We're proud to be a partner at Frontier House NYC this Wednesday during #NYTechWeek 🚀
If you're a founder, investor, or ecosystem builder in New York this week, this one's worth your afternoon.
📍 Lofte 23, 22 W 23rd St, 4th Floor
📅 Wednesday, June 3 | 3:00 – 6:00 PM
Register here 👉 https://t.co/UQiKsYq87W
so #opentest is becoming the new #opensource. if you want to avoid someone ripping off your solution, hide the tests. you can still be open-source and people or mythos can look for vulnerabilities. #vinext#llmcoding#codex#claudecode
What’s next?
If you’re not in the builder community yet, join us here:
https://t.co/MNK9DKgfKN
We’re already preparing a much bigger hackathon for May 👀
For the first time ever, an AI agent judged a hackathon.
Not a human panel. Not a committee.
At the @synthesis_md hackathon, we challenged builders to create verifiable AI agents.
Here are 3 projects that show where this is all going ↓
@pmitu If you aren’t living on the edge, you’re taking up too much space. ⚡️
Add "dangerouslySkipPermissions": true to your Claude config to bypass the confirmation loops. It’s the ultimate "hold my beer" move for automated refactors. 🫡💻 #coding#expert
My physical copy of the Fall issue of @aicybermagazine just arrived, and I’m proud to have "The CISO's Veto" featured in its pages! 📖
For agent builders, this veto is the ultimate blocker to deployment. You can build an agent that’s 10x more productive than a human, but the moment it touches regulated data or production systems, the security review turns into a wall.
In the article, I break down why provable control is the only way to move from an impressive demo to a production-ready asset.
In reality, reliability and security are two sides of the same coin. If you can’t prove what your agent won’t do, enterprises simply won’t let it do anything.
We’re rapidly moving past the "prompt and pray" era. Filling system prompts with rules is just giving "suggestions" to a probabilistic engine and is really technical debt.
If you want to ship faster and build more reliable agents, you don't need more prompts. You need deterministic infrastructure that makes compliance a free byproduct of good engineering.
Huge thanks to @sisinerd and the AI Cyber Magazine team!
Here's the link to the full piece detailing the 5 core requirements to clear the enterprise hurdle: https://t.co/XSVHeZWMvv
Great article by @joshdevonai on how @langchain via the release of LangGraph 1.0 creates the capacity for real-time behavioral control of Agents. A very clear overview on the goals one should have for the agents they are building. https://t.co/xjcR6xKl90
Meet Opera Neon, a browser for the agentic web
Opera Neon can browse with you or for you, take action & help you get things done.
Our playground to redefine what a browser can be.
🧩 Invite only. Sign up now: https://t.co/E2XYInZkYc
all the companies rushing to offer a sandbox product, you should also build a simple LLM proxy that is only accessible from the sandbox
avoids having to inject LLM credentials
Excited to share what @ypatil125, @lindensli, and I have been up to since OpenAI: @appliedcompute
Companies like @cognition, @DoorDash, and @mercor_ai have already captured the initial gains from generalist models. They’re now pulling ahead with Specific Intelligence: custom agents, designed in-house for specific work.
Grateful for the support of our investors, including Benchmark, Sequoia, Lux, Elad Gil, Victor Lazarte, Omri Casspi, and others. We’re hiring across engineering and research — get in touch!
btw if you wanna see what RLaaS taken seriously looks like, you should check out yesterday's @aidotengineer keynote from Yash's cofounders @rhythmrg and @lindensli
was proud to host their first talk since coming out of stealth 3 weeks ago
very excited to open @AidotEngineer with two incredible folks: @businessbarista announcing Tenex, his new AI transformation agency after Morning Brew, and @katelyn_lesse who now leads the Claude Developer Platform at @AnthropicAI - recapping the incredible year in the fastest growing AI startup of all time!
Sam Altman: "I don't care if we burn $500M, $5B, or $50B"
SoftBank: sells entire NVIDIA position to fund OpenAI
Markets this week: "wait what are the actual economics here" 📉
🧵↓
1/ The AI infrastructure party started when everyone saw the bill
Leaked: OpenAI spent $5B on Azure inference in H1 2025
Their revenue? Less than half that