@thegenioo Getting a similar experience regardless of harness - I largely use codex cli now, but when I want the agent to be more scrappy or quick I can switch to Pi without losing a sense of continuity
today we launched the Legal AGI Lab. AI agents are beginning to operate in highly regulated environments like healthcare and financial services.
but existing legal frameworks aren’t ready for this. And this creates a bottleneck for the agentic economy.
so we are conducting interdisciplinary legal & AI research on how agents should be governed, held liable, and measured and defining the legal architecture required for autonomous agents to operate safely in high-stakes environments.
Norm sits at a unique intersection: we build AI agents, we deploy them with institutional clients, and we power Norm Law, an AI-native law firm operating on live legal work. that feedback loop between building, testing, and deploying is what makes this research different.
I wrote some thoughts on what opportunities will remain for software as AGI looms. The jagged frontier represents risk which can be mitigated through software. We're in the business of risk-adjusted excess intelligence now. Link below.
@kashyap286 I moreso think of it as a reflection of their ability to take big swings
Even if the AI bets don't pan out the way Meta wants them to, it's all a reflection of the fact that this is a dynamic organization that has managed to scale without becoming complacent or stagnant
At Norm we're building the runtime for machine-executable Legal and Compliance. It's not about accelerating the process of drafting legal documents. It's about making law and other policies executable in the way that a computer program is.
@eshear When I was at Facebook the internal Facebook was really good for this. Getting caught up felt like doomscrolling the newsfeed without all the doomscrolling guilt because it felt productive.
@GergelyOrosz I find fast/easy code generation seems to incentivize redundant simpler code over abstractions - because the models are less likely to break things when there's less context required and minimal side effects
Prediction: in the next 2 months Anthropic is going to somehow build ClaudeCode into the Claude Desktop App, and start stealing market share from Cursor.
We are excited to announce $48 million in additional funding today from Coatue, Craft Ventures, Vanguard, Blackstone Innovations Investments, Bain Capital Ventures, New York Life Ventures, Citi Ventures, TIAA Ventures, and Marc Benioff, bringing the total funds raised to $87 million over the past 18 months.
Investors in this round collectively represent more than $15 trillion in assets.
Norm Ai is a legal & compliance automation platform that transforms how enterprises handle legal, regulatory and industry requirements. Rather than treating compliance separate from the business, Norm implements compliance checks proactively and delivers end-to-end enterprise-wide solutions.
The Norm Ai solution is built with the Legal Engineering Automation Platform (LEAP), our proprietary system for creating AI Agents with legal and regulatory domain expertise. LEAP empowers lawyers and ex-regulators at Norm Ai to convert complex policies into operational AI systems powered by Large Language Models.
We are also excited to announce today that Troy Paredes has joined as Senior Policy Advisor and Head of Capital Markets Strategy at Norm Ai to accelerate the company in setting the norms for regulatory and compliance AI. From 2008-2013, Mr. Paredes was a Commissioner of the U.S. Securities and Exchange Commission, having been appointed by President George W. Bush.
At the SEC, Mr. Paredes was actively involved across the SEC’s regulation and oversight of capital markets, and has been a strong proponent of leveraging technology in regulatory and compliance environments.