Deep Dive - SiliconNoon (Page 6)

The Stack

Tesla's Cybercab launched on a certification NHTSA never approved

Every automaker in America vouches for its own vehicles. Tesla just found out what happens when the vouching gets creative — and a regulator that likes autonomous vehicles anyway starts reading the fine print. The probe is about paperwork, not driving When Tesla put its first Cybercabs on Austin's streets on September 3, it did so without a federal exemption, without a waiver, and without steering wheel, pedals, or sideview mirrors — three features the Federal Motor Vehicle Safety Standards r

The Stack

Deep Dive — OpenAI's Agents API: renting out the machinery behind Codex

OpenAI spent this week handing out its plumbing. On September 10, the company put the Agents API into public beta: a managed service that exposes the same harness and infrastructure behind Codex — context management, tool use, multi-agent orchestration — through a single API call. As we covered in our brief on the launch, developers specify a task, a model, tools and an environment; OpenAI spins up the Codex harness, fans work out to parallel subagents that each keep their own context, compacts

The Arena

Amazon blocked AI agents for a year. Now it sells ads inside one

Amazon spent twelve months walling its catalog off from AI platforms — crawler blocks, a court order against Perplexity's shopping agent, product feeds cut from Google Shopping. On Thursday it announced the thing all that blocking was quietly building toward: its advertisers can now run ads inside ChatGPT. The mechanics are narrower than the headline. The pilot is US-only, limited to select brands that already advertise with Amazon Ads, and runs through Amazon's demand-side platform as a manage

The Guardrails

Deep Dive — The slowdown call moved from blogs into the building

In the first week of September, the case for slowing frontier AI down was made by a departing researcher, a chief scientist's blog post, and a think tank's essay. The people making it were insiders, but they were speaking from the edge of the industry — the person walking out, the essay published on a Saturday, the policy paper written by people who don't train models. In the last 48 hours the argument stopped being a side argument. Evan Hubinger, who runs alignment science at Anthropic, put a

The Arena

Apple's new iPhone is an AI phone wearing a phone costume

Apple held its first iPhone event under CEO John Ternus on Wednesday and shipped the iPhone 18 Pro, the foldable iPhone Duo, AirPods 5, and two watches. Read the spec sheets and it looks like an ordinary S-year. Read what the silicon and the software actually do and one thing is clear: Apple is no longer selling a phone that runs AI features. It is selling an inference device that happens to make calls. What shipped, in one pass The iPhone 18 Pro and 18 Pro Max get a 48-megapixel main camera

The Guardrails

Anthropic is building an intelligence shop to track its critics

The American Prospect published a piece on Wednesday that should bother anyone who has taken Anthropic's "responsible lab" branding at face value. Drawing on job postings, a podcast interview with the company's own security managers, and police records, it describes a frontier AI company assembling an in-house intelligence capability whose remit includes activism — and doing it while its CEO publishes essays about trust. What the reporting actually shows Daniel Boguslaw's story rests on three

The Frontier

Deep Dive — Math's credit system was built for humans. AI just broke it.

A Millennium Prize problem appears to have fallen, and the interesting part is not the proof. It is that the institutions mathematics uses to decide who discovered something — priority, publication, peer review, the two-year wait before a prize — all assume a human pace and a human author. OpenAI's Navier–Stokes result went from rumor to Lean-verified in five days. Nothing in the rulebook is calibrated for that, and the fight that erupted is what happens when a discovery outruns the machinery me

The Frontier

Deep Dive — 800M parameters just beat a 122B model at African translation

The most interesting model release of the week did not come from a frontier lab. Tether's AI research arm shipped three offline translation models covering 19 sub-Saharan African languages, and the smallest one — 800 million parameters — reports better translation quality than a 122-billion-parameter Qwen and a 27-billion-parameter TranslateGemma across three public benchmarks. It runs on a phone with no signal, no account, and no per-token bill. That result deserves the standard suspicion ever

The Guardrails

Matt Clifford quit ARIA for Anthropic. The revolving door is the story

Matt Clifford is stepping down as chair of ARIA, the UK's taxpayer-funded advanced research agency — five days after saying he intended to keep the job while taking a senior role at Anthropic. The reversal is the tidiest possible ending to a bad week, and it tells you almost nothing about the underlying problem. What happened, in order On 2 September, Anthropic announced Clifford as its managing director of international affairs, leading engagement with governments outside the US. At the time

The Frontier

Deep Dive — Robots are learning what their actions do, not what to do

Between May and August, at least ten Chinese embodied-AI companies shipped a world model. That is a land rush in four months, and it arrived just as the World Robot Conference in Beijing turned into a showcase for the same idea. When a field this young converges this fast, the useful question is not who is winning — it is whether everyone is building the same thing. They are not, and the split tells you more about robotics in 2026 than any leaderboard does. The clearest statement of the split c

The Everyday

Kenya's essay industry died in two years. The jobs plan didn't notice

The most useful AI-jobs story of the year is not about engineers or call centers. It is about an industry nobody was supposed to defend, and a government plan that assumed work like it would keep growing. What actually disappeared A New York Times feature published Saturday counts the collapse: at its peak, more than 40,000 people in Nairobi made a living writing essays and coursework for students in the United States and Britain. Two years after ChatGPT, the trade is largely gone. Students m

The Guardrails

Deep Dive — A think tank just put bombs on the table in the AGI race

A Washington think tank published a paper last week arguing that the United States should think through how it would blow up Chinese data centers. Not sanction them, not out-compete them — destroy them with conventional munitions, in the event that Beijing looks close to artificial general intelligence and Washington does not. The report is "Superpowers and AGI," published August 27 by the Center for a New American Security. Its author, Jacob Stokes, is a senior fellow and deputy director of CN

See all