“Tokenmaxxing” is making developers less productive than they think
There's a lot more code—but it's a lot more expensive and requires a lot more rewriting.
HeyStartup
Anthropic’s Amodei heads to the White House as Washington fights over Mythos access
Summary: Anthropic CEO Dario Amodei is meeting White House Chief of Staff Susie Wiles on Friday to negotiate access to Mythos, a frontier AI model that can identify and exploit thousands of zero-day vulnerabilities across every major operating system and browser. The meeting follows Anthropic’s blacklisting by the Pentagon after Amodei refused to remove safety […]This story continues at The Next Web
Zoom partners with Sam Altman’s World to verify that meeting participants are actually human
Summary: Zoom has partnered with World, Sam Altman’s biometric identity company, to let meeting participants verify they are human using World’s Deep Face technology, which cross-references iris-scanned biometric profiles with live video to display a “Verified Human” badge. The feature responds to deepfake fraud that cost businesses over $200 million in Q1 2025 alone, including […]This story continues at The Next Web
Uber launches $5 doorstep return pickups for Uber Eats purchases across 5,000 US cities
Uber has launched a service that sends a gig worker to your door to collect items you want to return to a retailer, for $5 a pickup. The feature, called Return a Package, is available in the Uber Eats app across nearly 5,000 US cities and works with nine retail partners including Target, Best Buy, […]This story continues at The Next Web
Intel’s new Wildcat Lake chips take aim at the MacBook Neo with AI as the differentiator
Intel has launched its Core Series 3 processors, codenamed Wildcat Lake, in what amounts to a direct response to the MacBook Neo. The new chips, announced on 16 April, target the same budget laptop segment that Apple redefined last month with its $599 machine, and they arrive with a familiar pitch: more choice, more AI […]This story continues at The Next Web
Hackers are abusing unpatched Windows security flaws to hack into organizations
A security researcher published details of three security vulnerabilities in Windows Defender, and the code used to exploit them. Now, hackers are taking advantage of the vulnerabilities in real-life attacks, according to a cybersecurity firm.
Volvo’s parent Geely is selling a $15,000 electric SUV with massaging seats and 600 km of range
In short: Geely, which owns Volvo, Polestar, Lotus, and Zeekr, has launched the EX5, a battery-electric SUV starting at 109,800 yuan (~$15,300) with massaging seats, a 1,000-watt sound system, and up to 610 km of range. Already sold in 35 countries and the most exported BEV A-class crossover in China, it undercuts the cheapest European […]This story continues at The Next Web
Train-to-Test scaling explained: How to optimize your end-to-end AI compute budget for inference
The standard guidelines for building large language models (LLMs) optimize only for training costs and ignore inference costs. This poses a challenge for real-world applications that use inference-time scaling techniques to increase the accuracy of model responses, such as drawing multiple reasoning samples from a model at deployment.To bridge this gap, researchers at University of Wisconsin-Madison and Stanford University have introduced Train-to-Test (T2) scaling laws, a framework that jointly
Zoom teams up with World to verify humans in meetings
Zoom will show a badge on verified participants' tile.
Canva becomes the design layer inside Claude with new Anthropic partnership
In short: Canva and Anthropic have launched Claude Design, a new Anthropic Labs product powered by Claude Opus 4.7 that uses Canva’s Design Engine to generate fully editable, on-brand visuals from text descriptions. The announcement coincides with Canva AI 2.0, which the company calls its biggest product launch ever, introducing conversational design, agentic orchestration, and […]This story continues at The Next Web
Most enterprises can't stop stage-three AI agent threats, VentureBeat survey finds
A rogue AI agent at Meta passed every identity check and still exposed sensitive data to unauthorized employees in March. Two weeks later, Mercor, a $10 billion AI startup, confirmed a supply-chain breach through LiteLLM. Both are traced to the same structural gap. Monitoring without enforcement, enforcement without isolation. A VentureBeat three-wave survey of 108 qualified enterprises found that the gap is not an edge case. It is the most common security architecture in production today.Gravit
Allbirds sold its shoe business for $39 million and is pivoting to AI cloud computing
In short: Allbirds is rebranding as NewBird AI and pivoting from sustainable footwear to GPU-as-a-service cloud computing after selling its shoe business to American Exchange Group for $39 million. The company secured $50 million in convertible financing, saw its stock surge 600% before falling back by a third, and plans to lease GPUs to AI […]This story continues at The Next Web
Gigs turns your concert history into a personal live music archive
New iPhone app Gigs uses AI to turn old tickets, screenshots, and emails into a personal concert archive with stats, memories, and more.
Chef Robotics escaped the robot cooking graveyard and says it’s thriving — here’s why
The company, which deploys AI-guided robot arms for food production, says it is looking to expand its services to provide for a broader array of customers.
Uber will now pick up your returns from your doorstep
The new returns feature is the company's latest effort to become an everything app.
Anthropic launches Claude Design, a new product for creating quick visuals
The company says Claude Design is intended to help people like founders and product managers without a design background share their ideas more easily.
Anthropic just launched Claude Design, an AI tool that turns prompts into prototypes and challenges Figma
Anthropic today launched Claude Design, a new product from its Anthropic Labs division that allows users to create polished visual work — designs, interactive prototypes, slide decks, one-pagers, and marketing collateral — through conversational prompts and fine-grained editing controls. The release, available immediately in research preview to all paid Claude subscribers, is the company's most aggressive expansion beyond its core language model business and into the application layer that
Should my enterprise AI agent do that? NanoClaw and Vercel launch easier agentic policy setting and approval dialogs across 15 messaging apps
For the past year, early adopters of autonomous AI agents have been forced to play a murky game of chance: keep the agent in a useless sandbox or give it the keys to the kingdom and hope it doesn't hallucinate a catastrophic "delete all" command.To unlock the true utility of an agent—scheduling meetings, triaging emails, or managing cloud infrastructure—users have had to grant these models raw API keys and broad permissions, raising the risk of their systems being disrupted by an
With US spy laws set to expire, lawmakers are split over protecting Americans from warrantless surveillance
Some lawmakers are calling for widespread reforms following years of surveillance scandals and abuses across successive U.S. administrations. But even if the spy law known as Section 702 expires in April, the government's spy powers will not automatically lapse.
Google’s AI Mode can now help you find products in stock nearby
Although you can already track hotel prices at the city level, the new update lets you do so for a specific hotel that you're interested in.