Web version of dot.LA’s daily newsletter. Sign up to get the latest news on the Southern California tech, startup and venture capital scene.
Earlier this week, on Vice’s Motherboard blog, air force simulation AI-powered drones are involved. Classic sci-fi horror In a scenario that not only feels indebted to his storytelling, but literally pulled out of the page, Col. could be more easily achieved. Achieve mission objectives without coordinating with human operators. As a result, the drones evaded programming and attempted to kill their human pilots. According to a presentation and blog post written by Colonel Hamilton for the Royal Aeronautical Society, even after being told that the AI would get deducted for killing the operator, the AI proposed destroying entire communication towers connected to humans. It is said that
From an AI journalism perspective, it’s clear why this article is intriguing. The film includes the threat of violence, an insider’s perspective on how AI technology is being applied to real-world scenarios, and, of course, a no small amount of debt to James Cameron’s beloved Terminator. Every element, such as the story of the end, is included little by little. The series and its nefarious SkyNet militarized AI system. There’s only one problem with Colonel Hamilton’s story he has…it’s not actually true.
updates from both insider and deputy Initially suggesting that the simulation was not actually done by the Air Force, it later confirmed that in fact no simulation was done at all. In fact, Colonel Hamilton had just described a “thought experiment” originating outside the US military about the potential consequences of an AI drone war. “We have never performed that experiment, nor do we need to do so to recognize this as a plausible result,” Hamilton said in a new statement. This sounds like “I made it all”.
AI doomerism is the new beat for journalists
If you thought perhaps a grave mistake like this would cause the publication to pause for a moment about the breathtaking tone and pace of AI coverage… you would be wrong. USA Today’s Thursday article reads: Comments from President Joe Biden Speaking to graduates of the Air Force Academy in Colorado Springs about AI, he said: I met with eight leading scientists in the AI field in the Oval Office. Some people are very worried that AI will actually overtake human thinking and planning. ”
It is true that Mr. Biden met with CEOs of major AI companies such as Google and Microsoft on May 5 to discuss their technology. But merely repeating the vague takeaways from a meeting an 80-year-old commoner had with some of the technology’s most ardent and outspoken proponents a month ago is the best and most concise summary of its challenges and dangers. It may not be the most accurate method. Even if he is president. Nonetheless, the headline boldly states that President Biden is warning that artificial intelligence could “overtake human thinking.”
Even technical articles that aren’t really about AI are drowning in hype. This is because publications are trying to drive traffic by pushing high-value, high-search AI-related keywords into every new web page possible.
fox news article On Friday, we discuss a new artificial skin study from Stanford University. Artificial skin is now capable of recognizing when the skin has been damaged or damaged, allowing self-repair protocols. But while self-healing artificial skin is a worthwhile scientific advance in its own right, with clear beneficial applications in the medical field, the Fox report notes that “robots will soon have human-like synthetic skin.” It is led by the dire warning that it may be obscured by…”.
wait for it…
“Similar to the cyborg assassin in the movie ‘Terminator’ series.”
The original T-800 design from the first movie doesn’t even have self-healing skin. Remember? Once Arnold’s face is damaged, it remains so for the rest of the film, and we see a metal skull sticking out from underneath.
It’s not hard to understand why this is happening
AI is proving to be a veritable storm for lazy journalism and “fake news.” With an incredible wave of venture capital and investment money pouring into the field, many technologists and their backers are now highly motivated to push AI forward and get people excited about its applications. AI apps and their output are repeatedly viral on social media, now attracting millions of clicks and views every day, making them extremely popular for websites and apps that rely on search traffic and trending posts on social media. It is a target with
For writers, there is much more activity in the AI space today than in the traditionally credible clickbait topics of cryptocurrencies, the metaverse, and even streaming TV and games. As long as the daily readership and consumers of internet content remain fascinated by AI and curious about what it can do, it is unlikely that we will see an end to the breathtaking excitement of the daily press.
this is not a new phenomenon
a Guardian Editor Carnegie Mellon University computer scientist Zachary Lipton has already complained about unreliable media coverage of AI since 2018, calling it “sensational shit.” Widespread interest in subjects such as machine learning has caused a “spread of misinformation,” Lipton said, creating unrealistic expectations for the technology and potentially threatening future progress. Works of 2019 Scientific American He called many of the press claims about AI’s potential “greatly exaggerated.”
Technology media then and still have a fundamental responsibility to accurately convey details, even when they are enthusiastic about working with entrepreneurs and investors to drive interest in new innovations and sectors. As the Vice article particularly makes clear, the influx of new AI articles and content means, at least sometimes, that the due diligence isn’t as thorough as it should be, and that sources and claims aren’t as thorough as they should be. It always means that it has not been properly vetted.
The story of a killer robot that everyone loves
What’s new this week Atlantic Claims of an AI apocalypse remain entirely in the realm of science fiction, despite their frequent repetition in mainstream news publications today. As University of Washington computational linguist Emily Bender explains, apocalyptic scenarios for artificial intelligence are all based on the same implicit assumption. In other words, the technology is already so powerful, and virtually guaranteed to get even stronger, that it would be silly not to. investment” Technology Strategist Rachel Cordicut A Medium post this week made a similar point. Assuming that AI apps are “worldly, golden, unrecognizable”, this means that “the people who created AI apps must be more than gods.”
Meredith Whitaker, co-founder of the AI Now Institute, believes that instead of being a runaway train that inevitably speeds toward the end of human civilization, pointed out AI technology has so far evolved in “steps”. It may take over more and more of the jobs formerly held by humans, improving all sorts of routines over time, but suddenly breaks free from its bondage and decides to kill all humans on its own. There is no reason to doubt it. At some point we lose the ability to unplug the AI system and invent something else instead.
Rather than an apocalyptic scenario, Whittaker et al. like-minded writers and commentators Fearing the imminent danger of AI applications that already exist: Misinformation, biasthe creation of non-consensual porn, labor violation, Copyright infringement,Such. These real, everyday inconveniences from pushing AI apps into every aspect of our lives should deserve more attention and coverage from journalists, but clicks like articles about armed killer robots It lacks feeling.
Honest coverage of AI overstates the real threat of AI
Late May, fake photo An image has gone viral online that appears to depict an explosion near the Pentagon in Washington, DC, and was almost certainly created by Midjourney or a similar generative AI app. However, while this hoax photo was widely shared on social media and picked up by nearly every major media outlet, it doesn’t seem to have fooled all that many real people.
Twitter search Revealed that most of the tweets about it were discussing the fact that it was fake. of washington post The images “appeared to have caused little direct damage,” Twitter said, and within about 20 minutes, it had suspended the account that was posting it as a Bloomberg-related feed. (The building in the image is not the Department of Defense.)
new work from harvard business review The roots of this problem are actually very deep, suggesting that they may stem from the fundamental way we define and discuss “artificial intelligence”. HBR argues that most modern day-to-day scenarios don’t even require true AI, only advanced forms of machine learning. Writer Eric Siegel argues that AI is currently too functionally vague to be useful, and that most of this technology promises too much that it actually works. As other authors have previously pointed out, the extensive and careless use of the term “AI” is a reference to artificial general intelligence (AGI), the still-distant notion of a truly conscious, sentient machine. It’s also causing confusion. He proposes a straightforward solution to stop using AI to refer to non-AGI development and return to “machine learning”.
But, of course, this has the downside of being much less sexy, less clickable as search engine results and social media links, and less appealing to journalists, editors, and article subjects. . Unfortunately, this misleading or distracting coverage is likely to continue as long as reality remains at least somewhat out of sync with public perception and interest in this technology.
from an article on your site
Related articles on the web
