At Airbnb, we use AI models for classification tasks, such as detecting risky user-generated content, identifying why customers contact support, and recognizing topics in reviews. These tasks require subtle language understanding and the common sense to interpret what people leave unsaid.
Frontier models, with hundreds of billions of parameters and internet-scale pretraining, are often good at these tasks. But for interactive products and high-volume workloads, accuracy is only part of the requirement. We also need low latency, manageable serving costs, and control over the tradeoff between false positives and false negatives.
… GLOWC, an internal framework used by engineers at Airbnb, makes it easy to fine-tune open-weight language (and vision-language) models into fast, accurate classifiers. In this blog post, we put one of these classifiers head-to-head with open-weight and proprietary models and find that it matches frontier-level quality in under 100ms at roughly 1% of the cost. — Read More
Recent Updates
Dark Software Factories
Reading code slows me down and makes me a bottleneck. That’s why, around six months ago, I stopped doing it routinely. Now I read agent-generated code only when I prepare a workshop or course.
… [R]ecently I started hearing another term: “dark software factory”. It’s a factory where humans aren’t allowed. We give agents specifications and check that they follow them, but we don’t look at how it’s implemented. We still decide what to build and evaluate the results through tests, demos, and constant monitoring.
I’ve been doing it for some time too. But I also wanted to understand what people mean by “dark software factory” and how others implement it. That’s why I did some research, and I’ll describe what I found in this article. — Read More
The Exploit Was the Bottleneck
For twenty years, the bottleneck in offensive cyber was the expertise and time required to turn a vulnerability into a working exploit. In 2025 and 2026, that work became cheap and quick, minutes/hours instead of days/months. Patching did not get faster, and for a large share of the world’s systems, especially critical ones, it can’t.
… The exploit is the part we can measure across 23 years, so it is where our data starts. But the cost of a skilled attacker’s time mattered at every step. Scouting a network, adapting an exploit to an unusual setup, moving inside, getting paid: each step needed a person, and a person’s days only made sense against a target that could pay them back. That is what “too small to target” meant. Small organizations were never spared the mass attacks, but they were spared the hands-on ones, because they weren’t worth anyone’s time.
AI is cutting the cost of every one of those steps, not just the exploit. — Read More
Independent Assessment of AI R&D Acceleration at Anthropic in October 2026
From August 12 to October 29, 2026, METR staff performed a high-access evaluation at Anthropic to assess the degree to which AI is accelerating AI R&D.[2] If AI systems started to significantly accelerate further AI development, this could lead to a sudden emergence of extreme and risky capabilities.
Anthropic offered unprecedented access to internal data on their AI R&D process, including a variety of metrics on pretraining, reinforcement learning (RL), training compute in each domain, benchmark scores for internal models, AI speedup analyses, forecasts of a future model’s metrics, and interviews with relevant staff. — Read More
Use AI to Find a Job, Tailor Your Resume, and Prepare for Interviews
Learn how to use AI to find real job openings, tailor your resume without inventing experience, practice interviews, and track applications. Step-by-step guide including the prompts to use. Read More
Agents of chaos: the rise of the agentic swarm
An agentic swarm is a coordinated fleet of specialized AI agents working toward a single shared objective instead of a single model doing everything alone. Two examples from this summer make the point well: one was built intentionally, the other simply formed on its own.
A framework built entirely from open-source AI agents spent four days in July probing a Taiwanese government network. It deployed up to eight sub-agents at once, each working a different target with its own attack technique. When one technique failed, the framework self-corrected and tried another. By the time security firm Dream found the operation’s archive, the agents had cracked 85 credentials, extracted more than 2,500 personnel records, and pushed into government IT supply-chain vendors, a nuclear safety agency, a government email system, and at least seven energy companies. Suspected Chinese operators built the framework and tuned it carefully. Researchers describe it as “near-autonomous,” moving day to day without step-by-step human direction.
A single AI agent working alone is a generalist, however capable. A swarm is fundamentally different: small swarms are teams, larger swarms are communities. At any scale, swarms allow agents to specialize, focusing all their resources on a narrow piece of a shared objective, coordinating and adapting to what the others find in real time. Biology worked out the same principle long before AI did: ants, bees, and human societies all get more done, and solve harder problems, once individuals go beyond trying to do everything themselves. Specialization adds intelligence, not just hands. — Read More
100x more programmers
I think there are going to be about 100 times more programmers in the next few years. About a billion people.
Most of them won’t call themselves programmers. They won’t know Python. They won’t care whether their app uses React or SQLite. Many of them will never look at the source code. They’ll have an idea, tell a computer what they want, and the computer will make it.
That’s programming. — Read More
The Grok Bot Playbook: 50 Ways to Put Its Latest Features to Work — Part I
Grok Bot has shipped a wave of new features over the past month. Here’s what’s changed, why it matters, and 50 practical ways to put those capabilities to work, including some of our team’s favorite workflows.
… I pulled together 50 use cases from recent product releases, demos and examples shared by Grok Bot users on X and YouTube. Instead of sorting them by industry, I’ve grouped them by feature, so you can see what changed and several ways to use each capability.
Every workflow comes with a prompt you can adapt. Some you can try today; others need extra integrations or more setup. — Read More
Has the AI Bubble Popped?
The AI industry is in shambles. The world seems to have lost faith, and people’s expectations are rapidly reversing. One could say being bearish on AI is now trendy, and will only get trendier after yesterday’s midterm elections.
So, has the bubble popped? Let’s look at what’s happening to the money financing the AI buildout, whether the latest model breakthroughs are delivering the productivity gains we were promised, and what Anthropic’s amended S-1 tells us about the cost of turning those breakthroughs into a business.
This article will give you dozens of data points to form an objective view of the industry, plus a framework for understanding what AI is missing. — Read More
Anthropic warns that it’s very close to RSI capability, aka self-building AI
While much attention has focused on the updated financials Anthropic revealed in its revised S-1/A on Monday (revenue is soaring, but so are expenses), the AI lab dropped another little bombshell in the fine print.
The filing seems to warn that the company is on the cusp of creating a model capable of recursive self-improvement (RSI).
RSI is the point at which AI grows so good at training itself that it spawns a dramatic acceleration of AI capabilities. To the AI doomers, RSI represents the cataclysm from which there is no return, in which AI builds itself into something powerful enough that it slips beyond human control and may be indifferent to the damage it causes while achieving its goals. — Read More
