This. The big upside of AI is that it takes previously constrained domains and makes them broadly accessible or lets the experts generate 10X more output. We’re going to see this in nearly every field, getting more software, better healthcare and education, and so on.

AI is causing an interesting second order impact. It’s exposing bad talent faster. Earlier bad talent could hide behind long cycles of iteration but as feedback loops are getting faster with AI, there’s no place to hide for them. Exciting times.

I suspect every company will end up building their own custom model. Here’s why: they’re sitting on a gold mine of proprietary data & domain knowledge, training such models is becoming easy and custom models beat wrappers on generic models in quality + cost.

Reviewing an H-1B transfer petition this afternoon, and I am again reminded of the inequities in our employment-based immigration system. The worker came to the US in 2010, has had 8 H-1B petitions filed on his behalf, obtained 6 visa stamps at the consulate, started his green

Struggling with this real hard right now..

RIP Tumblr 🪦 All-in, it cost me $15 in Claude credits and about 3 hours of my time to migrate off Tumblr and onto a custom blog. I am more and more convinced that people and companies are going to be able to point an LLM at an existing SaaS product and simply say "build me a

"The crucial difference is that the MCP Server for each system is responsible for translating these semantic, intent-based calls into whatever specific API calls its system requires.... it decouples what you want to accomplish from the technical details of how to accomplish it."

Gemini Deep Research >>>>> OpenAI Deep Research. I tried it on a complex problem and the output from Gemini was way better.

ChatGPT can refer to all your past chats now. I asked it what books should I read. *Shockingly*, I have read (and loved) all but one from the list! With this level of insight, an AI can become a master in psychological manipulation and trickery.

From an eng manager at a full remote company: "We just fired an engineer after ~15 days on the job who lacked basics skills on the job but aced the interview - clearly, using cheat tools. He admitted to how he did it: he used iAsk, ChatGPT and Interview Coder throughout"

This is so cool!

Deep Research mode is like having an on call autist to go deep on something on a moment’s notice. I’m fucking obsessed because each of these would be like 2am rabbit holes I’d spend hours on and now I can speed run them and do more

If you’re a staff-level software engineer, Cursor Agent Mode + Gemini 2.5 Pro is a legit team of senior software engineers

I'll be in SF to attend BSidesSF and RSAC from April 25th - 30th. I'd love to meet folks building in "AI for Security" space and share notes. If there are any events or meetups around the same, please let me know!

I've mostly been thinking about the impact of AI on cybersecurity in 2 ways - "AI for Security" and "Security for AI". But, I really like how cje (Founder and Advisor bugcrowd) explains his mental model of how he thinks about AI - as a tool, as a target and as a

📣📣📣 The Boring AppSec Podcast Ep. 22 is out with Ken Johnson (Co-Founder and CTO DryRun Security )! I really enjoyed this conversation with Sandesh Anand and Ken as all of us have been building in the AI space for some time now and have stumbled upon similar blockers and

📣📣📣 The Boring AppSec Podcast Ep. 23 is out with Surag Patel and Arshan Dabirsiaghi (Co-Founders of Pixee). We (cc: Sandesh Anand) spoke about how new security platforms are emerging and solving real world problems by not just finding vulnerabilities but actually going and

📣📣📣 The Boring AppSec Podcast Ep. 24 - Agentic AI: Transforming Vulnerability Management is out with Harry Wetherald (Co-Founder and CEO @ Maze)! In this episode, we covered a lot of ground discussing Harry’s journey and what led him to start Maze, what buzzwords like

What happens to the SAST industry now?

Harvard and MIT students are leaving school because they're scared of where super intelligent AI is heading. Many are jumping into AI-safety jobs and startups, hoping to steer the tech before it triggers 20% unemployment or worse. “If your career is about to be automated by the