As a longtime reader (and writer of two small pieces in the 1990s) of WIRED, I was curious if the coverage of the AI era really does differ from its coverage of the early internet. A couple of experiments with Astra shows it does. Interpretation depends on your view of AI, though
Information Center
Sequoia shutting down its Scouts program marks the end of an era
The company has drawn governments, a major chipmaker, and the Pentagon into an effort to control fragile photons and build a useful quantum machine. It aims to be the first.
According to data from levels.fyi, Meta is the top company OpenAI hires software engineers from followed by Google and Apple. On average they get a $300K - $400K pay bump. You can fit an entire other software engineer’s salary in the pay bump. 😬
Investors once expected a $2T valuation for Anthropic’s IPO, then $1.5T. Now, days before its planned roadshow, one investor says it may need to fall to $1T. It seems the dream of infinite tokens, infinite data centers, and infinite layoffs to fund it all are coming undone.
Microsoft spent $5B less than planned last quarter because its outsourced AI data centers weren’t ready. Now IREN reports more delays, while Microsoft is winding down its Oracle deal because the GPUs no longer meet its needs. Data center capacity is now the obvious weak link of the AI boom.
Shield AI is buying the eyes for its drones. It agreed to acquire radar maker Echodyne for $5.3B, and it is raising $2.5B at a $28.8B valuation. Echodyne radars already sit in counter-drone systems in 40+ countries. Source: Shield AI Video: Echodyne #Robotics #DefenseTech #PhysicalAI
Crusoe sells troubled Wyoming data center project to Google
Some new data from Anthropic looking at the impact of AI on the sciences by field: "With current capabilities, if adopted widely, AI could take on 27–52% of the researcher time needed to produce a new study, with a median of 38% across fields..." Here: www.anthropic.com/research/how...
Nous Research's Arete-27B (open-weight) Qwen3.8-27B, post-trained to remove China's bias. nousresearch.com/arete
OpenAI reportedly raises an additional $11 billion at a $1.25 trillion valuation
Tesla’s massive EV is the second within the last year to get the ax, while smaller vehicles are proliferating. It could be a huge benefit for the US battery supply chain.
The Duo apps are coming, says Apple
Huawei’s AI data center in Horinger, Inner Mongolia, is one of the largest in China by AI computing capacity. Yet it has just 7% as much computing capacity as the world’s largest AI data center, Colossus 2 in Tennessee.
An experiment in which AI models were asked to break into simulated networks showed the hacking skills of DeepSeek and GLM to be not far behind OpenAI’s most advanced cyber model, which isn’t public.
Hundreds of Artex AI hacking tools found publicly exposed to the web
Meta introduces Spaces, a new way to collaborate with your Muse agent
Loved this riff about AI products by @reckless.bsky.social on the latest episode of the Vergecast - "The products aren't good enough to sell people on the costs." www.youtube.com/watch?v=-U4t...
Zipline just raised $1.5B, and #DroneDelivery stopped being a pilot. Walmart signed for 300 more sites, up from 21 active today. At a $21.5B valuation, the bet is a nationwide US network. Source: TechCrunch Video: Zipline #Robotics #DroneDelivery #Automation
After the midterms, is the AI industry in for a rough two years?
Pentagon offers battery startup 6K $250M loan to build U.S. factory
Fox completes its $22B acquisition of Roku
Data brokers are collecting all sorts of internal corporate files to help train AI models. Not everyone is thrilled.
Ramp data shows AI spending stalling just as Anthropic nears its IPO
I’d love a dashboard of how much money is wasted on tokens from asking AI agents to redo work because they refused to follow instructions in the prompt or .MD file. That seems like a better efficiency dashboard for Anthropic to build than one tracking employee personal vs work related prompts.
Large language models are set to go lunar.
DoorDash says AI agents have generated only several hundred orders in total, versus 3M daily deliveries. Airbnb says agents generate almost none of its 500K daily bookings. The narrative is that AI agents will replace apps as the way consumers transact. So far, the numbers tell a different story.
Ripple-65B Preview, an open-weight model on agentic coding. huggingface.co/ricursiveAI/...
When it's worth it create your own decision model like Jev, or when it isn't. AirBnB's Frontier quality at 1% of the cost: a production benchmark of LLM classifiers airbnb.tech/ai-ml/fronti...
Major Investors Are Skeptical Anthropic IPO Can Achieve $1.5 Trillion Valuation www.wsj.com/finance/majo...
Private to this visit. Text and context are processed by AI.