Artificial Intelligence Crisis- Major breakthroughs in AI have seen machines being entrusted with business and safety-critical decisions, from guiding vehicles to diagnosing diseases. Yet a reproducibility crisis is creating a cloud of uncertainty over the entire field, eroding the confidence on which the AI economy depends. Reproducibility, the extent to which an experiment can be repeated with the same results, is the basis of quality assurance in science because it enables past findings to be independently verified, building a trustworthy foundation for future discoveries. This is crucial because previous breakthroughs are the barometer by which to measure all subsequent progress. Without the capacity to reproduce past results, the entire basis on which machines are increasingly making legal, corporate and even medical decisions is called into question. This could stop us from being able to benefit from some of the greatest advances in the field, from the AIs that power smart cities to those that find new drug treatments. For example, deep reinforcement learning (RL), whereby machines try every possible solution until they find the right one, could enable driverless cars to endlessly crisscross in virtual reality until they learn to safely change lanes in the real world. Yet experts found that RL results are not easily reproducible, raising questions over whether it can be relied on to ensure road safety. An analysis of 30 AI papers similarly found that the majority of them were difficult to reproduce because key records of their methodologies were missing, from training data sets to study parameters. As a result, Google researcher Ali Rahimi has likened AI to alchemy. The way in which alchemy produced new innovations such as glass alongside false cures such as leeches is directly analogous to the way that AI has discovered potential cancer cures yet failed to distinguish masks from faces. Lack Of Traceability The fundamental problem is that data science is not governed by the same generally accepted standards of quality assurance as other fields of science. As a result, the data trail charting the road from the origins of AI to its latest iterations is shrouded in mystery. There are currently no universal standards governing the data capture, curation and processing techniques that give vital meaning and context to AI experiments. This is the equivalent of climate scientists investigating global warming without any rules on how to document the locations or units of temperature readings. This is particularly concerning as there are so many iterations involved in developing machine-learning tools and there is no universal benchmark of good practice for implementing and recording them all. A single experiment to create a facial-recognition system involves a complex layer cake of processes, from training runs to software updates, file changes and tweaks to the algorithm. If any of these steps is not meticulously recorded, it would be painstakingly difficult to modify the AI or reproduce the original results. Share this:The post How Do We Address The Reproducibility Crisis In Artificial Intelligence? appeared first on Statii News. from http://news.statii.co.uk/how-do-we-address-the-reproducibility-crisis-in-artificial-intelligence/
0 Comments
A.I. Songwriting- “IT’S CHEATING.” That’s the response you’ll hear from self-proclaimed music purists talking about technological innovation in song creation. Sampling, synthesizers, drum machines, Auto-Tune—all have been derided as lazy ways to make chart-topping hits because they take away the human element. (With apologies to Vanilla Ice, Gary Numan, Prince, and T-Pain.) The new argument among fans and musicians will be about the use of artificial intelligence in songwriting. According to several estimates, in the next decade, between 20% and 30% of the top 40 singles will be written partially or totally with machine-learning software. Today, recording pros can use A.I.-powered programs to cue an array of instrumentation (from full orchestral arrangements to hip-hop beats), then alter it by mood, tempo, or genre (from heavy metal to bluegrass). “It’s like the future of self-driving cars,” says Leonard Brody, entrepreneur and cofounder of Creative Labs, a joint venture with Creative Artists Agency that invests in programs to help audio creators get their works delivered to the public. “Level 1 is an artist using a machine to assist them. Level 2 is where the music is crafted by a machine but performed by a human. Level 3 is where the whole thing is machines.” A.I. claiming ownership of a third of the top 40 may be surprising to the casual listener, but it’s a low bar for Drew Silverstein, CEO of Amper, an A.I.-based music composition software company in New York City. Amper’s product allows musicians to create and download “stems”—unique portions of a track like a guitar riff or a hi-hat cymbal pattern—and rework them. Silverstein sees predictive tools as an evolution in the process of music creation. “Starting from quill and parchment centuries ago, then moving into analog and tape and mobile [devices]—A.I. is really just the next step,” he says. Silverstein isn’t the only one with that view. Large technology companies also offer A.I.- powered tools and services for music-making. Among them: IBMWatson Beat, Google Magenta’s NSynth, Sony’s Flow Machines, and Spotify’s Creator Technology Research Lab. The resources, intended for use by artists and labels, use algorithms to analyze libraries of songs and sales charts to predict what may have the best chance of charting (and when). Though the latest developments in A.I. are helping fuel its use in popular music, it’s not really a new idea. More than two decades ago, David Bowie helped create the Verbasizer, a program for Apple’s Mac that randomized portions of his inputted text sentences to create new ones with new meanings and moods—an advanced version of a cut-up technique he used, writing out ideas, then physically slicing and rearranging them to see what stuck. Bowie made use of the Verbasizer for his 1995 album Outside. “What you end up with is a real kaleidoscope of meanings and topic and nouns and verbs all sort of slamming into each other,” said the influential pop star in a 1997 documentary featuring the tool. Share this:The post A.I. Songwriting Has Arrived. Don’t Panic appeared first on Statii News. from http://news.statii.co.uk/a-i-songwriting-has-arrived-dont-panic/ Artificial Intelligence needs structural changes before it can really match up with humans. AI Won’t Replace Us- The late Stephen Hawkins worried that AI could end mankind. It seemed reasonable. Elon Musk warned machines that learned to operate without a human telling them what to do could “destroy humanity as a matter of course without even thinking about it” if it “[had] a goal and humanity just happens [to be] in the way.”. But reality has proven that while AI can beat humans at games, it still fails at common tasks an infant can do, such as holding an object. In fact, to solve this problem, researchers from OpenAI used 6144 CPUs and 8 GPUs to collect about one hundred years of experience and trained the AI for 50 hours. As a result, the robotic hand can handle unknown objects — as long as they are “within reason.” The fundamental gapAs Antonio Bicchi, a professor of robotics at the Istituto Italiano di Tecnologia said, the research had a number of limitations, such as the hand is always facing up so that the objects always fall in the palm. We can’t tell for certain if another 100 years of training data would make the AI even better, or if it needs a new set of training data. What we can say is that humans are exceptionally good at incremental learning. Once a human learns to play with a ball, they can master any ball game quite easily. Or when we learn a foreign language, learning other languages becomes easier every time. But an AI must learn everything from scratch. AI can’t use “other AIs.” It always starts from zero. AI cannot be “combined” with other AIs to do more complex tasks — at least not yet. So, while AI masters skills at a superhuman level, it only masters one task. The missing pieceRecent developments in AI were boosted with the invention of deep learning algorithms and improved computer powers, which seemed to mimic how the brain operates by simulating perceptrons. But the brain is much more than that. We don’t know how the real brain works and, according to Sam Rodriques, we’ll never know until we drill holes in the human skull and plant probes to study how it really works “behind the scenes” (or bones). But what we do know is that the brain is much less rational than we used to think. Studies prove that we decide first, then we try to find reasons for why we decided the way we did. In fact, people with brain damage who were incapable of developing emotions could perfectly describe what they should be doing in logical terms, yet they found it very difficult to make even simple decisions, such as what to eat. Our choices are arguably always based on emotion. Yet, there is no AI system that works like this. AI can’t reason, which has led to hidden AI bias in many projects. Of course, there is work in progress to solve these cases, but AI was primarily designed as a black box since we did not know how to code it in the first place. In addition, just throwing more power in and building bigger machines is not the way if the machine takes the wrong path, and it can be expensive to know where it is really going since we can’t know what it has learned. The post AI Won’t Replace Us Until It Becomes Much More Like Us appeared first on Statii News. from http://news.statii.co.uk/ai-wont-replace-us-until-it-becomes-much-more-like-us/ AI NIPS- THE FUTURE OF humanity will be shaped by artificial intelligence. Now some of the best brains working on the technology are riven by a debate about a four-letter acronym that some say contributes to the field’s well-documented diversity problems. NIPS is the name of AI’s most prominent conference, a venue for machine learning research formally known as the Annual Conference on Neural Information Processing Systems. Researchers at tech companies including Google and leading universities allege that the name contributes to an atmosphere unwelcoming to women. The acronym has long inspired anatomical jokes about nipples; others dislike the word’s racist connotations. “It encourages juvenile behavior and it’s a distraction from the science we’re trying to do,” says Anima Anandkumar, a Caltech professor and director of research at chipmaker Nvidia. Late Wednesday she tweeted a link to an online petition asking the NIPS board to rename the event. There are now more than 800 signatories, including researchers at Amazon, Microsoft, and Google. They include Jeff Dean, who as Google’s top AI boss leads one of the world’s largest and most influential AI research groups, dubbed Google Brain. Dean told WIRED Thursday that he plans to raise his concerns with members of the NIPS board when he gets the chance. “I think enough people are made to feel uncomfortable by the current name that the NIPS board should change the name,” he tweeted Wednesday. University of Washington grad student Maarten Sap’s attempt to follow the brouhaha helped illustrate the complaints from Dean and others: When he plugged “nips” into Twitter’s search function, it led him to pornographic tweets. AI is projected to reshape everything from health care to war, but the community of people working on the technology is markedly different from the society it is supposed to serve. WIRED and startup Element AI found that in recent years at NIPS and two other leading academic conferences, only about 12 percent of people presenting work were women. That suggests the field is even less diverse than the notoriously monocultural tech industry. Some researchers fear this raises the risk of incidents like those in which image recognition systems have been found to have skewed views of women or black people. The name NIPS became a flashpoint in the debate over how to make AI more inclusive earlier this year after complaints about sexist behavior at the conference, which takes place each December. A blog post about an on-stage remark about sexual harassment at last year’s conference led to investigations into accusations of physical harassment by Google’s director of statistics research Steven Scott, and University of Minnesota professor Brad Carlin. Both subsequently left their their jobs. That incident helped inspire an open letter in March from 112 Johns Hopkins University faculty asking the NIPS board to seek a name less “vulnerable to sexual puns.” In April the board said it would consider alternatives, and in August it surveyed attendees from the past five years on crowdsourced options including SNIPS or ICOLS. Share this:The post AI RESEARCHERS FIGHT OVER FOUR LETTERS: NIPS appeared first on Statii News. from http://news.statii.co.uk/ai-researchers-fight-over-four-letters-nips/ Oracle co-Chief Executive Mark Hurd views it as a marathon, one in which Oracle customers are slowly migrating to the cloud. “This is not a switch flip,” Hurd told Barron’s in San Francisco on Wednesday. “This will take months, quarters, years, to ripple through” Oracle’s business. Hurd thinks Oracle just gave customers a compelling reason to make the switch. This week at its annual developers conference, Oracle announced enhanced security to its cloud technology that includes “autonomous robots” that “find and kill” malicious attacks from bot-equipped hackers, he said. He believes it’s a game-changer—less than 1% of all Oracle databases are encrypted—in a tech marketplace dominated by news of security snafus atFacebook (FB), Alphabet ’s Google (GOOGL), and elsewhere. “You can make the argument that this is the biggest technical release in our [41-year] history,” Hurd said. “There is no question about its impact on helping cloud sales.” Still, despite a rally that briefly lifted shares this summer, Oracle stock has slumped amid questions about its cloud business. Oracle stock is down 9% since Oct. 1. Oracle reported $6.61 billion in fiscal first-quarter revenue for cloud services and license support in September, which fell short of analysts’ estimates of $6.68 billion. Oracle also came up shy in overall revenue, at $9.19 billion, missing the $9.24 billion forecast by analysts polled by FactSet. The company has missed revenue estimates seven of the past 13 quarters. Deepening concerns, the putative head of Oracle’s cloud product strategy, Thomas Kurian, said he isn’t coming back after the company initially said in early September that he was taking “extended time off” from the company. Kurian reportedly clashed with Oracle co-founder, Chairman and Chief Technology Officer Larry Ellison over cloud strategy. Oracle declined to comment on Kurian’s departure. Adding to the tension, Ellison has boasted at previous OracleWorlds that Oracle was coming after the juggernaut Amazon Web Services, as Oracle shifts from a traditional model of licensing and maintenance to subscription-based cloud computing. In February, Oracle said it would quadruple its number of giant data-center complexes over the next two years. But the case has been just the opposite the past few years, based on market share of the $38.8 billion worldwide market for DBMS, defined as business software for data management deployed onsite or in the computing cloud. While Oracle remains the market leader, its market share eroded to 37% in 2017 from 43.6% in 2013, according to numbers compiled by market researcher Gartner. Amazon’s slice, meanwhile, has rocketed to 9.3% from 0.9% in the same time frame, and Microsoft’s has improved to 21.7% from 18%. “Hurd talked about the slow rate of adoption in the cloud as if it were a feature and not a bug,” said Gartner analyst Merv Adrian, who is attending OracleWorld and had a briefing with Hurd. “He put a brave face on things. Hurd is saying some large customers are moving cautiously in shifting operations to the cloud.” The post Oracle CEO Mark Hurd Says His Customers Will Move to the Cloud. Eventually. appeared first on Statii News. from http://news.statii.co.uk/oracle-ceo-mark-hurd-says-his-customers-will-move-to-the-cloud-eventually/ Can these key catalysts keep up their momentum?Microsoft Calendar- Software giant Microsoft (NASDAQ:MSFT) has continued to impress investors in 2018 amid its ongoing progress becoming a more cloud-centric company. The company recently wrapped up its fiscal 2018 with a 14% year-over-year increase in revenue and an impressive 21% boost to operating income. Shares have surged 28% year to date. In less than two weeks, investors will get to see whether the company has been able to keep up its strong momentum into fiscal 2019. Microsoft reports its first-quarter earnings for fiscal 2019 on October 24. Ahead of Microsoft’s earnings release, here’s an overview of some of the key areas investors will want to watch. Commercial-cloud revenueCore to Microsoft’s momentum recently is the company’s rapidly growing commercial-cloud revenue. As a revenue category that lumps together some of Microsoft’s biggest commercial-cloud products — including Office 365 commercial, Azure, and Dynamics 365 — Microsoft’s commercial-cloud revenue has been a major catalyst for the company. In Microsoft’s most recent quarter, commercial-cloud revenue increased 53% year over year, to $6.9 billion, accounting for 23% of the quarter’s total revenue. For Microsoft’s first quarter of fiscal 2019, investors should look for more year-over-year growth for the segment of around 50%. Microsoft CFO Amy Hood indicated in the company’s fiscal fourth-quarter earnings call that the segment continues to fire on all cylinders, saying: “Customer commitment to our cloud platform continues to increase. In FY18, we closed a record number of multi-million dollar commercial cloud agreements and more than doubled the number of $10 million-plus Azure agreements.” AzureCloud-computing service Azure, which falls under Microsoft’s commercial-cloud revenue categorization, also will be worth taking a close look at. Not only is Azure Microsoft’s biggest contributor to the company’s commercial-cloud revenue, but it’s growing at a mind-boggling rate. In Microsoft’s fiscal fourth quarter, Azure revenue surged 89% year over year, or 85% year over year in constant currency. Given that this was a slight deceleration compared to the 93% revenue growth (89% growth in constant currency) that Azure saw in the third quarter of fiscal 2018, investors shouldn’t be surprised to see some more modest deceleration in fiscal Q1. Of course, investors should look for Azure to grow at a high year-over-year growth rate of around 80%. Investors also should look for an update on LinkedIn, which Microsoft acquired in 2016. The social network for professionals has been an excellent performer for the software giant. Fiscal fourth-quarter revenue from LinkedIn was up 37% year over year. Even more impressive, this was the segment’s fifth quarter in a row of accelerating revenue growth. Article Credit: The Motley Fool The post Microsoft Earnings: Mark Your Calendar appeared first on Statii News. from http://news.statii.co.uk/microsoft-earnings-mark-your-calendar/ Analytics Organization- When I left the MIT Sloan School of Management in 1993, little did I know it at the time, but I was armed with the knowledge and mindset of a paradigm that would become known as Big Data. For the record, I owe tremendous credit to professors at MIT such as Steve Graves and Erik Brynjolfsson whose insights and perspectives left an indelible mark on my thinking. Although I could not fully articulate all of the details of what Big Data meant at the time, I did have an unwavering conviction that the lifeblood of a high-velocity organization would eventually be based on real-time analytics of and execution on mass quantities of data. And, to take it one step further, Big Data would revolutionize how companies operated and would drive the creation of whole new businesses (and perhaps even industries). When I left MIT in 1993 for Intel, anyone could clearly see that computational power was advancing at a staggering rate and that the World Wide Web along with the broader infrastructure of the Internet were going to play a central role in this revolution. I knew that whatever “it” was – was going to be very big. But in all honesty, in my wildest dreams, I could never have predicted how big this digital transformation would become and how it would continue to evolve. One afternoon in the mid-90’s, I wrote down a vision of an operating model for a corporation that would employ vast amounts of data and automated algorithms to enable the company to autonomously optimize and run its operations in real-time: integrating across operational domains that encompassed day-to-day tactical decision-making, operational planning and management and the long-term strategic decision-making. At the essence of this vision was an organization whose very existence and operational ethos was built on data and analytics. This is about as close as I ever got to a crystal ball. Imagine my delight and amazement when a few short years later, I had the opportunity to join a small, but growing company (Amazon.com), led by a visionary leader (Jeff Bezos) that shared this vision almost to a tee! I had found myself in the business and career analog of the “kid in the candy store”! Candidly, I thought that life could never be better – until I found myself a decade ladder in an even bigger role at Google. During my entire professional career, I have built analytics functions and groups that operated at the core of any operation that I’ve managed. Fortunately, along the way, I have had a disproportionate share of success, but have made some mistakes as well. I’d like to share some of those key learnings along the way. This article will be mostly retrospective in nature – I’ll save the predictions for a future article. Share this:The post Building An Analytics-Centric Organization appeared first on Statii News. from http://news.statii.co.uk/building-an-analytics-centric-organization/ Bitcoin die- No one knows how long cryptocurrencies will last, but it’s a decent bet they might outlast you. Passing your digital holdings on to loved ones after your death isn’t as simple as bequeathing cash or other property, though, particularly since wills aren’t designed for confidential information. Because a private key is all that’s necessary to transfer funds from a wallet, including it in your will might be a terrible idea. “I would strongly advise against anyone putting any information they consider private into their will,” says estate planning attorney Gordon Fischer. “Wills, after your death, become court documents and are generally public documents, accessible by anyone.” A private key is like an unchangeable password, which is generated when you create a new cryptocurrency wallet. It should always be kept as safe and secure as possible. Although a will might not enter public records immediately, it’s unwise to risk exposing the keys to your crypto wallets at all. Your family might not recognize the significance of a private key right away, and by the time they do, your digital wealth could be pilfered by crafty crooks or other unsavory characters. Fischer notes that trusts, however, are “generally private documents.” With conventional assets, there’s an established procedure for claiming them through probate court, but with cryptocurrencies the process is less certain. Complicating matters is that many cryptocurrency exchanges don’t let their customers name beneficiaries. Coinbase, the largest trading platform, puts the burden on the heirs to claim any assets left by the deceased (so, hopefully your family knows which exchange you use). Another major bourse, Gemini, declined to comment for this story. Of course, the problem isn’t confined to the cryptocurrency world: Popular stock trading app Robinhood, which recently began offering crypto trading, doesn’t offer basic beneficiary support. Sadly, as a result, proving that heirs are entitled to crypto inheritance could quickly become a protracted legal nightmare. Share this:The post What happens to your bitcoin after you die? appeared first on Statii News. from http://news.statii.co.uk/what-happens-to-your-bitcoin-after-you-die/ The contracts that were seen as a step toward bringing crypto to Wall Street remain a tiny market. Bitcoin Future- On a Sunday evening last December, as the cryptocurrency craze consumed the world, traders waited eagerly at their computers to witness the debut of a flashy new financial product. Bitcoin futures were set to start trading on Cboe Global Markets Inc.’s exchange at 5 p.m. Chicago time. Futures allow an investor to place bets on the price something will hit at a later date without having to buy the asset itself. Industry enthusiasts hoped the contracts would help bring Bitcoin trading into the financial mainstream, ushering in big investors with bulging pocketbooks. In the months leading up to the debut, Bitcoin’s price went ballistic. It surged around 600 percent between when Cboe revealed its plans in early August and CME Group Inc. started trading its own version in mid-December, which coincided with Bitcoin’s record high of about $20,000. It has since lost more than half its value. It’s not surprising that the cryptocurrency peaked within hours of the kickoff of CME’s contracts, according to Michael Unetich, vice president of cryptocurrencies at Chicago-based Trading Technologies International Inc. Some people were expecting tens of thousands of contracts per day to trade, he says, “and the market just wasn’t ready for that to happen.” Ten months later, some of the hopes for Bitcoin futures look more like pipe dreams. Cboe and CME combined traded about 9,000 contracts a day in the third quarter. “It has not been what you would call a roaring success,” says Craig Pirrong, a finance professor at the University of Houston and an expert on futures trading. “Institutional players have stayed on the Bitcoin sidelines, and as long as they are, the futures contracts are likely not to generate substantial amounts of volume.” The average of about 5,000 daily contracts at CME in the third quarter is up from around 3,500 in the prior quarter. Still, by comparison, CME traded more than 18 million contracts daily in the second quarter on products tied to everything from oil and gold to interest rates and the S&P 500. “We’re not seeing huge flows” for Bitcoin contracts, CME Chief Executive Officer Terry Duffy told Bloomberg Television in July. The post Anybody Want Bitcoin Futures? Anybody? appeared first on Statii News. from http://news.statii.co.uk/anybody-want-bitcoin-futures-anybody/ Bitcoin Anniversary- Bitcoin has come a long way in just 10 short years. Despite the bitcoin price struggling to regain its highs this year, the long-term trend is hard to ignore. The bitcoin price is up a ridiculous 10.6 million percent since CoinDeskbegan tracking it back in July 2010 — rising from 0.06 cents to $6,421 over that time. Next week, on October 31, it will be the 10th anniversary of Satoshi Nakamoto’s release of the bitcoin whitepaper — leading many to look ahead to what the next 10 years will bring. Bitcoin has dominated investor interest over the last ten years, but that could be very different in the coming decade. It’s only in the last five or so years that bitcoin has had any major competitors. Ethereum, the world’s second largest cryptocurrency by market capitalization, was first conceived in 2013. Since then, hundreds of cryptocurrencies have exploded onto the scene, many of which were born and died in the cryptocurrency gold rush of 2017. Bitcoin has maintained its so-called dominance over the sector, however. Bitcoin dominance — its share of the total cryptocurrency market capitalization — hit year-to-date highs in mid-September. Though it has fallen somewhat from those highs, it has remained above 50%. Bitcoin dominance barely dipped below 80% until March last year, according to CoinMarketCap data. But it has lost ground to the likes of ethereum, ripple (XRP), bitcoin cash, litecoin, and dash since then. “I believe that bitcoin’s influence and dominance of the cryptocurrency sector will drastically reduce in its second decade,” said Nigel Green, founder and CEO of deVere Group, a financial advisory group. “This is because as mass adoption of cryptocurrency grows, more and more digital assets will be launched – by organizations in both the private and the public sectors. This will increase competition for bitcoin and dent its market share.” Green expects the bitcoin and wider cryptocurrency market will expand by “at least” 5,000% in the next 10 years — meaning it could be worth $20 trillion, up from $400 billion today. Share this:The post Bitcoin Price Bulls Come Out Fighting Ahead Of 10 Year Anniversary appeared first on Statii News. from http://news.statii.co.uk/bitcoin-price-bulls-come-out-fighting-ahead-of-10-year-anniversary/ |
About Us
Lates News about computer, businesss softwares and much more. Archives
December 2018
Categories |