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.
CEO of applications Fidji Simo told staff the company will prioritize coding and enterprise users over the wide array of projects it has been pursuing, the WSJ reports, including deepfake machine Sora, browser Atlas, and gadgets.
It gave a similar explanation when it delayed the release of ChatGPT’s “adult mode.”
Nvidia has uploaded the GTC 26 replay, and the embed above should be queued up to the Olaf part. (I met Olaf yesterday!) Right after that, you can help me witness what appears to be an entirely AI-generated music video that attempts to summarize the keynote. It is quite the thing.
During GTC on Monday, Nvidia announced it’s launching NemoClaw, an agentic AI platform that adds privacy and security protections to the autonomous AI platform OpenClaw by running it in an “isolated sandbox” environment:
NemoClaw uses Nvidia Agent Toolkit software to optimize OpenClaw in a single command. It installs OpenShell to provide open models and an isolated sandbox that adds data privacy and security to autonomous agents. This provides the missing infrastructure layer beneath claws to give them the access they need to be productive, while enforcing policy-based security, network and privacy guardrails.
The NVIDIA Space-1 Vera Rubin Module is the ticket. CEO Jensen Huang:
We’re working with our partners on a new computer called Vera Ruben Space 1, and it’s going to go out to space and start data centers out in space. Now, of course in space there’s no conduction, there’s no convection, there’s just radiation, and so we have to figure out how to cool these systems out in space. But we’ve got lots of great engineers working on it.
Also see: every other tech billionaire eying the sun’s limitless power and glossing over potential problems. Space!
The world’s first $5 trillion company (currently 4.47 trillion) is strutting its stuff in San Jose, California. We’re expecting an AI and robot show... but fingers crossed for N1 consumer laptop chips too! Jensen’s keynote starts 11am PT / 2pm ET.
How Yahoo escaped its Verizon death spiral and became profitable again.
Just days after Meta acquired the “social” network for AI agents, the platform has updated its terms of service, stating that users must be 13 years of age or older to use the site and are “solely responsible” for actions taken by their AI agents, whether they “act autonomously or otherwise, and irrespective of whether such actions or omissions were intended.”
He was commenting on deepfakes about Iran war being used to create online chaos. It’s quite the turn from a man who’s been aggressively deregulating the AI industry, blocking states from implementing their own safety guardrails, and personally using AI to spread political disinformation, smear opponents, and fabricate endorsements.
The job requires the ‘ability to recognize, express, and shift between emotions in a way that feels authentic and human.’
The company recently hosted a series of speakers at AIPCon, including Cameron Stanley, the Department of War’s Chief Digital and Artificial Intelligence Officer, who gave a chilling demo of Palantir’s Maven Smart System, where anyone or anything can be targeted for a military strike with a “Left click, right click, left click.”
Three years ago Buzzfeed embraced AI, using it to generate articles and quiz responses. Turns out, nobody wants that. The company posted a loss of $57.3 million in 2025 and its stock dropped to just $0.70. But CEO Jonah Peretti is still planning to bring “new AI apps” to market. 🤦
Amazon has expanded its Alexa Plus personality styles with a new “Sassy” option. The chili pepper icon hints at this Alexa’s “unfiltered personality,” but don’t worry, Amazon is keeping the spice mild, with only “occasional censored profanity.”
Chris K.:
Programming your robot to swear then programming it to censor its own swearing is something
Get the day’s best comment and more in my free newsletter, The Verge Daily.
The company postponed its next AI model, codenamed Avocado, from this month until at least May, the NYT reports. Performance apparently falls short of rivals like Google.
Meta’s spent billions trying to catch up, and Avocado will be its first major release since hiring Scale’s Alexandr Wang to revamp its efforts.
Narayen told investors today that AI-first products “should be our next billion-dollar business,” and that he will leave the role he’s held since 2007 once the board names a successor. He’s overseen its Creative Cloud rollout, big bets on the future of AI, and its abandoned $20 billion attempt to acquire Figma.
Narayen:
The next era of creativity is being written right now — shaped by AI, by new workflows and by entirely new forms of expression. Adobe has never waited for the future to arrive. We’ve anticipated it. We’ve built it. And we’ve led it.
Brief, Sweet, and Chill options launched in January for Alexa Plus, and now there’s also Sassy, an “unfiltered personality with razor-sharp wit, playful sarcasm, and occasional censored profanity.”
The clever comebacks-equipped voice is adults-only and requires additional verification checks, but it’s no Microsoft Tay, M3gan, or AIA. Where the Sweet version leads with “I’m radiating pure joy,” this one suggests mayhem and being “ready to wreck some things together.”
Techdirt’s Mike Masnick on the history of the NSA and mass surveillance in America, and why Anthropic’s fight with the Pentagon should worry us.
Ben Affleck’s AI startup, recently purchased by Netflix for around $600 million, has a different approach to gen AI.
Claude can now communicate across Excel and PowerPoint, saving you from needing to keep switching tabs or re-explaining datasets at every step. Anthropic said it’s Claude “carrying the conversation across apps without losing track of what’s happening in either.”
The newly-launched Meta Training and Inference Accelerator (MTIA) 300 chip is designed to train ranking and recommendations systems across Instagram and Facebook. And while the upcoming MTIA 400, 450, and 500 will be “capable of handling all workloads,” Meta says it will mainly use them for generative AI inference “in the near future and into 2027.”
[Meta Newsroom]
Mark Zuckerberg famously built Facebook, and the rest of Meta, on a “move fast and break things” philosophy, but one commenter has suggested that Grammarly’s new sloppelgangers could suggest a new approach for the industry.
tsmuse:
move fast and loot everything
Get the day’s best comment and more in my free newsletter, The Verge Daily.
