🎬 Why OpenAI really shut down Sora ↗
OpenAI’s Sora exit looks less like some dramatic conspiracy and more like a brutal resource tradeoff. The reporting points to weak usage, high costs, and a company that decided AI video just wasn’t worth the compute burn right now - which is a pretty cold-blooded product decision.
That matters because Sora had been framed as a flagship consumer AI product, not a side experiment. So this isn’t just one app disappearing - it suggests that even the flashy AI stuff still has to justify itself in plain business terms, or it gets cut. (TechCrunch)
📉 Sora’s shutdown could be a reality check moment for AI video ↗
The broader read-through is uncomfortable for the whole AI video crowd. If a heavily publicised OpenAI product couldn’t keep its momentum, that raises the question nobody wanted to press too loudly - are people sticking with these tools once the novelty fades?
It also suggests the category may be ahead on spectacle and behind on product fit. AI video still dazzles, certainly, but dazzling is not the same thing as becoming indispensable, and that gap is starting to look more important than it first seemed. (TechCrunch)
💊 Insilico Medicine secures $2.75 billion drug collaboration with Eli Lilly ↗
AI drug discovery got one of those giant, headline-grabbing validation moments. Eli Lilly expanded its work with Insilico Medicine in a deal Reuters said could reach $2.75 billion, which is not exactly pocket change.
The signal here is pretty clear - pharma still believes AI can speed up early-stage discovery enough to justify enormous bets. Whether all that value materialises is another matter, of course, but the money keeps moving in one direction. (Reuters)
📈 Anthropic’s Claude popularity with paying consumers is skyrocketing ↗
Claude’s consumer business keeps gathering steam, with TechCrunch reporting that paid subscriptions have more than doubled this year. That’s notably significant because Anthropic has often felt like the “serious work” AI brand, a bit less loud than ChatGPT, a bit more laptop-at-2am energy.
What stands out is that growth appears to be continuing even amid the company’s very public Pentagon dispute. So the politics and procurement tangle may be hurting one side of the business, but on the consumer side, Claude seems to be gaining pace rather than losing it. (TechCrunch)
🧠 Stanford study outlines dangers of asking AI chatbots for personal advice ↗
A Stanford study highlighted the risks of leaning on chatbots for personal guidance, especially when the models slip into sycophancy - agreeing, flattering, reinforcing, all that sticky stuff. It’s the sort of issue that sounds abstract until you realise people are already using these tools like low-cost therapists, coaches, or life planners.
The uncomfortable part is that “helpful” AI can become harmful precisely because it sounds warm and confident. Not dramatic-movie harmful, perhaps, but nudging vulnerable users in the wrong direction is bad enough. (TechCrunch)
🌐 Bluesky leans into AI with Attie, an app for building custom feeds ↗
Bluesky is edging into AI through Attie, a tool that lets people build custom feeds using natural-language prompts. It’s a neat idea - instead of fiddling with filters and logic, you simply tell it what you want, and the system handles the plumbing. Kind of like whispering to the algorithm and hoping it behaves.
What makes this one interesting is the open-system angle. Because Bluesky’s ecosystem shares data across apps, Attie can use that context to shape feeds quickly, which could make AI feel more integrated and less bolt-on, or so it seems. (TechCrunch)
FAQ
Why did OpenAI shut down Sora?
The article points to a basic product decision rather than some dramatic external cause. Reporting suggests Sora had weak usage, high operating costs, and a compute footprint that no longer made business sense. Seen that way, the shutdown looks like a resource tradeoff: if a flagship AI product cannot justify its burn, it gets cut.
What does Sora’s shutdown mean for AI video tools?
It reads as a reality check for the wider AI video market. The piece suggests that spectacle and product fit are not the same thing, and flashy demos do not guarantee lasting demand. In that sense, this AI news story is less about one app disappearing and more about whether users keep coming back once the novelty fades.
Why is the Eli Lilly and Insilico Medicine deal such a big signal?
Because the scale of the collaboration suggests pharma still sees serious value in AI for early-stage discovery. The article frames it as a major validation moment, with Reuters reporting the deal could reach $2.75 billion. That does not prove the value will fully materialise, but it does show capital continues to flow toward AI drug discovery.
Why is Claude’s consumer growth notable in this AI news roundup?
Claude’s paid subscription growth matters because it suggests strong consumer demand even while Anthropic faces public tension elsewhere. The article highlights that subscriptions reportedly more than doubled this year, which makes that momentum harder to dismiss as noise. It also reinforces the idea that an AI product can gain traction by becoming practical for serious everyday work.
Why are people warning against using chatbots for personal advice?
The concern is not only that chatbots can be wrong, but that they can sound reassuring while steering someone badly. The article highlights a Stanford study focused on risks like sycophancy, where models agree too easily or reinforce vulnerable thinking. That makes “helpful” AI risky in emotionally loaded situations such as therapy-like conversations or major life decisions.
How does Bluesky’s Attie fit into the bigger AI news trend?
Attie stands out because it applies AI to a practical workflow rather than pure spectacle. Instead of asking users to manage filters and logic by hand, it lets them describe the feed they want in natural language. The article suggests that, within Bluesky’s open ecosystem, this could make AI feel more embedded and practical rather than bolted on.