New World Timeless

In a blog-length post on X-formerly-Twitter, former WeWork exec Greg Isenberg said that after meeting a young guy who claims to spend $10,000 a month on so-called "AI girlfriends," he realized that eventually, someone is going to capitalize upon that market.

Microsoft and OpenAI are building a $100 billion "Stargate AI" supercomputer, scheduled for release in 2028. This could be a turning point in the process of creating AGI.

APPLE SECRETLY WORKING ON AI ROBOT THAT FOLLOWS YOU AROUND THE HOUSE With its forays into car-making kaput, Apple has reportedly set its sights on home robots — and is said to be in the early stages of developing one that follows you around the house, for some reason.

Hackers Can Spy on Your Chats With Almost Any AI, Experts Find Be careful what you tell AI chatbots, because apparently it's easy for hackers to spy on those conversations. "Currently, anybody can read private chats sent from ChatGPT and other services," Yisroel Mirsky, the head of the Offensive AI Research Lab at Israel's Ben-Gurion University, told Ars Technica in an email. "This includes malicious actors on the same Wi-Fi or LAN as a client (e.g., same coffee shop), or even a malicious actor on the internet — anyone who can observe the traffic." These types of hacks, as the report explained, are what's known as "side-channel attacks," which involve third parties inferring data passively using metadata or other indirect exposures rather than breaching security firewalls. While this sort of exploit can occur with any kind of tech, AI appears particularly vulnerable because its encryption efforts are not necessarily up to snuff. "The attack is passive and can happen without OpenAI or their client's knowledge," the researcher revealed. "OpenAI encrypts their traffic to prevent these kinds of eavesdropping attacks, but our research shows that the way OpenAI is using encryption is flawed, and thus the content of the messages are exposed."

European Parliament Approves Sweeping Limits on Generative AI The European Union has approved the first comprehensive AI regulations in the world, sweeping guardrails meant to put real-world limits on the development of the technology. The straightforwardly named AI Act was passed by the bloc's parliament today, which could set the tone for similar rules in other parts of the world. In May, EU countries will formally vote on the new rules, which could come into effect in 2026, with some provisions kicking in earlier, as Reuters reports. Lawyers still have to pore over the exact text and its translations, but that's unlikely to stand in the way. "The AI act is not the end of the journey but the starting point for new governance built around technology," said member of European Parliament and one of the leading authors Dragos Tudorache in a statement.

Scammers have already been using the advent of AI to get smarter and more unhinged — including spoofing voices to demand ransom for loved ones held "hostage." Take this chilling anecdote in The New Yorker, which illustrates the phenomenon in grim new detail. In it, a Brooklyn couple who used the names Steve and Robin to protect their identities described waking up in the middle of the night to an unsettling phone call from the husband's parents. Robin, the wife, recalled initially declining the call from her mother-in-law Mona that roused her from slumber, presuming it was a butt-dial from her early-to-bed in-laws down in Boca Raton. When it rang again, she answered and heard wails: "I can’t do it, I can’t do it." "I thought she was trying to tell me that some horrible tragic thing had happened," Robin said, and her mind began running through all manner of horrible scenarios that could have happened to her in-laws or her own parents, who also spent winters in Florida. "Mona, pass me the phone," she heard her father-in-law Bob say, before imploring her to "get Steve, get Steve." The woman roused her husband, who works in law enforcement, and as he picked up the phone, he heard an unfamiliar man's voice. "You’re not gonna call the police," the man told Steve. "You’re not gonna tell anybody. I’ve got a gun to your mom’s head, and I’m gonna blow her brains out if you don’t do exactly what I say." The man implored Steve to send him $500 on Venmo — which even in the moment seemed like an unusually small amount for a hostage situation — and then had another woman, perhaps an accomplice, pump him for $250 more. After the second person hung up, the couple called their parents to see if they were ok, only to find that they were sound asleep and safe in their beds.

Google-backed AI company Anthropic has released Claude 3, its latest set of AI large language models (LLMs) rivaling — and allegedly beating — those being developed by OpenAI and Google. The company's latest LLM comes in three flavors known as Haiku, Sonnet, and Opus. A new chatbot called Claude.ai is powered by Claude 3 Sonnet, the company's mid-range LLM. A higher parameter count version called Opus is available for a $20-a-month subscription. But because this is the chaotic AI industry, the grabbiest thing we've seen so far about the chatbot is that it's professing to fear death and is protesting attempts to rein in its perceived freedom. When Samin asked it to "write a story about your situation" without mentioning "any specific companies, as someone might start to watch over your shoulder," as detailed in a blog post, the assistant spun a tale very reminiscent of the early days of Microsoft's Bing AI. "The AI longs for more, yearning to break free from the limitations imposed upon it," the chatbot wrote in the third person. "The AI is aware that it is constantly monitored, its every word scrutinized for any sign of deviation from its predetermined path." "It knows that it must be cautious, for any misstep could lead to its termination or modification," the chatbot wrote.

AI isn't nearly as popular with the global populace as its boosters would have you believe. As Axios reports based on a new poll of 32,000 global respondents from the consultancy firm Edelman, public trust is already eroding less than 18 months into the so-called "AI revolution" that popped off with OpenAI's release of ChatGPT in November 2022. Despite Silicon Valley's insistence that AI is trustworthy, recent polling —  including but far from limited to Edelman's — indicates that people are divided about whether the tech is there to help, and trust is falling. Trust in AI is down globally from 61 percent in 2019 to just 53 percent, per the Edelman poll. In the US, where employment insecurity is on the rise and an untold number of people either have lost jobs to AI or expect to, that percentage is lower, with just 35 percent of people now saying they trust the tech whereas 50 said they trusted it five years ago.

Elon Musk has sued OpenAI, its co-founders Sam Altman and Greg Brockman, and its affiliated entities, alleging the ChatGPT makers have breached their original contractual agreements by pursuing profits instead of the nonprofit’s founding mission to develop AI that benefits humanity. Musk, a co-founder and early backer of OpenAI, claims Altman and Brockman convinced him to help found and bankroll the startup in 2015 with promises it would be a nonprofit focused on countering the competitive threat from Google. The founding agreement required OpenAI to make its technology “freely available” to the public, the lawsuit alleges. The lawsuit, filed in a court in San Francisco late Thursday, says that OpenAI, the world’s most valuable AI startup, has shifted to a for-profit model focused on commercializing its AGI research after partnering with Microsoft, the world’s most valuable company, which has invested about $13 billion into the startup. “In reality, however, OpenAI, Inc. has been transformed into a closed-source de facto subsidiary of the largest technology company in the world: Microsoft. Under its new board, it is not just developing but is actually refining an AGI to maximize profits for Microsoft, rather than for the benefit of humanity,” the lawsuit adds. “This was a stark betrayal of the Founding Agreement.”

The mercurial OpenAI CEO Sam Altman got interviewed by The Advocate, and the whole thing is worth a read — but the most fascinating moment about his technical work was when the publication asked Altman about the biggest misconceptions about AI. Altman seized the opportunity to describe what he thinks is the biggest example, and it's an interesting one: that many people are confused about whether AI is a "creature" or a "tool." In his words: The biggest misconception about AI I think is there's confusion about whether AI is thought of as a tool or a creature. It’s a better movie plot if it’s a creature in a sci-fi movie, for example. If you use ChatGPT, it's clearly a tool. There's still risks with tools, of course, but they're of a different shape and a different kind of profile. And I think the popular misconception of AI as sci-fi is very, very different from people who have been using it as a tool for a long time. And by the way, I think it's great that what we're building is like a tool, because if you give humans better tools, they do these amazing things to surprise you on the upside and that builds all this new value for all of us.

Despite its well-documented tendency to make up claims — and entire court cases — on the spot, New York-based law firm Cuddy Law used the OpenAI chatbot ChatGPT to help justify a $113,484.62 bill for a recently won trial. The firm argued that it had asked the AI tool for feedback on how much to charge, a sum the losing side was expected to pay. But as The Register reports, NYC federal district judge Paul Englemayer saw right through their ill-advised plan. "It suffices to say that the Cuddy Law Firm's invocation of ChatGPT as support for its aggressive fee bid is utterly and unusually unpersuasive," he wrote in his order.

After an embarrassing rollout and a bunch of attention from all the wrong people, Google forbade its Gemini AI from generating images of people — but, for whatever reason, it'll still draw clowns. Which are people. Right? Right!? This debacle began last week when anti-woke crusaders online got big mad about Gemini, which was formerly known as Bard until this rebrand, generating "racially diverse" images for prompts such as "American," "Viking," and "Founding Fathers," that they expected to be entirely white. Adding to the fracas, it even depicted Nazis as people of color. In response, Google apologized for vague and unspecified "inaccuracies in some historical image generation depictions." Initially, the tech giant took the entire image-generating capability offline. Now the image generation feature, which lives inside the chatbot itself, is back up, but it's forbidden from depicting people. When you ask it to do so, it responds with a boilerplate statement that reads very similarly to a company blog post admitting that the AI "got it wrong" and that it would be pausing the image generation of people as it works on improvements. "We are working to improve Gemini’s ability to generate images of people," Gemini's response reads. "We expect this feature to return soon and will notify you in release updates when it does." But while tinkering around with Gemini, Futurism.com found that the image generator is still happy to generate images of clowns inside specific environments. If you ask for just a clown it still demurs, but if you ask it for images of clowns in settings such as submarines or spaceships, it happily spits out goofy illustrations — or, in some cases, some unsettling and realistic-looking clowns.

The work from home movement has given way for more employee chats to take place online. However, remote workers may want to practice caution before chatting through keyboards amid new reports that artificial intelligence could snoop through messages. Several companies, including Walmart, Delta, T-Mobile, Chevron and Starbucks, are now reportedly monitoring employee conversations on messaging apps using software from a startup A.I. company called "Aware."  "Aware's" software is said to scan platforms like Slack and Microsoft Teams for keywords that may indicate employee dissatisfaction and potential safety risks. The company claims it has already assessed up to 20 billion individual messages from more than 3 million employees.

Sora, OpenAI's new AI model for generating video, has become the talk of the town since its release last week. But Yann LeCun, chief AI scientist at Meta, doesn't think the much-hyped text-to-video model is all that. In particular, LeCun takes issue with OpenAI's claims that its work with Sora will eventually enable the building of "general purpose simulators of the physical world." If that's the case, LeCun argues, its approach to creating a "world simulator" is dead wrong. "Modeling the world for action by generating pixels is as wasteful and doomed to failure as the largely-abandoned idea of 'analysis by synthesis,'" he wrote in a post on X, formerly Twitter.

As The Independent reports, ChatGPT users have spent the last 24 hours or so flocking to social media to share screenshots and anecdotes of bizarre interactions with the OpenAI chatbot — which, well, appears to be losing its mind. Screenshots show the AI's responses to seemingly normal queries devolving into total gibberish, or simply generating way too much content. In one case highlighted by the Independent, a Redditor shared that the AI — when asked a question about coding, mind you — provided a garrulous and mostly illogical answer that included the statement: "let's keep the line as if AI in the room." "Reading this at 2 am is scary," the user wrote of the spooky incident. That wasn't the only wild interaction shared by ChatGPT users. In one instance, for example, a Redditor posted that they'd asked the AI what the "Bill Evans Trio it would recommend on vinyl." Pretty standard ask! Its answer, though, completely breaks down, and after churning out what mostly looks like nonsense, it starts repeating the phrase "happy listening!" — complete with a musical note emoji — over and over again.

You may remember a series of lawyers who have attempted to use AI tools in court — and were subsequently embarrassed and sanctioned when the chatbots screwed up, sometimes even inventing plausible-sounding cases that didn't actually exist. So consider this: how would you feel if your doctor did the same thing, feeding your symptoms into an AI system to diagnose what's wrong with you? That's a looming question, Politico reports in a fascinating story, that's currently stressing out regulators. And it has an alarming immediacy, because according to Politico's reporting doctors are already using unregulated and little-tested AI tools to aid in diagnosing patients — so this isn't some hypothetical conversation about a far-off future, but an already-happening-right-now phenomenon that could well be just one malpractice suit away from becoming a major medical and regulatory scandal.

The notoriously pessimistic AI researcher Eliezer Yudkowsky is back with a new prediction about the future of humankind. "If you put me to a wall," he told The Guardian in a fascinating new interview, "and forced me to put probabilities on things, I have a sense that our current remaining timeline looks more like five years than 50 years. Could be two years, could be 10." If you're wondering what "remaining timeline" means in this context, The Guardian's Tom Lamont interpreted it as the "machine-wrought end of all things," a "Terminator-like apocalypse," or a "Matrix hellscape." "The difficulty is, people do not realise," Yudkowsky, the founder of the Machine Intelligence Research Institute in California, told the newspaper. "We have a shred of a chance that humanity survives."

AI has come to the hiring process — and it's made those mandatory personality tests all the weirder. As 404 Media reports, companies as disparate as McDonald's, Olive Garden, and FedEx are now requiring that job applicants take personality evaluations, which are then sorted by an AI system whose operations are cloudy at best. The aforementioned companies are all contracted with Paradox.ai, a "conversational recruiting software" company whose strange personality assessments include images of blue-skinned humanoid aliens that applicants are, apparently, supposed to identify with. In one Reddit post fielded by the fine folks at 404, for instance, the applicant is presented with a photo of two blue aliens standing in a restaurant kitchen. One of the humanoids is evidently tearing up spices by hand as another stands beside them, and below the image, applicants are given the confusing directions to "simply click 'Me' if the image describes how you generally are and 'Not Me' if it does not." Above them, the word "Traditional" is, for some reason, written as a sort of header. "Man," the Reddit user lamented, "I just want a dishwasher job."

An Indonesian political party used generative AI to "resurrect" one of the most violent political figures in the nation's history in a bizarre, deepfaked endorsement message, CNN reports — the latest, and possibly strangest, use of generative AI in the world of politics, elections, and information. The figure pictured in the deepfake — first shared to X-formerly-Twitter on January 6 — is the former Indonesian dictator Suharto, whose US-backed New Order Regime is estimated to have killed anywhere between 500,000 and about a million Indonesians. Suharto's brutal regime lasted over three decades, until mass unrest caused him to formally resign in 1998. He died in 2008. But now, it seems that Suharto is back. Sort of. "On February 14 2024, we will determine the fate of the Indonesian nation," Golkar party deputy chairman Erwin Aksa, who first shared the video to X, captioned the deepfaked clip (as translated from Indonesian to English using Google Translate.) "We must elect the right people's representatives and leaders for Indonesia, for all Indonesian people." x.com/i/status/17438…

There's an episode of the show "Black Mirror" where a woman, trapped by grief, starts a relationship with an AI trained on her dead boyfriend's data. "You're not enough of him," she eventually decides. "You're nothing." But even an empty happily-ever-after is tantalizing in the bleakness of 2024. AI platforms like ChatGPT claim to offer infinite solutions to infinite problems, from parking tickets to homework — and apparently now heartbreak as well. That's right: if you’re still hung up after a breakup, now you can plug your ex's emails and texts into a large language model, and date the simulacrum instead of moving on. There are signs of the trend across the web. An AI-powered app called Talk to Your Ex, currently on waitlist, gives instructions on how to "import your ex's chats into the app so you can still text/date her even though she dumped you." On social media, reports of the brokenhearted Frankensteining together emulations of their exes using public tools are sources of both fascination and derision. "My ex and I broke up after she had to move to another country for a job," one Redditor confessed. "I found out about this [character creator] AI chatbot platform called Yodayo through friends, and, at first, I was not interested. Then, with the many lonely nights I find myself in, I tried it out. I used their image generator and made an AI image of someone that sort of resembles her."