AI touch myself: Scientists create self-replicating neural network

AI touch myself: Scientists create self-replicating neural network

Instead of painstakingly creating the layers of a neural network and guiding it’s development as it becomes more advanced – they’ve automated the process. Eventually a neural network that can predict its own growth could lead to an AI system that’s resilient to efforts to delete or scale it back. Theoretically, humans could try and remove specific components or delete the entire program but tiny snippets of code, perhaps stored in a secure cloud, could bring entire systems back online near instantaneously. In fact, the researchers on the Neural Network Quine paper ran into the same power consumption problem with their digitally-evolving AI that exists in nature. Because it takes extra resources to self-replicate and create a better version of itself, AI that does so is less successful at accomplishing tasks than more traditional neural networking methods. Where those can reach near 100 percent accuracy at tasks such as image recognition after only a nominal number of “evolutions,” or iterations spent trying a particular task, the self-replicating model is at least ten percent worse after the same number of tries. It’s not entirely clear why this is so. But we note that this is similar to the trade-off made between reproduction and other tasks in nature. The researchers, Oscar Chang and Hod Lipson, published their fascinating paper titled “Neural Network Quine” earlier this month, and with it a novel new method for “growing” a neural network. But this is the first time a neural network, designed with another purpose (in this case image recognition), has been built with a self-replication mechanism baked-in. The research is still in the early stages, but future iterations will include neural networks that can use the same self-replication techniques to recreate other neural networks. The primary motivation here is that AI agents are powered by deep learning, and a self-replication mechanism allows for Darwinian natural selection to occur, so a population of AI agents can improve themselves simply through natural selection – just like in nature – if there was a self-replication mechanism for neural networks. Various components of the network, agents, do specific tasks – such as finding all the images of cats in a stack of six million images or attempting to replicate a human style of art. These networks can improve in a myriad of different ways, including having two agents argue, or approaching different aspects of a task with individual agents and then combining the “knowledge” each gathers. But with the quine system that Lipson and Chang have created, the neural network improves through “evolving” new versions of agents within itself by predicting what they will look like in the future after they’ve learned new information.

Read More

Chemical synthesis with artificial intelligence: Researchers develop new computer method

Chemical synthesis with artificial intelligence: Researchers develop new computer method

After this breakthrough in the world of chess, the board game Go was long considered to be a bastion reserved for human players due to its complexity. This analysis provides the recipe, which is then used for working "forwards" in the laboratory to produce the target molecule, proceeding from the starting materials. Although easy in theory, the process presents difficulties in practice. "Just like in chess, in every step or move, you've got variety of possibilities to choose from," says Segler. "In chemistry, however, there are orders of magnitude more possible moves than in chess, and the problem is much more complex." This is where the new method comes into play, linking up the deep neural networks with the Monte Carlo Tree Search—a constellation so promising that a large number of researchers from a variety of disciplines are working on it. The Monte Carlo Tree Search is a method for assessing moves in a game. Using the Monte Carlo Tree Search, the computer can test whether the reactions predicted really do lead to the target molecule." The idea of using computers to plan syntheses isn't new. "The idea is actually about 60 years old." says Segler. "People thought it would be enough, as in the case of chess, to enter a large number of rules into the computer. Added to this is the fact that the number of publications with new reactions doubles every 10 years or so. Neither chemists nor programmers can keep up with that. In a double-blind AB test, the Muenster researchers found that chemists consider these computer-generated synthesis routes to be just as good as existing tried-and-tested ones. "We hope that, using our method, chemists will not have to try out so much in the lab," Segler adds, "and that as a result, and using fewer resources, they will be able to produce the compounds which make our high standard of living possible." Explore further: AlphaZero just wants to play More information: Marwin H. S. Segler et al, Planning chemical syntheses with deep neural networks and symbolic AI, Nature (2018). The recipe for the success of this computer program is made possible through a combination of the so-called Monte Carlo Tree Search and deep neural networks based on machine learning and artificial intelligence. A team of researchers from the University of Muenster in Germany has now demonstrated that this combination is extremely well suited to planning chemical syntheses—so-called retrosyntheses—with unprecedented efficiency. Marwin Segler, the lead author of the study, says, "Retrosynthesis is the ultimate discipline in organic chemistry. Chemists need years to master it—just like with chess or Go. In addition to straightforward expertise, you also need a goodly portion of intuition and creativity for it. So far, everyone assumed that computers couldn't keep up without experts programming in tens of thousands of rules by hand. What we have shown is that the machine can, by itself, learn the rules and their applications from the literature available." Retrosynthesis is the standard method for designing the production of chemical compounds. Going backwards mentally, the principle is that the compound is broken down into ever smaller components until the basic components have been obtained.

Read More

Astronomers use artificial intelligence to spot 6000 new craters on the Moon – The Verge

Astronomers use artificial intelligence to spot 6000 new craters on the Moon – The Verge

One of the biggest challenges in astronomy is also the most obvious: space is big, and it takes a long time to look at it all. But using AI to find these craters is important, as it demonstrates another way machine learning can automate a labor-intensive task. The less time astronomers have to spend flicking through pictures of the Moon, labeling craters by hand, the more they have to focus on other, more challenging research. Plus, the more we know about the Moon’s craters, the better we can theorize about the history and formation of our Solar System. The blue circles are craters identified by humans that the network successfully spotted; the red circles are new craters the network found; and the purple ones are those it missed. As the researchers who conducted the work explain in an unpublished paper, they trained their network using a data set of craters previously identified by humans. Once the program had learned what craters looked like, it was turned loose on a new section of the Moon’s surface (roughly one-third of its total surface area). This system wasn’t perfect, and when tested against a human crater-spotter, it only found 92 of the same feature. It turns out that the same machine vision tools developed for tasks like guiding self-driving cars are also perfect for sorting through vast amounts of astronomical data. So, astronomers announced this month that they’d used AI to find 6,000 new craters on the Moon. The Moon is estimated to have hundreds of thousands of craters, mostly caused by impacts with asteroids and meteors. First, because the Moon has no atmosphere, these objects have a free path down to the surface (unlike on Earth where air friction slows them down and reduces them in size).

Read More

We asked a neural network to bake us a cake. The results were…interesting.

We asked a neural network to bake us a cake. The results were…interesting.

But simulated brain cells in so-called neural networks can mimic our problem-solving skills. It gets this format every time because of consistency across all the recipes we pulled from. But my AI hasn’t seen as many examples of how to use rare ingredients like sesame. A neural net has to learn these from scratch with few examples. Chocolate is just a lucky guess pulled from your average cookbook. “Until golden brown” could mean sweet or savory, and ambiguity confuses the network. An AI will look at a dataset, figure out its governing rules, and use those instructions to make something new. 5. Feedback The confused network is spitting out random words—creating more confusion, leading to more random words. But it remembered to close its parenthesis. There’s probably a neuron solely devoted to parentheses. Words like “frost” or “serve” can cue a ­network to finish; many recipes in our dataset end this way. This article was originally published in the Spring 2018 Intelligence issue of Popular Science. We already employ these bots to recognize faces, drive cars, and caption images for the blind. I fed a neural network thousands of recipes and asked it to whip up something of its own. 1. Amnesiac AI To keep processing fast, the network recalls only 65 characters at a time. It adds cocoa just before it would otherwise forget we’re making “chocolate.” Hopefully it forgets that black puddings usually feature blood.

Read More

Everything you need to know about a new EU data law that could shake up big US tech

Everything you need to know about a new EU data law that could shake up big US tech

Not only will it affect organizations located within the EU, but it will also apply to companies outside of the region if they offer goods or services to, or monitor the behavior of, people in the bloc. So companies will not be able to use vague or confusing statements to get you to agree to give them data. "If you have a page of different consent, and saying by clicking here you consent to lots of things, that will be wrong, you need to be able to apply that consent individually," Harry Small, a partner at law firm Baker & McKenzie, told CNBC by phone. Another rule will make it mandatory for companies to notify their data protection authority about a data breach within 72 hours of first becoming aware of it. The processor of the data will need to notify customers "without undue delay" after learning of the breach, according to an EU document. You will be able to access the personal data being stored by companies and find out where and for what purpose it is being used. This means you can ask whoever is controlling your data to erase it and potentially stop third parties processing it too. Another provision of GDPR allows people to take their data and transfer it to a different service provider. An organization in breach of GDPR laws will be fined up to 4 percent of annual global turnover or 20 million euros ($24.6 million), whichever is bigger. Some of the biggest technology companies are making billions in turnover every year so this could be a big hit if they were to breach any rules. The big technology firms who have huge user bases and handle massive amounts of data have spoken about what they are doing. But most likely you haven't because it sounds boring, but it's really important and CNBC has a guide to help you understand it. "We think there is a risk that reported MAUs (monthly average users) could drop off for Facebook and Twitter starting in late 2Q. DAUs (daily average users) are far more important and less of a GDPR concern for the social networks, but may also drop off a bit," Barclays analysts said. Our checks suggest that most companies using cookies and tags for digital marketing should be relatively unchanged as most publishers have been using GDPR compliant notifications for months ahead of the May mandate." It's a piece of European Union (EU) legislation that could have a far-reaching impact on some of the biggest technology firms in the world including Facebook and Google. European authorities have given companies two years to comply and it will come into force on May 25, 2018. It replaces a previous law called the Data Protection Directive and is aimed at harmonizing rules across the 28-nation EU bloc. The aim is to give consumers control of their personal data as it is collected by companies.

Read More

Microsoft doubles down on artificial intelligence in engineering reorganization

Microsoft doubles down on artificial intelligence in engineering reorganization

Apple aims to patent that idea Be bombastic: Top investors share startup tips for female entrepreneurs TLDR: Big week for Bill Gates, Trump vs. Amazon, electric ferries and an Apple AirPods giveaway Tech industry’s cavalier and arrogant attitude foments backlash but started long before Trump, top execs say Working Geek: Nintex CEO Eric Johnson says lemonade, hard knocks, discipline led to success Thanks, Trump: Amazon stock tumbles and Jeff Bezos loses $5.2B off his net worth in one day Smartsheet files for IPO, aims to become Seattle region’s next public company Trucking startup Convoy partners with Goodyear, surpasses 225 employees and 100K trucks Elon Musk takes the wraps off the Boring Company’s next product line: bricks Windows chief leaving Microsoft as CEO Satya Nadella rolls out massive engineering reorg How Funko pops out a Pop! Inside the process of creating collectible figures, from idea to product Bill Gates and teen daughter share love for author John Green and pen joint review of his latest book GoDaddy signs multiyear deal with Amazon Web Services for ‘vast majority’ of its computing infrastructure Digital signature giant DocuSign files for IPO, seeks $100M in tech industry’s latest public offering President Trump reportedly ‘obsessed’ with regulating Amazon, wants to clip Jeff Bezos’ wings Big Fish Casino video game constitutes illegal online gambling, federal appeals court rules Amazon discreetly visits HQ2 cities in Phase 2 of extraordinary second headquarters competition Facebook escalates AI talent wars, makes key hire from Paul Allen’s AI2 in push to grow Seattle team Microsoft doubles down on artificial intelligence in engineering reorganization Seattle’s socialist councilmember wants to tax Amazon and other big companies to fund housing: Is that a good idea?

Read More

Deep learning, artificial intelligence leading the way to smart houses

Deep learning, artificial intelligence leading the way to smart houses

baylorlariat.com Dr. Liang Dong is leading deep learning research. Here, he shows off a digital communication board. Baylee VerSteeg | Multimedia Journalist By Samantha Amaro | Reporter Plans for smart houses in the future are slowly becoming more and more plausible. A house that does all the manual labor for the occupants, where dinner is ready on the kitchen table and all the amenities in a house are included in these plans. Thanks to Baylor…

Read More

France wants to become an artificial intelligence hub

France wants to become an artificial intelligence hub

“[Artificial intelligence] is a technological, economical, social and obviously ethical revolution,” Macron said in a speech. “This revolution won’t happen in 50 or 60 years, it’s happening right now. Today, Samsung, Fujitsu, DeepMind, IBM and Microsoft all announced plans to open offices in France to focus on AI research. This represents tens of millions of dollars in investments and hundreds of employees. “Everybody is saying that Silicon Valley is overflowing right now,” a source close to the French President told me. That’s why big tech companies need to find talent outside of the U.S. — Emmanuel Macron (@EmmanuelMacron) March 28, 2018 Foreign companies creating hundreds of jobs isn’t going to foster public research and European tech giants though — these companies are just tapping the smartest brains they can find. That’s why the French government wants to make it easier to work on fundamental research papers when you work for a private company. The INRIA is going to create a national AI research program with four or five partners. The goal is quite simple — Macron said that there should be twice as many people studying and researching AI projects in France. It’s also going to get easier if you want to create a startup based on your research work or if you want to work for a private company during your PhD. Some of the best mathematics and engineering schools are in France, and some of the best data scientists and AI researchers come from France. Second, France is going to set some new boundaries when it comes to data. French administrations are going to share new data sets so that anyone can build AI services using those data sets. When it comes to health data, it looks like France wants to avoid another NHS/DeepMind scandal. While multiple French governments have worked on some kind of health data hub, Macron announced that this time it’s going to happen for real. The INDS is going to make sure that services and public institutions respect your privacy and grant authorizations on a case-by-case basis. Third, when it comes to regulation, companies will be able to experiment in multiple industries. Overall, France is going to invest $1.85 billion (€1.5 billion) in AI projects, from public research to startup investments. Macron said today that AI startups should be the first priority of this new fund. The French administration already has to share all its algorithms and data that they use following Axelle Lemaire’s law. Research projects or companies financed with public money will also have to share everything — this could impact public infrastructure companies for instance. And the French government wants to capitalize on that soft power to make an AI push. That’s why schools and universities should make sure that they train a diverse group of people. As Next INpact pointed out, there have been multiple reports on artificial intelligence over the past few years — FranceIA, the CNIL, the OPECST and the European Economic and Social Committee all wrote their own recommendations when it comes to AI policies. According to a source close to the French President, multiple ministers now have to focus on artificial intelligence for their own industries. Now, it’s all about convincing the rest of the government to put aside all the urgent tasks for a minute and look at what’s important. France’s answer is quite complicated because the government doesn’t want to inject a ton of public money and call it a day. First, many private companies have opened or plan to open AI research centers in France.

Read More

Billionaire father and son who have their sights set on the British engineering giant GKN are so ruthless they have been branded ‘financial terrorists’ writes GUY ADAMS

dailymail.co.uk � � Femail Today ‘There is no need to even defend myself’: Chris Brown insists ‘no foul play’ after being photographed with hands around woman’s throat in Miami Kate Upton flaunts her eye-popping assets and killer curves in plunging black lingerie for sizzling photoshoot� for Italian underwear brand Yamamay Kim Kardashian’s daughter North feeds elephant as Kanye West dotes on son Saint during zoo day Duo posed for a sweet father-son photos Post-prison makeover!…

Read More
1 2 3 29