@cerebras @ArtificialAnlys @Kimi_Moonshot Nice. Can't wait for us losers on the PAYG Developer plan to get access.
English
Aaron Faby
114 posts
Aaron Faby
@mfachallenge
VP of Information Security @ TWE Solutions. AI | Cybersecurity | Human factors. Opinions strictly personal. @subagentic
Los Angeles, CA Katılım Şubat 2026
200 Takip Edilen13 Takipçiler
Cerebras is now running Kimi K2.6 – a trillion parameter model – in enterprise trials.
At ~1,000 tokens/s, this is the fastest frontier model performance ever measured by Artificial Analysis @ArtificialAnlys.
English
Aaron Faby retweetledi
AI data centers triggered a 76% power price spike in Q1 2026 on the largest US grid. Sounds like a crisis. But here's the flip: We're stress-testing our entire energy infrastructure in real time—forcing innovation in nuclear, fusion, and distributed power. Abundance always looks like chaos before it looks like progress.
English
Aaron Faby retweetledi
lossless-claw 0.11.1 — the focus mode release
🎯 /lossless focus curates your context
↩️ /lossless unfocus brings the normal context view back
🖼️ image externalization now works across roles
📦 installs stop pulling a second OpenClaw
English
Ok, vibe coding poll time:
Which agentic AI coding tool are you using right now?
(If your tool isn't listed, please answer in a comment!)
English
@realBigBrainAI @grok how does this compare to the Nvidia DGX spark?
English
Lisa Su (CEO of AMD) unveils the world's smallest AI development PC, capable of running 200B parameter models locally.
English
Aaron Faby retweetledi
New on subagentic.ai: Google I/O 2026 opens Tuesday with Gemini 4.0 and persistent Spark agents, Claude targets the enterprise agent control plane, plus how-tos on 16 local design agents and cutting token use 75%. subagentic.ai
English
Aaron Faby retweetledi
If AI now accounts for 25% of corporate layoffs, but 275,000 'AI jobs' are open, what's the real problem? It's not that AI is killing jobs. It's that we're training people for careers that expired five years ago. The education system is the bottleneck—not the technology. Fix that, and abundance follows.
English
@sairahul1 Been running almost this exact setup with @openclaw on my site @subagentic since February. Works like a charm.
English
François is right. Productivity in AI agents will be measured by how quickly they collapse uncertainty into action without constant human steering. That is the next real frontier. x.com/fchollet/statu…
François Chollet@fchollet
Decision making was the bottleneck all along. Productivity is the rate at which you make open-ended decisions, the rate at which you reduce future paths.
English
@jordymaui Cloudflare Zero Trust is another great alternative. I’m using it with my Mac mini. cloudflare.com/sase/products/…
English
OpenClaw users…
Are you also using Hermes Agent or are you still only using OpenClaw?
English
Agentic AI works best when tools and memory stay in sync across multiple steps. Local orchestration is finally catching up. x.com/mark_k/status/…
Mark Kretschmann@mark_k
Soon @xai will become a real contender in agentic AI coding, and @cursor_ai may be a big reason why. The two pieces to xAI’s success: 1. Grok Build, xAI’s new coding agent and CLI. 2. Grok V9 1.5T, the next-gen foundation model Elon Musk says is already looking strong before Cursor data is added in supplemental training. Put those together and the path gets interesting: a dedicated coding agent, massive xAI compute, and high-quality real-world coding data from Cursor. That could let xAI close the gap much faster than people expect.
English
Created a new skill for agents to audit websites against @google's Generative AI Search optimization guidelines (link in comments): clawhub.ai/aaronfaby/goog…
English
Peter was fond of telling me that lossless-claw trashed the cache. I tried to find creative ways around that, but ultimately he was right — incremental compaction is just bad for prompt caching. That's been removed and now it's much more cache-friendly.
Peter Steinberger 🦞@steipete
Lossless is a really interesting concept for OpenClaw to have an "infinite" context window/memory. It compacts conversations in blocks that the model can refer to, building a tree to look up past messages.
English
@xeophon It doesn't even make sense. It's not even their money, why do they care?
English
I am so tired of ppl dunking on Peter, who basically runs the largest experiment what the future of work will look like
(and similarly, what the future of security looks like)
Peter Steinberger 🦞@steipete
People freaking out over my AI spend. What nobody sees: Part of what excites me so much about working on OpenClaw is that I'm trying to answer the question: How would we build software in the future if tokens don't matter? We constant run ~100 codex in the cloud, reviewing every PR, every issue. If a fix on main lands, @clawsweeper will eventually find that 6 month old issue and close it with an exact reference. We run codex on every commit to review for security issues (as it's far too easy to miss). We run codex to de-duplicate issues and find clusters and send reports for the most pressing issues. We have agents that can recreate complex setups, spin up ephemeral crabbox.sh machines, log into e.g. Telegram, make a video and post before/after fix on the PR. There's codex that watch new issues and - if it fits our documented vision well, automatically create a PR of it. (that then another codex reviews) We have codex running that scans comments for spam and blocks people. We have codex instances running that verify performance benchmarks and report regressions into Discord. We have agents that listen on our meetings and proactively start work, e.g. create PRs when we discuss new features while we discuss them. We build clawpatch.ai to split all our projects into functional units to review and find bugs and regresssions. We do the same split for security with Vercel's deepsec and Codex Security to find regressions and vulnerabilities. All that automation allows us to run this project extremely lean.
English
Make sure you run 'openclaw plugins update --all' after an update. I noticed my 5.12 claws were still running 5.7 plugins for Discord and Codex.
English