I’m not here to tell you AI is coming for your job. You’ve heard that a hundred times already, and frankly, nobody wants to here the same thing again.
You’ve also probably read the top skills to learn in 2026. Learn Python. Learn AI. Learn prompt engineering. Sure all those are valid. But here’s the thing: everyone is saying that. And when everyone is saying the same thing, the real opportunity is usually one step ahead.
So what’s that step?
Agentic AI. And hang on, it’s not some buzzword to add to your LinkedIn bio. It’s a fundamental shift in what AI does, how it thinks, how it works, and what it’s capable of. Right now, very few people understand it deeply enough to actually build with it.
That gap is exactly where opportunity lives. — Read More
Recent Updates Page 23
Don’t choose the WRONG career in 2026 (Data Scientist vs. ML Engineer vs. AI Engineer)
π0.7: a Steerable Model with Emergent Capabilities
We’ve trained a new model, π0.7, that exhibits a step-change in generalization. π0.7 is a general-purpose model that can perform a wide range of dexterous tasks with the same performance as fine-tuned specialists, but even more importantly, it can follow new language commands and perform tasks that were never seen in its training data. In our experiments, we see π0.7 exhibiting the first signs of compositional generalization, recombining skills from various tasks to solve new problems, like using new kitchen appliances and even enabling a new robot to fold laundry for which there is no laundry folding data.
… A true generalist model should perform all of the skills out of the box, and be able to recombine them to solve new tasks. π0.7 demonstrates initial signs of such general capability: it can perform dexterous manipulation skills like those we’ve previously shown with our RL fine-tuned π*0.6 specialist models, with the same speed and robustness, it can compose and recombine the skills it learned to solve new tasks, and it can generalize across robot platforms, scenes, and tasks more effectively than our prior models. The examples below illustrate this breadth of capability, from fine manipulation to long-horizon household behaviors all with one model, straight out of the box. — Read More
What Is Vibe Engineering? How AI Turns Ideas Into Working Prototypes Instantly
For most people, ideas used to die before they were ever built.
… “How are you actually going to build this?”
And we didn’t have a real answer.
Fast forward to today, that exact situation looks very different.
If you have an idea now, you don’t immediately worry about whether you can build it or not. You open an AI tool, start describing what you want, explore possibilities, and within minutes, you have something that resembles a working prototype. The barrier between imagination and execution has almost disappeared.
This shift is what we call vibe engineering. — Read More
UK gov’s Mythos AI tests help separate cybersecurity threat from hype
Last week, Anthropic announced it was restricting the initial release of its Mythos Preview model to “a limited group of critical industry partners,” giving them time to prepare for a model that it said is “strikingly capable at computer security tasks.” Now, the UK government’s AI Security Institute (AISI) has published an initial evaluation of the model’s cyberattack capabilities that adds some independent public verification to those Anthropic reports.
AISI’s findings show that Mythos isn’t significantly different from other recent frontier models in tests of individual cybersecurity-related tasks. But Mythos could set itself apart from previous models through its ability to effectively chain these tasks into the multistep series of attacks necessary to fully infiltrate some systems. — Read More
Frontier Systems for the Physical World
The dominant paradigm in AI today, insofar as it is used in production-ready settings, is organized around language and code. The scaling laws governing large language models are well-characterized, the commercial flywheel of data, compute, and algorithmic improvement is spinning, and the returns to incremental capability gains remain large and mostly legible. This paradigm has earned the capital and attention it commands.
But a set of adjacent and related fields has been making meaningful strides in its gestation phase. These areas of activity include VLAs, WAMs, and other approaches to generalist robotics models, physical and scientific reasoning in the pursuit of AI scientists, and novel interfaces for human-computer interaction (including BCIs and neurotech) that take advantage of advances in AI to rethink how we interact with machines. Beyond technical progress, each of these areas has seen the beginnings of an influx in talent, capital, and founder activity. The technical primitives for extending frontier AI into the physical world are maturing concurrently, and the pace of progress over the past eighteen months suggests that these fields could soon enter a scaling regime of their own. — Read More
Being a Staff+ Data Scientist in 2026
I became a data scientist in 2013 when the title was young. It was so new that most companies had no idea what a data scientist should be doing, only that they desperately needed one or they would be left behind. Sound familiar?
I’ve tried to survey the job description of data science a couple of times with varying degrees of success, most recently to go with some informal recommendations for creating data science degree programs. Together with a group of colleages we tried to summarize what data scientists do and the data science subtypes of maker, oracle, detective, generalist. But in the face of changing expectations this doesn’t feel like enough anymore. It’s time for a refresh. — Read More
Managing context in long-run agentic applications
In complex, long-running agentic systems, maintaining alignment and coherent reasoning between agents requires careful design. In this second article of our series, we explore these challenges and the mechanisms we built to keep teams of agents working productively over long time spans. We present a range of complementary techniques that balance the conflicting requirements of continuity and creativity.
… Language model APIs are stateless: to provide continuity between requests, the caller must provide the complete message history with each request. Agent frameworks solve the state management problem for users by accumulating message history between API calls. This fills the agent’s context window, which provides a hard limit on how much information the agent can handle. Even approaching an agent’s context window limit can degrade the quality of responses. For short-run applications, no extra context window management is typically required.
Complex security investigations can span hundreds of inference requests and generate megabytes of output, requiring special handling. Multi-agent applications, like ours, add further complexities. For each agent to optimally execute its role, it requires a tailored view of the investigation state. Each view must be carefully balanced. If agents are not anchored to the wider team, the investigation will be disconnected and incoherent. Conversely, sharing too much information stifles creativity and encourages confirmation bias.
Our solution uses three complementary context channels: Director’s Journal, Critic’s review and Critic’s Timeline. — Read More
Earlier Article
Stop Treating AI Memory Like a Search Problem
Back in October, my AI assistant stored a memory with an importance score of 8/10. Content: “Investigating Bun.js as a potential runtime swap.”
I never actually switched to Bun. To be fair, it was a two-day curiosity that went nowhere. But this memory persisted for six months, popping up each time I asked about my build process and quietly pushing the AI toward a Bun solution with confidence.
There was nothing wrong with the system; it was doing exactly what it was supposed to do. That was the issue. — Read More
Meta Is Warned That Facial Recognition Glasses Will Arm Sexual Predators
More than 70 civil liberties, domestic violence, reproductive rights, LGBTQ+, labor, and immigrant advocacy organizations are demanding that Meta abandon plans to deploy face recognition on its Ray-Ban and Oakley smart glasses, warning that the feature—reportedly known inside the company as “Name Tag”—would hand stalkers, abusers, and federal agents the ability to silently identify strangers in public.
The coalition, which includes the ACLU, the Electronic Privacy Information Center, Fight for the Future, Access Now, and the Leadership Conference on Civil and Human Rights, is demanding Meta kill the feature before launch, after internal documents surfaced showing the company hoped to use the current “dynamic political environment” as cover for the rollout, betting that civil society groups would have their resources “focused on other concerns.” — Read More
Rick's Cafe AI 