 |
|
|
|||||
Teaching machines in the way that animal trainers mold the behavior of dogs or horses has been an important method for developing artificial intelligence and one that was recognized Wednesday with the top computer science award.Two pioneers in the field of reinforcement learning, Andrew Barto and Richard Sutton, are the winners of this year’s A.M. Turing Award, the tech world’s equivalent of the Nobel Prize.Research that Barto, 76, and Sutton, 67, began in the late 1970s paved the way for some of the past decade’s AI breakthroughs. At the heart of their work was channeling so-called “hedonistic” machines that could continuously adapt their behavior in response to positive signals.Reinforcement learning is what led a Google computer program to beat the world’s best human players of the ancient Chinese board game Go in 2016 and 2017. It’s also been a key technique in improving popular AI tools like ChatGPT, optimizing financial trading and helping a robotic hand solve a Rubik’s Cube.But Barto said the field was “not fashionable” when he and his doctoral student, Sutton, began crafting their theories and algorithms at the University of Massachusetts, Amherst.“We were kind of in the wilderness,” Barto said in an interview with The Associated Press. “Which is why it’s so gratifying to receive this award, to see this becoming more recognized as something relevant and interesting. In the early days, it was not.”Google sponsors the annual $1 million prize, which was announced Wednesday by the Association for Computing Machinery.Barto, now retired from the University of Massachusetts, and Sutton, a longtime professor at Canada’s University of Alberta, aren’t the first AI pioneers to win the award named after British mathematician, codebreaker and early AI thinker Alan Turing. But their research has directly sought to answer Turing’s 1947 call for a machine that “can learn from experience”which Sutton describes as “arguably the essential idea of reinforcement learning.”In particular, they borrowed from ideas in psychology and neuroscience about the way that pleasure-seeking neurons respond to rewards or punishment. In one landmark paper published in the early 1980s, Barto and Sutton set their new approach on a specific task in a simulated world: balance a pole on a moving cart to keep it from falling. The two computer scientists later coauthored a widely used textbook on reinforcement learning.“The tools they developed remain a central pillar of the AI boom and have rendered major advances, attracted legions of young researchers, and driven billions of dollars in investments,” said Google’s chief scientist Jeff Dean in a written statement.In a joint interview with the AP, Barto and Sutton didn’t always agree on how to evaluate the risks of AI agents that are constantly seeking to improve themselves. They also distinguished their work from the branch of generative AI technology that is currently in fashionthe large language models behind chatbots made by OpenAI, Google and other tech giants that mimic human writing and other media.“The big choice is, do you try to learn from people’s data, or do you try to learn from an (AI) agent’s own life and its own experience?” Sutton said.Sutton has dismissed what he describes as overblown concerns about AI’s threat to humanity, while Barto disagreed and said “You have to be cognizant of potential unexpected consequences.”Barto, retired for 14 years, describes himself as a Luddite, while Sutton is embracing a future he expects to have beings of greater intelligence than current humansan idea sometimes known as posthumanism.“People are machines. They’re amazing, wonderful machines,” but they are also not the “end product” and could work better, Sutton said.“It’s intrinsically a part of the AI enterprise,” Sutton said. “We’re trying to understand ourselves and, of course, to make things that can work even better. Maybe to become such things.” Matt O’Brien, AP Technology Writer
Category:
E-Commerce
Before Reddit there was Digg, which popularized up- and down-votes on online posts. Now the founders of both platformssocial media veterans Kevin Rose and Alexis Ohanianare relaunching the early Reddit rival with a focus on “humanity and connection” they hope will be boosted by the use of artificial intelligence.Rose founded Digg, which launched in 2004 and let people up- and down-vote (“Digg” or “bury”) content from users and from sources around the web. At its peak, it had 40 million monthly usersa high number for the time considering that Facebook only hit 100 million in 2008.Digg was divvied up and sold in 2012, with many of its assets and patents acquired by LinkedIn. Reddit, which launched in 2005 and was cofounded by Ohanian, took a similar approach to let users vote on what they thought was the best and worst content on the site.But much has changed since 2012not just when it comes to advances in artificial intelligence but also how people treat each other online.“The social space online is definitely harsher, it feels like, than it’s ever been before,” said Justin Mezzell, who will serve as the new company’s CEO. “It feels really difficult to connect. I think the platforms have gotten more disconnected. You know, if ever there was a true town hall of the internet, it feels like it has been deconstructed in a pretty big way.”Digg’s new leaders say they want to use artificial intelligence to “handle the grunt work” of running a social media site while allowing humans to focus on building meaningful online communities. The question, Mezzell said, is how to get people to “show up and have conversations, to learn from each other, to share something they’re passionate about and do it earnestly?” Especially when some of today’s social media algorithms “exist really just optimize for outrage.”Rose said Digg will take a more nuanced approach to content moderation than banning or not banning content, which is a process that can be easy to get around.“There is a world where, you know, you show up in (a) meditation (group) and you’re swinging four-letter words all over the place, and you hit submit,” he said. And “we come back and we say, hey, you can post this, of course, but only 2% of the audience is going to see it, because the way that the moderator set the tone.”“That is unique. That is different. That’s not like a hard-defining rule,” Rose added “It’s more like just sensing the voice and how it fits within the entire ecosystem and the model that’s behind the scenes for that community.”The new Digg will launch in the coming weeks as a website and mobile app. Barbara Ortutay, AP Technology Writer
Category:
E-Commerce
Forty-three days after taking office for the second time, President Donald Trump delivered a joint address to Congress on Tuesday, in the same chamber that an angry mob of his supporters ransacked four years earlier in an attempt to overthrow the government. Like most Trump speeches of late, this one was a lengthy, rambling affair that clocked in as the longest-ever joint address to Congress by a healthy margin. Many Democratic lawmakers elected not to attend at all, and several who did show left well before Trump wrapped for the night. Apparently, one can only spend so much off-the-clock time in the same room as a euphoric, seal-clapping Lauren Boebert before deciding to try and beat traffic instead. The speechs substance will be familiar to anyone who has seen clips of a Trump rally over the past 10 years: a jumble of unhinged culture-war screeds and inscrutable conspiracy theories, sprinkled with the occasional gesture toward making America great that prompts the sycophants to pop out of their chairs like reactionary jack-in-the-boxes. But to a greater extent than most joint addresses to Congress, which newly elected presidents typically use to preview their loftiest aspirations, this felt more like a victory lap from a lame-duck president who sees his victory as a license to plunder the country as much as the law allows, and sometimes beyond it. If Tuesdays agenda is any indication, for the next four years, Trumps plan for governing is to make every decision based on how much he thinks he and his cronies stand to profit from it. Trump of course spent a considerable amount of airtime touting the accomplishments of the Department of Government Efficiency, Elon Musks ambitious project to make the entire federal government as buggy and nonfunctional as Twitter became shortly after he purchased the site. DOGE has already uncovered hundreds of billions of dollars of fraud, said Trump, who rattled off examples of foreign aid expenditures he wants to scuttle in the jeering cadence of a comedian who knows his audience does not need to hear the punchline to understand the racist joke. $8 million to promote LGBTQI+ in the African nation of Lesotho, which nobody has ever heard of, he said at one point, soaking in the laughter that followed. Later, Trump highlighted his and Musks ongoing efforts to gut the federal agencies charged with implementing the laws Congress passes. For nearly 100 years, the federal bureaucracy has grown until it has crushed our freedoms, ballooned our deficits, and held back Americas potential in every possible way. he said. The nation founded by pioneers and risk-takers now drowns under millions and millions of pages of regulations and debt. In order to unshackle our economy, Trump promised that under his leadership, the executive branch would eliminate ten existing regulations for every new one it creates, building on his first-term record of ending unnecessary rules and regulations like no other president had done before. Set aside, for a moment, the fact that seemingly every time Musk and the DOGE teens announce some new source of cost savings, their estimates turn out to be wrong by an order of magnitude at least. Grousing about purportedly frivolous expenditures and onerous regulations are time-honored traditions among wealthy conservatives, whose definition of wasteful government spending includes all government spending that does not redound directly to their benefit. Musk and Trump want to cut foreign aid because they want the government to do fewer things that require their tax dollars, and know that in a Republican Party animated by bigotry and xenophobia, humanitarian assistance for developing countries makes for an easy political target. And by kneecapping agencies ability to do the day-to-day work of governing, Trump and Musk would ensure that deep-pocketed corporations relentlessly chasing shareholder value are free to abuse and exploit consumers without fear of meaningful consequences. The balance of Trumps speech continued in this same vein: He framed his second-term tax agenda as offering cuts for everybody, which glosses over the fact that, according to a Wharton School analysis, the top 10% of earners would receive about 56% of the proposed cuts’ value. He reiterated his pledge to take back the Panama Canal, presumably to the delight of billionaire investor Larry Fink, whose firm, BlackRock, just bought key ports on either side of it. When discussing tariffs on Canada, Mexico, China, and probably more countries to be named soon, Trump promised to take in trillions and trillions of dollars and create jobs like we have never seen before, ignoring the reality that the real-world burdens will fall first on farmers who cant sell crops and working people paying more for cars, cell phones, and t-shirts. In the two days after Trump announced that his tariffs would take effect, the Dow fell some 1300 points, which Trump characterized as a little disturbancebut, he added, Were okay with that. I am not sure the millions of normal people whose retirement savings the president is staking on a harebrained trade war are quite as sanguine. Last August, Trump held a press conference surrounded by foodstuffs in which he promised to immediately bring down pricesstarting on day one, he added as if to clear up any ambiguity. Yet his speech only occasionally referenced what some two-thirds of voters describe as a very big problem; when he did bring it up, it was mostly by framing DOGEs scorched-earth approach to governance as a cure-all for everything ailing the country, from spiking egg prices to the high cost of vehicle financing. By slashing all of the fraud, waste, and theft we can find, we will defeat inflation, bring down mortgage rates, lower car payments and grocery prices, protect our seniors, and put more money in the pockets of American families, he said after a lengthy riff about alleged rampant Social Security fraud. Again, for everyone whose weekly bills have not plummeted since Trump took office, I do not think just trust Elon Musk will be an especially persuasive message. Perhaps the most oafishly venal policy Trump discussed on Tuesday was his idea for a “gold card,” which would extend green card-style privileges and an easy path to citizenship to foreign nationals willing to pay a $5 million fee. We will allow the most successful job-creating people from all over the world to buy a path to U.S. citizenship, he said, promising that the cards would go on sale soon, as if he were a late-night TV pitchman trying to get you to buy a wearale blanket with cat ears affixed to the hood. For all the scorn that Trump displays for immigrants fleeing violence and poverty in their home countries, he is happy to extend the benefit of the doubt to anyone with the means to write a seven-figure check. Trump has never had any real interest in governing; like everything else hes done in his career, his decision to seek the GOP nomination in 2016 was mostly an elaborate branding exercise that succeeded beyond his wildest dreams when he accidentally won 304 electoral votes. Ten years later, he is (presumably) winding down his political career by running an even more transparent version of the same playbook, scrounging up every last opportunity to reshape American society in ways that will make wealthy people like him even wealthier. For Trump, it does not matter how many others get hurt in the process, because enriching himself is one of the privileges he enjoys as president. If it werent, why would anyone want the job in the first place?
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||