Coté

Also: human-in-the-loop theatre, MCP-on-Go pain, and Google Translate at 20. From: The Spanish napkin Related to your interests Where AI Accountability Goes To Die When You Stick A ‘Human In The Loop’ And Call It ‘Governance’ - 🤖: “Stuart Lauchlan and HFS / Altimetrik’s Humans at the helm of AI study of 505 G2000 execs find enterprise AI is cost-driven, accountability-asymmetric (CIOs carry the blame, not the strategy), and the ‘human in the loop’ is mostly theatre - only 25% say humans would prevail in a disagreement, 18% have clear visibility into AI reasoning.

This feels like me: NEWMARK: I guess, ideally, I would have more social skills—meaning, some. COWEN: We’re simulating social skills just fine here. NEWMARK: That’s the phrase I use. At least on my part, what looks like social skills is just fakery. I can do it for short amounts of time, maybe 90 minutes. I’ve given up, though, on actually accumulating social skills, getting better at it. More to the point, I try to get into positions where other people can show social skills.

Also: AI’s biggest critic loses the plot, Notepad++ on Mac after 20 years, and dark modes as cosplay. From: Monkigras, 2026 Related to your interests From Infrastructure to Agents: A Hands-On Guide to Secure Private AI with Broadcom Claude Code and the rise of personal software - Low-code, except high code made by non-programmers. “Developer loyalty is at zero right now”: Google doesn’t care which AI coding tool you use - “I come across as a wildcat in this one.

In this episode, Whitney and I talked with Emily Long, CEO and co-founder of Edera about their hardened container runtime - the kind you can swap in without re-platforming or kicking off yet another zero-trust migration. The pitch is basically: stop chasing detect-and-respond alerts and just fix the foundation. We also got into the messy reality of going from COO to CEO (the COO title is, it turns out, kind of made up) and what it’s like raising a deep-tech Series A as an all-woman founding team - the downside-vs.

The author at SCALE, March 2023. I've been having a terrible time getting talks accepted at conferences. Talking at conferences has, you know, been a huge part of my life for about 20 years, so this is kind of a bummer. Care to join me in some self-pity-partyingcare? I've got the wrong haircut What gives with my high rejection rate? My theories are: I don't have enough new material.

Also: AGENTS.md as model upgrade, Europe-as-vassal, and a Slack chatbot nobody bothers to verify From: WHY ARE YOU LIKE THIS Related to your interests What We Learned Testing Frontier AI Security Models Against Our Own Code - “What we learned is that the useful mental paradigm for this technology is not to treat it like a vulnerability ‘scanner.’ The models are non-deterministic; they do not work like static analyzers.

Also: AI vendor lock-in, half of GenAI projects fail, tinkers vs. endpoints, and a blogroll worth keeping Related to your interests MCP vs. APIs: Why You Need Both for AI Applications Beware Software Brain - 🤖: “The failure mode is exporting that logic to human life. The ask is no longer that computers adapt to people but that people make themselves legible to the machine: open your files, email, calendar, and messages so the AI becomes more valuable.

Here’s coverage of our recent Tanzu Platform 10.4 announcements. This is a big agentic AI release, bringing in to Tanzu Platform all sorts of features to secure, standardize, and otherwise make agentic AI work more enterprise-y. My think on 10.4 is listed below too. There’s also the short video I made on the 10.4: Tanzu Platform 10.4 Clippings "Introducing Tanzu Platform 10.4: Extending Platform as a Service to Agentic Applications," Darin Zook, Tanzu blog, April 15th, 2026.