There are a few iconic things that will always signal we're watching a James Bond movie: A Martini, shaken not stirred; our hero behind the wheel of a gorgeous Aston Martin; and, of course, an opening sequence that features 007 pointing a gun directly at the audience and shooting, with a classic theme song blaring as a blood red animation drips down the screen.
It's becoming increasingly clear that OpenAI put staggeringly little thought into the rollout of Sora 2, its latest text-to-video generating app, a "move fast and break things" approach that has resulted in plenty of drama. Last week, the Sam Altman-led company released the TikTok-style app that churns out endless feeds of low-rent and mind-numbing AI slop. It's an " unholy abomination" that intentionally encourages users to generate deepfakes of others,
Meta has been getting content moderation wrong for years, like most platforms that host user-generated content. Sometimes it's a result of deliberate design choices-privacy rollbacks, opaque policies, features that prioritize growth over safety-made even when the company knows that those choices could negatively impact users. Other times, it's simply the inevitable outcome of trying to govern billions of posts with a mix of algorithms and overstretched human reviewers.
Without directly naming SafetyCore, Google explained that the optional setting can blur photos that may contain nudity and display a warning before you view or share them. Sensitive Content Warnings appears to use SafetyCore to analyze images locally on your device. Google has emphasized that SafetyCore runs entirely on your phone -- images don't leave your device, and Google doesn't know if nudity was flagged.
Google appears to have blocked AI search results for the query "does trump show signs of dementia" as well as other questions about his mental acuity, even though it will show AI results for similar searches about other presidents. When making the search about President Trump, AI Overviews will display a message that says, "An AI Overview is not available for this search":
The cost of making high-quality media is collapsing. The cost of getting anyone to care about it is not. As generative A.I. turns production into a near-commodity, cultural power is shifting from studios and galleries to the platforms that allocate attention and the algorithms that determine who gets paid. The new patrons are not moguls with checkbooks; they are recommendation systems tuned for engagement and brand safety.
"I found out that I didn't get booked for a Harry Potter convention recently, as I'm now doing OnlyFans. They explained it was because it's a 'family show and OnlyFans is affiliated with porn,'" she wrote. "This was baffling to me as some actors who do conventions (most actors, actually) have done TV and films in which they've done sex scenes and nudity. I'm just playing with my hair!"
The administration's announcement of the campaign explains it will target social media platforms, short video services, the livestreaming platforms used by Chinese e-commerce sites to host infomercials and even delve into comments left across the internet. Some of the material Beijing wants to root out - such as content that incites violence - would likely fall foul of content moderators in many nations.
For decades, Mark Lemley's life as an intellectual property lawyer was orderly enough. He's a professor at Stanford University and has consulted for Amazon, Google, and Meta. "I always enjoyed that the area I practice in has largely been apolitical," Lemley tells me. What's more, his democratic values neatly aligned with those of the companies that hired him. But in January, Lemley made a radical move. "I have struggled with how to respond to Mark Zuckerberg and Facebook's descent into toxic masculinity and Neo-Nazi madness," he posted on LinkedIn. "I have fired Meta as a client."
Mastodon, an open source, decentralized alternative to X, is rolling out a somewhat controversial feature by adding quote posts, which will launch next week. The feature, which allows a user to quote someone else's post and re-share it with their own response or commentary, has contributed to a culture of " dunking " on X, where users often deride other people by responding with snark or insulting humor. To address this concern, Mastodon says it's implementing quote posts with safety controls.
Over the past 15 years, I briefed each of these companies about this complex issue. The first such briefing, back in December 2010, focused on my research on how Al-Qaeda and other terrorist groups were using YouTube to radicalize young people in the West. At Google's invitation, I met with senior Google representatives, including its head of public relations and policy, senior policy manager, senior policy counsel and free speech attorney, at the company's Washington, D.C., office.
