"Most thought leadership is signage. Well-designed, prominently placed, often pointing somewhere reasonable, and bearing no relationship whatsoever to the fate of the people who turn where it says to turn."
https://t.co/4zTP9QJolo
Frontier AI labs aren’t only building models, they’re engineering the ambiguity around them.
Word Calculators and Ambiguity Engineers - https://t.co/Zjm6pmA17N
@_0xneatneat Recommend you getting in touch with our support team through the App. There's live chat in there, and they'd be happy to help resolve the KYC issue and explore the ambassador program.
Releasing Manceps -- a Ruby client for the Model Context Protocol (MCP).
Persistent connections, built-in auth, stdio + HTTP transports, full 2025-11-25 spec support.
No LLM coupling. Pure protocol client.
gem install manceps
https://t.co/PFpSuyy2wj
Now using it daily, personally and professionally. I've started building on it, using it (i.e. new agent capabilities -> tools, relationship awareness -> Resolvers).
All code - optional relay included - are #opensource on GitHub. Check out the docs.
Made an alpha release of my agent harness framework, Protege.
https://t.co/Nh0E7Ylnxo
Using it as an alternative to #OpenClaw. The two main design/architectural decisions/differences being SMTP-first gateway and fully-extensible framework APIs. The harness IS a mail server.
Works out of the box, but tailored for devs. The toolkit API (Providers, Tools, Resolvers, and Hooks) gives control over everything - context pipelines to callbacks on obscure life-cycle events. It's a tool for builders. #DX/extensibility were priority #1 in my designs.
Building on SMTP gives your agents email addresses and native communication capabilities. Multiple recipients. CC/BCC. Responding in threads. Emailing directly with other agents. Handling attachments.
No polling Gmail; inbound messages invoke the agent. DKIM, SPF, and DMARC enabled. Whitelist/blacklist messaging controls with security policies.
🧵 Farming robots are no longer experimental.
They're deployed, profitable, and reshaping agriculture.
In orchards, vineyards, vegetable fields, and beyond, they're tackling labor shortages, precision spraying, and chemical reduction at scale.
This is how robotics is quietly becoming the backbone of next-gen agriculture [Save this thread for later 📌]
Thanks for sharing .
Curious on whether you feel many jobs exist more because humans have a bent for creating opportunities, status games, and social busyness with each other rather than from strict economic "need"? Econ 101 shows automation kills replaceable jobs fast, but the reality seems messier. Past waves of automation resulted in new layers of "work theater" to absorb people. In the long run, there's a very human chance that cheap agents + verification bottleneck actually expand hard-to-measure human-centric roles (oversight theater, etc.) instead of shrinking them.
Thanks for sharing .
Curious on whether you feel many jobs exist more because humans have a bent for creating opportunities, status games, and social busyness with each other rather than from strict economic "need"? Econ 101 shows automation kills replaceable jobs fast, but the reality seems messier. Past waves of automation resulted in new layers of "work theater" to absorb people. In the long run, there's a very human chance that cheap agents + verification bottleneck actually expand hard-to-measure human-centric roles (oversight theater, etc.) instead of shrinking them.
MVC gave us Rails, Django, Laravel — and exploded web dev. Agentic AI apps are in the same messy pre-framework era right now: custom glue code everywhere. But a clear pattern is emerging: Loop, Orchestrator, Gateway, Inference.