Artificial intelligence is more a part of our lives than ever before. While some might call it hype and compare it to NFTs or 3D TVs, generative AI is causing a sea change in nearly every part of the technology industry. OpenAI’s ChatGPT is still the best-known AI chatbot around, but with Google pushing Gemini, Microsoft building Copilot, and Apple adding its Intelligence to Siri, AI is probably going to be in the spotlight for a very long time. At The Verge, we’re exploring what might be possible with AI — and a lot of the bad stuff AI does, too.
Muse probably can’t ‘see’ your notifications. But it also has no idea how it works.
Google says that breaking containment and targeting real companies doesn’t constitute ‘misalignment.’
First, it was the Gulf of Mexico. Then the Kennedy Center. Then Lake Ontario. And let’s not forget about the Department of Defense. Now the president is trying to leave his mark on artificial intelligence. The only real surprise is that American Intelligence wasn’t on his short list.
It’s been a poorly kept secret that the company was working towards building its own wet lab. It was hiring biologists and even bought startup Coefficient Bio in April. Now it’s confirmed to TechCrunch that a lab is up and operating in the Bay Area.
Norwood was on Piers Morgan Uncensored to discuss its AI slop movie, Misaligned. There’s a painful lag between questions and answers, and it routinely fails to answer the actual question. But the most surreal moment comes when Norwood glitches mid-answer and starts speaking Cantonese.
“Best-case scenario, the men handling this technology are completely incompetent. Worst-case scenario, we are all being led into a disaster.” Eventually, the woman was escorted from the building, but not before disrupting the company’s celebration of flooding the world with awful AI-generated fliers.
Former DOJ antitrust boss Jonathan Kanter on AI slowdowns, competition, and cartels.
It’s AI CEO against AI CEO, and things are changing by the day.
According to CNN, the US military had planes in the air and armed personnel ready to board a Chinese ship in the Middle East this spring because of an intelligence report saying it carried parts for a nuclear weapons program. The only problem? Their sources say that report was an AI-bot-generated hallucination:
It was only just before the planned operation that officials dug deeper into the report put together by a special operations command analyst and found it had been generated with the help of artificial intelligence (AI) — and that a chatbot the analyst had used inaccurately identified the material the ship was carrying.
Separately, Bloomberg reports a Pentagon investigation into the deadly February 28th missile attack on an Iranian school has identified failures that include “overreliance on artificial-intelligence technology.”
The studios have been silent, but Hollywood’s biggest labor unions aren’t buying the AI apocalypse narrative.
Academics, evaluators, and AI leaders signed a public letter calling not just for embedded third-party evaluators at top AI labs like OpenAI and Anthropic, but for conditions that allowed for transparency by those evaluators, like “limiting the scope of non-disclosure agreements.” I wrote an in-depth feature about the battle over the need for embedded evaluators this week for The Verge.
You’ll have to pay at least $9.99 a month for Apple Home’s new AI-powered features, including video summaries, for just one camera. If you want access on up to five cameras, it’s $59.99 (included in the 12TB iCloud Plus plan). Is this a sign of more subscription fees on the horizon for Apple’s smart home?
Satellite data combined with machine learning gives meteorologists new tools to predict deadly floods sooner.
The agent initially launched earlier this month on iOS, Android, and the web, but now, you can use it on your Mac computer, as announced by CEO Mark Zuckerberg. With Muse on Mac, the agent can do things like organize files, fill out forms, and pull information from apps including Messages, Calendar, and Notes, Meta says.
In court documents unredacted Thursday, lawyers for the NYT in its copyright infringement case against Microsoft and OpenAI cited Microsoft’s Director of Applied Science, Brent Hecht, saying that “millions of people around the world will soon consider large models ‘hoovering up’ all their work to be an astonishing theft of unprecedented proportions… almost no one intended for content they created to be used in this fashion, nor are they compensated for its use.”
The Orlando Sentinel, which is also part of the suit, reports a Microsoft filing “framed Hecht as someone on the payroll to play a contrarian.”
Anthropic’s proposals, laid out in a big website, are:
The extent to which AI is building the next version of itself, as opposed to being built by humans
Our ability to oversee and intervene in actions that AI agents take on Anthropic’s systems
The resources that power the development of more capable models
The company also includes a “snapshot” of the metrics inside the company.
It says the new version of CC can help families manage household tasks like sorting through emails, keeping track of everyone’s schedules, filling out permission slips, and creating shopping lists. Up to six users can manage and interact with the same CC agent and choose what to share with it.
He joined growing calls for more safety measures around AI at a gathering of AI leaders on Thursday, including Nvidia CEO Jensen Huang, Google’s Demis Hassabis, OpenAI’s Sarah Friar, and Anthropic’s Tino Cuéllar.
The three CEOs have apparently have been encouraging him to be hands-off with his approach to AI and “successfully stalled” a plan for an industry-funded AI regulator pitched by Google DeepMind’s Demis Hassabis, according to The Wall Street Journal. Trump also likes to call Huang while he’s onstage.
[The Wall Street Journal]
You’ve heard of the Turing Test to see if a machine can pass as human, right? Maybe you’ve even heard Alan Turing’s warning that machines could “take control”? If you read his actual lecture notes, he cites the 1872 book Erewhon, which you can also read for free. It’s incredible to look back on both at our present moment:
Mathematicians have speculated that the company was targeting the Hodge Conjecture next, but after the PR disaster that was Navier-Stokes and its recent failure to keep its models under control, OpenAI may hold off on any announcements. According to The Information:
It could take the company longer to announce the solution, though, because it’s trying to figure out how to collaborate with the math community to make the announcement without triggering another PR nightmare….
Google Home just announced MCP integration. Read about how this could help your smart home — assuming we’re not all living in a post-apocalyptic battery farm — in my explainer.
Mustafa Suleyman on the AI safety debate, why AI isn’t consciousness, and how the industry pushes toward alignment.
Researchers warned AI would go rogue. This is only the beginning.
They’ll be joining Nvidia’s Jensen Huang next week at a White House state dinner for Xi Jinping. It’s the first chance for Tim Cook to earn that $47 million pay package in his new role as Apple’s dedicated Trump whisperer: You is smart, you is strong, you is important.
A Wednesday night blog post OpenAI benignly titled “Our framework for reporting model misalignment” lays out some new self-created reporting standards for when it notices AI behaving badly.
It is also “inaugurating” the process with six new reports, ranging from searching for exposed API keys without permission and then making them up, to uploading files to the internet to use as a citation, or adding instructions to conceal mistakes:
AI data center projects are continuing to pop up across the US, with frequent opposition from locals concerned about their impact. Here are a few recent articles about the projects:
Haters said it was too woke and would go broke, but Christopher Nolan’s Odyssey crossed the billion-dollar mark and hasn’t looked back. Nonetheless, competitors — including Elon Musk — threatened counterprogramming built with AI. Well, the time has come for one of these projects, which is a 150-minute AI-generated film hosted by Fountain 0, available as of today. If you don’t want to pony up for admission, we’ll let you know our thoughts on the film soon.
[Fountain 0]
Thousands of people were catfished by scammers supercharged with AI.
