts AI's machine-like attributes of fast computation and information lookup, features that are some of modern AI's most effective. The emphasis on tricking humans means that for an AI to pass Turing's test, it has to inject pauses in responses to questions like, "do you know what is the cube root of 3434756?" or, "how far is Seattle from Boston?" In reality, AI knows these answers instantaneously, and pausing to make its answers sound more human isn't the best use of its skills. Moreover, the Turing Test doesn't take into account AI's increasing ability to use sensors to hear, see, and feel the outside world. Instead, it's limited simply to text. To make AI more useful today, these systems need to accomplish our everyday tasks efficiently. If you're asking your AI assistant to turn off your garage lights, you aren't looking to have a dialogue. Instead, you'd want it to fulfill that request and notify you with a simple acknowledgment, "ok" or "done." Even when you engage in an extensive dialogue with an AI assistant on a trending topic or have a story read to your child, you'd still like to know it is an AI and not a human. In fact, "fooling" users by pretending to be human poses a real risk. Imagine the dystopian possibilities, as we've already begun to see with bots seeding misinformation and the emergence of deep fakes. Instead of obsessing about making AIs indistinguishable from humans, our ambition should be building AIs that augment human intelligence and improve our daily lives in a way that is equitable and inclusive. A worthy underlying goal is for AIs to exhibit human-like attributes of intelligence -- including common sense, self-supervision, and language proficiency -- and combine machine-like efficiency such as fast searches, memory recall, and accomplishing tasks on your behalf. The end result is learning and completing a variety of tasks and adapting to novel situations, far beyond what a regular person can do.

Read more of this story at Slashdot.

schwit1 quotes Intelligent Living: Plenty is an ag-tech startup in San Francisco, co-founded by Nate Storey, that is reinventing farms and farming. Storey, who is also the company's chief science officer, says the future of farms is vertical and indoors because that way, the food can grow anywhere in the world, year-round; and the future of farms employ robots and AI to continually improve the quality of growth for fruits, vegetables, and herbs. Plenty does all these things and uses 95% less water and 99% less land because of it. Plenty's climate-controlled indoor farm has rows of plants growing vertically, hung from the ceiling. There are sun-mimicking LED lights shining on them, robots that move them around, and artificial intelligence (AI) managing all the variables of water, temperature, and light, and continually learning and optimizing how to grow bigger, faster, better crops. These futuristic features ensure every plant grows perfectly year-round. The conditions are so good that the farm produces 400 times more food per acre than an outdoor flat farm. Another perk of vertical farming is locally produced food. The fruits and vegetables aren't grown 1,000 miles away or more from a city; instead, at a warehouse nearby. Meaning, many transportation miles are eliminated, which is useful for reducing millions of tons of yearly CO2 emissions and prices for consumers. Imported fruits and vegetables are more expensive, so society's most impoverished are at an extreme nutritional disadvantage. Vertical farms could solve this problem.

Read more of this story at Slashdot.

DeepMind's latest AI program can attain "superhuman performance" in tasks without needing to be given the rules. From a report: Like the research hub's earlier artificial intelligence agents, MuZero achieved mastery in dozens of old Atari video games, chess, and the Asian board games of Go and Shogi. But unlike its predecessors, it had to work out their rules for itself. It is already being put to practical use to find a new way to encode videos, which could slash YouTube's costs. [...] MuZero could soon be put to practical use too. Dr Silver said DeepMind was already using it to try to invent a new kind of video compression. "If you look at data traffic on the internet, the majority of it is video, so if you can compress video more effectively you can make massive savings," he explained. "And initial experiments with MuZero show you can actually make quite significant gains, which we're quite excited about." He declined to be drawn on when or how Google might put this to use beyond saying more details would be released in the new year. However, as Google owns the world's biggest video-sharing platform -- YouTube -- it has the potential to be a big money-saver. DeepMind is not the first to try and create an agent that both models the dynamics of the environment it is placed in and carries out tree searches -- deciding how to proceed by looking several steps ahead to determine the best outcome. However, previous attempts have struggled to deal with the complexity of "visually rich" challenges, such as those posed by old video games like Ms Pac-Man.

Read more of this story at Slashdot.

Alphabet's Google this year moved to tighten control over its scientists' papers by launching a "sensitive topics" review, and in at least three cases requested authors refrain from casting its technology in a negative light, Reuters reported Wednesday, citing internal communications and interviews with researchers involved in the work. From a report: Google's new review procedure asks that researchers consult with legal, policy and public relations teams before pursuing topics such as face and sentiment analysis and categorizations of race, gender or political affiliation, according to internal webpages explaining the policy. "Advances in technology and the growing complexity of our external environment are increasingly leading to situations where seemingly inoffensive projects raise ethical, reputational, regulatory or legal issues," one of the pages for research staff stated. Reuters could not determine the date of the post, though three current employees said the policy began in June. The "sensitive topics" process adds a round of scrutiny to Google's standard review of papers for pitfalls such as disclosing of trade secrets, eight current and former employees said. For some projects, Google officials have intervened in later stages. A senior Google manager reviewing a study on content recommendation technology shortly before publication this summer told authors to "take great care to strike a positive tone," according to internal correspondence read to Reuters.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Phys.Org: A team of scientists at Freie Universitat Berlin has developed an artificial intelligence (AI) method for calculating the ground state of the Schrodinger equation in quantum chemistry. The goal of quantum chemistry is to predict chemical and physical properties of molecules based solely on the arrangement of their atoms in space, avoiding the need for resource-intensive and time-consuming laboratory experiments. In principle, this can be achieved by solving the Schrodinger equation, but in practice this is extremely difficult. Up to now, it has been impossible to find an exact solution for arbitrary molecules that can be efficiently computed. But the team at Freie Universitat has developed a deep learning method that can achieve an unprecedented combination of accuracy and computational efficiency. The deep neural network designed by [the] team is a new way of representing the wave functions of electrons. "Instead of the standard approach of composing the wave function from relatively simple mathematical components, we designed an artificial neural network capable of learning the complex patterns of how electrons are located around the nuclei," [Professor Frank Noe, who led the team effort] explains. "One peculiar feature of electronic wave functions is their antisymmetry. When two electrons are exchanged, the wave function must change its sign. We had to build this property into the neural network architecture for the approach to work," adds [Dr. Jan Hermann of Freie Universitat Berlin, who designed the key features of the method in the study]. This feature, known as 'Pauli's exclusion principle,' is why the authors called their method 'PauliNet.' Besides the Pauli exclusion principle, electronic wave functions also have other fundamental physical properties, and much of the innovative success of PauliNet is that it integrates these properties into the deep neural network, rather than letting deep learning figure them out by just observing the data. "Building the fundamental physics into the AI is essential for its ability to make meaningful predictions in the field," says Noe. "This is really where scientists can make a substantial contribution to AI, and exactly what my group is focused on." The results were published in the journal Nature Chemistry.

Read more of this story at Slashdot.