LLMs have evolved from language generators to autonomous agents capable of complex, long-horizon tasks. This evolution has produced paradigms including Prompt Engineering to elicit model capabilities, Context Engineering to manage information access, Harness Engineering to organize external tools and resources, and Loop Engineering to support continual reflection and self-improvement. Yet as tasks grow more complex, individual intelligence faces a fundamental limit: many tasks require heterogeneous expertise, interdependent subtasks, parallel execution, independent verification, and persistent state, exceeding any single agent’s organizational capacity. Augmenting one agent’s capabilities or context cannot resolve this architectural mismatch; intelligence must instead be distributed across specialized agents and organized at the system level. We call this System Intelligence: an agent system’s ability to organize and coordinate multiple intelligent components into a coherent, adaptive whole pursuing a shared objective. Achieving it requires more than adding agents; it demands explicit structures to organize work, coordinate heterogeneous agents, and maintain evolving execution states. We introduce Graph Engineering, an emerging paradigm for next-generation agent systems. Unlike prior paradigms that mainly optimize individual interactions or agent-level behavior, Graph Engineering constructs explicit, dynamic, evolving graph structures representing tasks, agents, and system states. These abstractions provide a unified foundation for organizing complex objectives, orchestrating heterogeneous agents, modeling system dynamics, and enabling scalable agent evolution. We systematically review the principles, methodologies, and applications of Graph Engineering for LLM agents. Related papers, open-source data, and projects are collected at this https URL. — Read More
#architectureRecent Updates
What Z.ai’s Ox Alpha reveals about AI economics
On Wednesday, Z.ai unveiled GLM-5.3-Flash, the first natively multimodal model of the GLM-5 series. The company said the model was released anonymously as Ox Alpha on OpenCode and OpenRouter where it completely overtook leaderboards and went viral for offering a capacity for 100 trillion tokens per day.
Notably, the company said in its announcement that all of the traffic from its model’s skyrocketing popularity was “served on Chinese AI chips.”
Chinese firms may be going after one enterprise pain point in particular: token costs. [T]his model’s popularity also points to another trend: not every daily use model needs to be state-of-the-art. — Read More
Autonomy and Innovation
While not every Western followed the cliché, by the 1930s cowboy serials had landed on a consistent visual cue: the hero of the show wore a white hat, and the villain wore a black one. At the end of the day, however, they both were cowboys with cowboy hats.
[H]ackers who are focused on patching vulnerabilities and protecting software are “white hat hackers”, while hackers who are focused on exploiting vulnerabilities for malicious reasons are “black hat hackers”. The actual takeaway is that all of this complexity is overwrought: just as a cowboy is a cowboy, a hacker is a hacker; the hat is not a statement of capability, but rather intentions, and those intentions are shaped by incentives.
This delineation between capability and intent and incentive is critical when it comes to AI. The point is the one I made in the introduction: when it comes to cybersecurity, the capability that is necessary for good defense is the exact same capability that is necessary for good offense; the color of the hat is a matter of who is actually prompting the AI. And, sometimes, not even that is clear. — Read More
Dylan Patel – Anthropic & OpenAI will have most of the world’s compute by 2028
Had a lot of fun chatting again with my twin brother Dylan Patel.
We went through lab economics over the next few years – the shift from inference to training as RSI draws near; and how Anthropic and OpenAI are on track to control most of the world’s usable FLOPs within the next few years (because they can monetize compute better and thus outbid everyone).
And then we discuss whether the >$10T of total AI capex we’ll see by the end of the decade will cause a sovereign debt crisis, where hyperscaler debt raises interest rates, drives non-AI exposed countries into bankruptcy, and crashes non-AI equities.
One question we weren’t able to resolve is whether there’s anything that can counter all the forces barrelling towards centralization in this industry – the economies of scale in training, the scarcity of compute, and eventually continual learning and RSI. — Read More
The turbulent AI era is here. The choices we make now are critical.
… AI will either be the greatest equalizer ever invented, or the worst source of injustice. The challenge is monumental. Even under the best circumstances, the transition to this new AI era will be one of the most turbulent times in human history. How will we use this technology to make the world a fairer place and keep it from widening the divide between rich and poor? How will we protect the people who are most vulnerable to the harms caused by artificial intelligence, including those who lose their livelihoods and the sense that they are in control of their future?
I believe that answering these questions and acting on the answers should be the world’s top priority. If the world takes the right steps AI will be a force for good and leave everyone better off. — Read More
Strategy in the age of AI – Eight points beyond the obvious
What happens when AI changes not only how companies compete, but the very game they are competing in?
[V]iewing AI through the traditional mechanisation/automation lens traps us in looking for operational benefits (much of the discourse today), when the real advantage lies in understanding its impact on the rules of competition. It changes where scarcity lies, how industries are organized, and what constitutes a defensible advantage.
As AI makes intelligence within individual modules abundant, advantage increasingly moves to those who can connect activities, learn across their boundaries, and improve the whole system together.
Vertical integration may therefore be returning, but in a new form. The prize is no longer necessarily ownership of every asset. It is ownership of the feedback loops that connect them. — Read More
SaaS Isn’t Dead. Sameness Is.
Everyone suddenly wants to tell you SaaS is dead.
Even Gartner now estimates that as much as $234 billion in enterprise application spending could be exposed to what it calls “agentic arbitrage” by 2030, roughly 20 percent of enterprise SaaS spending.
But Gartner itself hedges the apocalypse. It calls what is coming less an apocalypse than a metamorphosis.
I think that’s right.
The next SaaS company may run one service underneath ten thousand different applications. — Read More
How to evaluate LLMs before production
A language model can perform well on a clean benchmark and still struggle with the cases that matter in production.
But as a system moves closer to production, the evaluation problem changes.
Real inputs are often ambiguous. Labels may be inconsistent. Important context may be missing or truncated.
We encountered these challenges while evaluating an LLM-based system designed to reduce false positives in GitHub secret scanning.
Rather than determine whether an LLM could classify a string correctly, we needed to understand whether the system could reduce noisy alerts while preserving enough recall to remain safe for a security workflow. — Read More
Anonymous Ox Alpha processes 26T tokens on OpenCode, breaks OpenRouter launch record
Dax Raad (@thdxr)’s OpenCode said its users processed 26 trillion tokens through Ox Alpha during the anonymous AI model’s first four days, turning a free preview into one of the largest model trials on the coding agent.
The August 24th disclosure covered 327,000 unique users and 8,328,244 completed sessions, according to OpenCode’s usage dashboard. Ox Alpha ranked second among models tracked by OpenCode, behind DeepSeek V4 Flash at 33 trillion tokens and ahead of Xiaomi’s MiMo-V2.5 at 12 trillion. — Read More
OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show
At the Hot Chips conference on Tuesday, OpenAI shared a more detailed look at Jalapeño, including the first batch of benchmark results for the new system. Tested on SemiAnalysis’ InferenceX benchmark, Jalapeño registered both more tokens per user and more throughput per kilowatt than the currently available state-of-the-art inference processors.
“The bottom line is that the results show a very, very significant performance advance over state of the art,” said Richard Ho, OpenAI’s head of hardware, in a press call. — Read More
Rick's Cafe AI 