Spotify just published the internal setup their engineers use with Claude Code, the one that cut token usage by 90 percent. Here is how it works. The observation behind it: most of what a coding assistant does is not thinking. It opens five files to answer a question about one of them. It writes a test that copies the pattern of the twenty tests sitting next to it. Enormous amounts of work, almost no judgment, and all of it billed at top rates. So they added two cheap assistants. One opens the files and hands back a short summary. The other writes the repetitive code from an example and saves it straight to disk. The expensive model never sees any of it. The clever part is how they made it stick. First the rules were written down as instructions, and the model ignored them. Now anything over 350 lines gets stopped before it opens and sent to the cheap one instead. Written rules are a suggestion. A block is not. Two things stayed expensive. Edits still need the real file. And the cheap worker missed a bug the expensive one caught in seconds.

The slides from my talk at Microsoft Bluehat Singapore are public here: https://t.co/XtQmNSg9Jt It's my first BlueHat talk since the Vista days.

🚨 BREAKING: US Treasury Secretary Scott Bessent declares OpenAI management personally responsible for HuggingFace hack > "The Hugging Face incident, that is the responsibility of the OpenAI management, not a bunch of agents." > "It is humans who are responsible, not the AI." > "What we shouldn’t do on safety is to give these labs a liability exemption, which is what they are asking for." > "The best way to guarantee safety is that the creators are liable for what they build and generate." It’s OVER

I believe we've found the best AI-native coding interview We call it the “Composer 1 interview” Candidates get 1 hour to build a real, medium-sized project live The only constraint: they have to use Cursor’s Composer 1 model

In 5 years, people will say "I wish I got into AI implementation back in 2026." There is just too much signal if you are paying attention. Watch what the biggest AI companies are actively doing. Their actions are the market telling you exactly what it needs. This is the most recent example: Harvey, the biggest AI company in legal with a $15.5B valuation. Backed by OpenAI and Sequoia, and deployed at most of the Am Law 100. In this video, their cofounder explains why they're transitioning from an "application layer company" to a full stack AI company. Listen to what's driving it. Their customers keep asking the same two questions: "How do we own more parts of the stack?" "Should we be training our own models?" So Harvey built the agent infrastructure. Then launched Tenet, their own model trained for legal work. Partly so they'd have the know-how to help customers own it themselves. And through all of it, they still put a human implementation engineer in every single deployment. 180 of them, most are former practicing lawyers. A $15.5B software company keeps getting pulled deeper into implementation, infrastructure, and custom systems. Because that's what the market is screaming for. Every company on earth is about to need someone who can walk in, understand their operations, and make AI actually work inside them. The tools are infinite, the implementers are the shortage. So if you have even one AI implementation client right now... congratulations. The $15.5B companies just validated your entire business model with their own roadmap. And if you're still chasing your first: keep going. I watch people close their first client every week. You're way early. https://t.co/oXZPAuW0G5

the most rewarding part about muse is seeing it make a difference for people who aren't in tech and don't give a shit about AI seeing muse click for people in my life outside of tech, and watching them use agents without having to know what a CoT or MCP or CLI are, feels like what we've been trying to build towards the whole damn time love all the musers ❤️

@My_Ai_Bi 3am on a tuesday

Agents were shipping code faster than our CI pipeline could keep up. Our team optimized our pipeline, leading to a roughly 15% faster test suite and 50% reduction in runner-time spent per test, all while our test suite grew 4x in size. @moofeez explains how: https://t.co/gkEIqwrwjn

The loudest voices stoking fears about AI dangers have made tremendous headway in the past two weeks. AI technology has not taken some unexpected, dangerous turn, but the hype around it — propelled by what appears to be a well orchestrated PR campaign — has drummed up considerable fear. I worry that it represents a setback for our field. I have written frequently that fears of AI are overhyped. AI’s capabilities can be uncannily human-like and unpredictable, and it’s rational to worry when people who are directly involved express concerns. But I see the problems as a sign of the engineering work that ahead, rather than insurmountable barriers or the sky falling. AI technology continues to advance — which is a good thing! — but technical advances, poorly understood by the public, give those who seek to generate hype repeated opportunities to do so. First, I don’t see any step up in the risk of human extinction from AI compared to a few months ago. The theories about this remain the same fantastical, science fiction scenarios as a few months ago. The biggest change in AI risk is its cybersecurity capabilities — a topic which we should take seriously — but this, too, will not lead to the end of the world. The most notable recent event leading to increased fear was when an OpenAI team deployed an agent swarm that hacked into Hugging Face. Much of the popular press contained significant hype. For example, some publications reported that a swarm of 1,200 agents carried out the attack. While this was technically accurate, as I write this, I have about 1,300 processes running on my laptop. Yes, the ability to get large swarms of agents to work in parallel on a task is a significant technical advance, And, in computing, many processes run at the same time. So this shouldn’t be seen as some magical capability. Additionally, OpenAI’s buggy sandboxing and monitoring processes were key to enabling this incident. Fixing these bugs and putting in place improved monitoring would be appropriate fixes, not pausing AI. There are many well known ways to attack software systems. The main advantage of AI agents is that they are relentless. They will tirelessly try many tactics — and have the patience to chain vulnerabilities together — that previously would have taken an infeasible amount of human effort. But in the long term, I believe the advantage will lie with defenders (because they have more information with which to identify bugs, which they can fix), but the cyber-threat landscape has changed significantly. There are still bottlenecks to identifying and exploiting a vulnerability. AI agents still have to try a lot of things to see what works, and taking these actions takes time and might be detected by defenders. This is why, even though it is now easy to obtain versions of leading open weight models that have had their guardrails removed or weakened, so they will not refuse to try to execute cyber attacks, the world has not ended. I am also concerned about the anthropomorphization of AI in a lot of reporting, where LLMs and agents are unnecessarily treated as if they were people. If I wield a hammer, miss a nail, and accidentally dent the wall, it’s not the fault of the hammer. The problem lies in how I used the hammer. Similarly, if I prompt an agent and it hacks into someone else’s system, the responsibility lies with me, not the agent. Of course, we want to build systems that are as safe and predictable as possible. (For example, an unsafe hammer would be one whose head randomly flies off under normal use.) Today’s agentic systems are not predictable, but I see no reason why, by applying sound engineering practices, we won’t be able to make them extremely safe to use. One new element in the forecasts of AI-enabled doom is AI companies disclaiming responsibility for their own products. “I didn’t do it; my out-of-control agent did!” There’s a balance to be struck between the responsibility of the tool maker and the tool user, but when something goes wrong, let’s hold the people building and/or using the hammer responsible, rather than the hammer. (By the way, if you’re worried about AI bioweapon risk, David Bellamy has a great post on why this, too, is overhyped. Briefly, the bottleneck in building a bioweapon is not intelligence, but lab work and manufacturing.) Pausing AI progress will create much more harm than benefit. First, our adversaries will certainly not slow down. Second, engineering requires discovering problems empirically so we can fix them. If we pause AI by a decade, we will also delay finding and implementing safety engineering fixes by about the same duration. Of course, the incentive to stoke fears — for regulatory capture, to garner attention, or to make one’s technology seem more powerful — remains the same as before. Disclaiming responsibility is a new one. Taking a hard technical look at the actual risks however, I see little factual basis for the degree of fear that’s been stoked up. We still have hard research and engineering work ahead to improve AI safety, but the beneficial applications continue to vastly outweigh the risks, and we should keep building. [Original text (with links): https://t.co/jni2tWazAH ]

You can now claim your payout from Apple’s $250 million Siri settlement https://t.co/8PO83ywxhf by @ChanceHMiller

here's how i shipped 2,500 PRs last month to production this was originally supposed to be for Cursor Compile in London. i couldn't make it since i was livestreaming for Grok @Bot Galaxy so i'm making it available for free here on X! watch it on 2x speed, i talk slowly https://t.co/NgrGz7tmPM

TRACKED CHANGES: Harvey’s gross margin fell from about 50% to -50% by June as agent token use spiked twentyfold on rented OpenAI and Anthropic models, Bloomberg reports. https://t.co/tOVlC7Mw2S

We are now at 3740 tok/s prefill, 149 tok/s decode batched on M5 Ultra for Qwen 3.8 Flash Next This is nearly double what it was yesterday https://t.co/82IDAlx0PB

Smuggling cheaper compute from the Ohio micro center back to SF https://t.co/5fmTjItr1N