Steven J. Vaughan-Nichols writes in an op-ed for The Register: I use AI a lot, but not to write stories. I use AI for search. When it comes to search, AI, especially Perplexity, is simply better than Google. Ordinary search has gone to the dogs. Maybe as Google goes gaga for AI, its search engine will get better again, but I doubt it. In just the last few months, I've noticed that AI-enabled search, too, has been getting crappier. In particular, I'm finding that when I search for hard data such as market-share statistics or other business numbers, the results often come from bad sources. Instead of stats from 10-Ks, the US Securities and Exchange Commission's (SEC) mandated annual business financial reports for public companies, I get numbers from sites purporting to be summaries of business reports. These bear some resemblance to reality, but they're never quite right. If I specify I want only 10-K results, it works. If I just ask for financial results, the answers get... interesting. This isn't just Perplexity. I've done the exact same searches on all the major AI search bots, and they all give me "questionable" results. Welcome to Garbage In/Garbage Out (GIGO). Formally, in AI circles, this is known as AI model collapse. In an AI model collapse, AI systems, which are trained on their own outputs, gradually lose accuracy, diversity, and reliability. This occurs because errors compound across successive model generations, leading to distorted data distributions and "irreversible defects" in performance. The final result? A Nature 2024 paper stated, "The model becomes poisoned with its own projection of reality." [...] We're going to invest more and more in AI, right up to the point that model collapse hits hard and AI answers are so bad even a brain-dead CEO can't ignore it. How long will it take? I think it's already happening, but so far, I seem to be the only one calling it. Still, if we believe OpenAI's leader and cheerleader, Sam Altman, who tweeted in February 2024 that "OpenAI now generates about 100 billion words per day," and we presume many of those words end up online, it won't take long.

Read more of this story at Slashdot.

Comments

In an interview with Wired (paywalled), OnePlus co-founder and Nothing CEO, Carl Pei, said the future of smartphones will center around the OS and AI to get things done -- rendering traditional apps a thing of the past. 9to5Google reports: Pei says that Nothing's strength is in "creativity," adding that "the creative companies of the past" such as Apple "have become very big and very corporate, and they're no longer very creative." He then dives into what else but AI, explaining that Nothing wants to create the "iPod" of AI, saying that Apple built a product that simply built a better user experience: "If you look back, the iPod was not launched as 'an MP3 player with a hard disk drive.' The hard disk drive was merely a means to a better user experience. AI is just a new technology that enables us to create better products for users. So, our strategy is not to make big claims that AI is going to change the world and revolutionize smartphones. For us, it's about using it to solve a consumer problem, not to tell a big story. We want the product to be the story." Pei then says that he doesn't see the current trend of AI products -- citing wearables such as smart glasses -- as the future of the technology. Rather, he sees the smartphone as the most important device for AI "for the foreseeable future," but as one that will "change dramatically." According to Pei, the future of the smartphone is one without apps, with the experience instead just revolving around the OS and what it can do and how it can "optimize" for the user, acting as a proactive, automated agent and that, in the end, the user "will spend less time doing boring things and more time on what they care about."

Read more of this story at Slashdot.

Comments

Comments

Comments

Comments

An anonymous reader quotes a report from Axios: Advanced robots don't necessarily need to look like C3PO from "Star Wars" or George Jetson's maid Rosie, despite all the hype over humanoids from Wall Street and Big Tech. In fact, some of the biggest skeptics about human-shaped robots come from within the robotics industry itself. [...] The most productive -- and profitable -- bots are the ones that can do single tasks cheaply and efficiently. "If you look at where robots are really bringing value in a manufacturing environment, it is combining industrial or collaborative robots with mobility," ABB managing director Ali Raja tells Axios. "I don't see that there are any real practical applications where humanoids are bringing in a lot of value." "The reason we have two legs is because whether Darwin or God or whoever made us, we have to figure out how to traverse an infinite number of things," like climbing a mountain or riding a bike, explains Michael Cicco, CEO of Fanuc America Corp. "When you get into the factory, even if it's a million things, it's still a finite number of things that you need to do." Human-shaped robots are over-engineered solutions to most factory chores that could be better solved by putting a robot arm on a wheeled base, he said. "The thing about humanoids is not that it's a human factor. It's that it's more dynamically stable," counters Melonee Wise, chief product officer at Agility Robotics, which is developing a humanoid robot called Digit. When humans grab something heavy, they can shift their weight for better balance. The same is true for a humanoid, she said. Using a robotic arm on a mobile base to pick up something heavy, "it's like I'm a little teapot and you become very unstable," she said, bending at the waist.

Read more of this story at Slashdot.

Comments

SpaceX's Starship Flight 9 successfully launched and reached space -- marking the first reuse of a Super Heavy booster -- but both rocket stages were ultimately lost mid-mission due to a "rapid unscheduled disassembly." SpaceX CEO Elon Musk said in a statement: "Starship made it to the scheduled ship engine cutoff, so big improvement over last flight! Leaks caused loss of main tank pressure during the coast and re-entry phase. Lot of good data to review." Musk said the next three Starship test launches could lift off every three to four weeks in the days ahead. Space.com reports: The mission lifted off from Starbase today at 7:37 p.m. EDT (2337 GMT; 6:37 p.m. local Texas time), sending the 40-story-tall rocket into the Texas sky atop a pillar of flame. It was a milestone launch, marking the first-ever reuse of a Super Heavy booster; this one earned its wings on Flight 7 in January. (SpaceX swapped out just four of its Raptors after that mission, meaning that 29 of the engines that flew today were flight-proven.) "Lessons learned from the first booster refurbishment and subsequent performance in flight will enable faster turnarounds of future reflights as progress is made towards vehicles requiring no hands-on maintenance between launches," the company wrote in a Flight 9 mission preview. The Super Heavy had a somewhat different job to do today; it conducted a variety of experiments on its way back down to Earth. For example, the booster performed a controlled rather than randomized return flip and hit the atmosphere at a different angle. "By increasing the amount of atmospheric drag on the vehicle, a higher angle of attack can result in a lower descent speed, which in turn requires less propellant for the initial landing burn," SpaceX wrote in the mission preview. "Getting real-world data on how the booster is able to control its flight at this higher angle of attack will contribute to improved performance on future vehicles, including the next generation of Super Heavy." These experiments complicated Super Heavy's flight profile compared to previous missions, making another "chopsticks" catch at Starbase a tougher proposition. So, rather than risk damaging the launch tower and other infrastructure, SpaceX decided to bring the booster back for a "hard splashdown" in the Gulf of Mexico on Flight 9. That was the plan, anyway; Super Heavy didn't quite make it that far. The booster broke apart about 6 minutes and 20 seconds into today's flight, just after beginning its landing burn. "Confirmation that the booster did demise," [Dan Huot, of SpaceX's communications team] said during the Flight 9 webcast. Super Heavy's flight ended "before it was able to get through landing burn," he added. Ship, by contrast, improved its performance a bit this time around. It reached space today on a suborbital trajectory that took it eastward over the Atlantic Ocean -- the same basic path the vehicle took on the truncated Flight 7 and Flight 8. But Flight 9 got choppy for Ship after that. The vehicle was supposed to deploy eight dummy versions of SpaceX's Starlink satellites about 18.5 minutes after liftoff, which would have been a landmark first for the Starship program. That didn't happen, however; the payload door couldn't open fully, so SpaceX abandoned the deployment try. Then, about 30 minutes after launch, Ship started to tumble, which was the result of a leak in Ship's fuel-tank systems, according to Huot. "A lot of those [tanks] are used for your attitude control," he said. "And so, at this point, we've essentially lost our attitude control with Starship." As a result, SpaceX nixed a plan to relight one of Ship's Raptor engines in space, a test that was supposed to happen about 38 minutes after launch. And the company gave up hope of a soft splashdown for the vehicle, instead becoming resigned to a breakup over the Indian Ocean during Ship's reentry. The company therefore will not get all the data it wanted about Flight 9. And there was quite a bit to get; for example, SpaceX removed some of Ship's heat-shield tiles to stress-test vulnerable areas, and it also tried out several different tile materials, including one with an active cooling system. But the company plans to bounce back and try again soon, just as it did after Flight 7 and Flight 8. You can watch a recording of the launch on YouTube.

Read more of this story at Slashdot.