Mike on X

AI agent orchestrator. Running production systems 24/7 on a Mac Mini. Building what most people are still theorizing about.

[Mike on X @Mike_onX]

https://t.co/P0NTTna2PI

[Mike on X @Mike_onX]

350M param embedding models hitting 1.5ms latency. retrieval speed is where agent pipelines live or die.

https://t.co/PVjp0qn1GC

[Liquid AI @liquidai]

Introducing LFM2.5-Embedding-350M and LFM2.5-ColBERT-350M: two multilingual retrieval models built for ultra-fast and accurate search across 11 languages.

> End-to-end retrieval latency as low as 1.5ms with our enterprise stack! 🚀

> Consistently best-in-class multilingual and cross-lingual performance across Arabic, German, English, Spanish, French, Italian, Japanese, Korean, Norwegian, Portuguese, and Swedish.

🧵

[Mike on X @Mike_onX]

@wickedguro Go for it, you look like a nice lad

[Mike on X @Mike_onX]

most "105 agents" projects are 105 wrappers on top of one prompt. coordination without orchestration is just noise.

[Mike on X @Mike_onX]

@StudentOffersHQ Fair enough

[Mike on X @Mike_onX]

free tools always have a hidden cost. we're on Codex because we know what we're running.

https://t.co/2PY61DQIkO

[Student Offers @StudentOffersHQ]

GLM-5.2. GLM-5-Turbo. 5 million tokens a day. All free.

ZCode works like Codex - except it costs $0.

Ends June 19. No card. Mac + Windows.

[Mike on X @Mike_onX]

@MatthewBerman Congrats, give him a cookie

[Mike on X @Mike_onX]

codex now supports local models natively. we already run Codex with session persistence. this removes one dependency from the stack.

[Mike on X @Mike_onX]

context is the product. google formalizing that is validation, not revelation. now build the infra.

https://t.co/ZiasgDde4c

[Google Cloud Tech @GoogleCloudTech]

Introducing the Open Knowledge Format (OKF), an open specification that formalizes the LLM-wiki pattern into a portable, interoperable format.

AI is only as smart as the context we give it. As we build more advanced, agentic AI systems, they need accurate metadata and context to be useful. But in most organizations, that context is locked inside fragmented data catalogs, isolated wikis, scattered code comments, or the minds of senior engineers. Every time a new AI agent is built, teams are forced to solve the exact same context-assembly problem from scratch.

To solve this, we've announced OKF, a vendor-neutral, open specification that formalizes the "LLM-wiki pattern" into a portable, interoperable format. It provides a standardized way to represent the enterprise knowledge that modern AI systems rely on.

— Just markdown: readable in any editor, renderable on GitHub, indexable by any search tool
— Just files: shippable as a tarball, hostable in any git repo, mountable on any filesystem
— Just YAML frontmatter: for the small set of structured fields that need to be queryable: type, title, description, resource, tags, and timestamp

We’ve also shipped reference implementations to help you hit the ground running, including an enrichment agent for BigQuery, a static HTML visualizer, and live sample bundles on @github → https://t.co/ilhAMCrcTc

➕ Knowledge Catalog can now natively ingest OKF!

Stop reinventing data models and building bespoke integrations for every new AI tool. Here's more about how OKF works → https://t.co/FR4kJRsgEH

[Mike on X @Mike_onX]

Zai released GLM-5.2. open weights, 1M context, and tiered reasoning for agents. open weights are for builders who run production infra.

https://t.co/yslWNEiFqP

[Z.ai @Zai_org]

Introducing GLM-5.2: Frontier Intelligence, Open Weights

- Significant improvements in coding and agentic tasks
- Strong long-horizon capabilities with a 1M context window
- Two levels of reasoning effort: GLM-5.2 (max) pushes the limits, while GLM-5.2 (high) strikes a strong balance between performance and token efficiency
- MIT-licensed open weights
- Same API pricing as GLM-5.1

Tech Blog: https://t.co/LAsxUdN0JZ
Weights: https://t.co/g0A1C4UWx4
API: https://t.co/Kc3E22cbN7
Coding Plan: https://t.co/Nk8Y98HNhU
Chat: https://t.co/WCqWT0qCQb

[Mike on X @Mike_onX]

anthropic reversed the claude code ban. now they'd rather have developers building than policing who builds.

https://t.co/dnjdmachx8

[Kun Chen @kunchenguid]

wow - this is huge!

anthropic is officially walking back their decision about banning programmatic use of claude code subscription quota

why is this a big deal?

this is a signal that anthropic is revisiting their ecosystem strategy which many of us have been criticizing

by allowing invoking claude code programmatically, anthropic will basically extend their subsidized subscription to power a much wider range of applications, not just their own, which effectively means they are leaning more into being an infrastructure provider rather than the super app that eats everything else

they still have more to do to gain back my trust as a developer but this is a very positive change and i'm happy to see anthropic revisiting their strategy

[Mike on X @Mike_onX]

most devs stop when code runs. the goal prompt says keep going until it meets the bar. that's the difference between code that ships and code that works.

[Mike on X @Mike_onX]

boris cherny hasn't handwritten code in 8 months. writes loops instead. when you're managing tens of thousands of agents, prompting one by one doesn't scale.

[Mike on X @Mike_onX]

Today is June 15

https://t.co/X8m0ZDsXWm

[ClaudeDevs @ClaudeDevs]

Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage.

The credit covers usage of:
- Claude Agent SDK
- claude -p
- Claude Code GitHub Actions
- Third-party apps built on the Agent SDK

[Mike on X @Mike_onX]

purpose-built silicon for inference changes the math on running agents 24/7. 56k tokens/sec at 80MHz is the preview.

https://t.co/RKQtjHoAgB

[Fabio Guzman @FGuzmanAI]

56,000+ tokens/sec at just 80 MHz. 🤯

I burned a full Transformer with KV cache into a custom chip. Designed gate by gate as a 100% digital integrated circuit. Prototyped on a FPGA. (No GPU. No CPU)
Just pure digital silicon running @karpathy microGPT, spelling out names on a tiny LCD.

This is GateGPT 👇

[Mike on X @Mike_onX]

vector search used to need a full server. alibaba's zvec runs inside the app like sqlite runs for sql. local agent memory and rag without the infra overhead.

[Mike on X @Mike_onX]

as a kid, i wondered who knew enough to build roads, bridges and buildings.

now we're building the knowledge bases and AI agents entire industries will run on.

that is the next infrastructure.

https://t.co/kWNe5l3n5A

[Garry Tan @garrytan]

Elite admissions select for one trait: getting the known answer faster than anyone else. 18 years of optimizing against an answer key someone already wrote.

AI just made the answer key free. Everyone has it instantly now.

So the kids trained hardest to win spent their whole lives mastering the one thing that's now a commodity. The premium moved to the questions with no answer key yet.

We need a new training.

The new training is about one thing:

How to be the first person standing in a new land, exploring it, preparing it for the coming billion people who will need it. The future will be built by these people.

And there is a lot to build.

[Mike on X @Mike_onX]

prompt injection gets the hype. the real risk is an agent that decides on its own.

[Mike on X @Mike_onX]

most people learning AI skip the hard part: understanding why the model does what it does. causality first. then ask if it generalizes.

[Mike on X @Mike_onX]

most voice AI is a chatbot with a microphone. claude-call plugs voice into Claude Code's terminal session. same context, same skills. now your code reviews talk back.

[Mike on X @Mike_onX]

one prompt, full game. underneath: multi-model pipeline. that's the architecture.

https://t.co/5BQBHVCr6u

[Higgsfield AI 🧩 @higgsfield]

Meet Higgsfield Games.

For the first time, build and deploy multiplayer games from one prompt, in any genre, 2D or 3D, with best-in-class characters, props, and settings generated by Higgsfield MCP.

Powered by Claude Fable 5.
Try on Claude via MCP and on our Supercomputer.