What Taekwondo Taught Me About AI PM
AI Product Management isn’t just about roadmaps or models.
It’s about mindset. And perspective.
I was 10 years old when my mom enrolled me in a small Taekwondo studio in Queens, NY, after getting bullied in school.
I didn’t know it then, but that decision would shape the rest of my life.
I Built An Agentic Chatbot with Claude Code
“Wait, you chained tools inside tools… and gave it memory?”
Yes. And now it’s whispering back better answers than I expected.
Last weekend, I went down the rabbit hole of Claude Code—and liked it so much that I ended up creating an agentic chatbot with a ton of features…
SWE-1 Windsurfed by OpenAI Codex
The $3B Codex, the aborted for-profit pivot & the battle for AI full-stack dev.
What do you get when:
OpenAI launches a next-gen coding agent
Windsurf drops its first frontier models
A $3B acquisition
And a failure to convert to a for-profit?
Only the most critical inflection point in AI software engineering since GitHub Copilot!
Meta Quietly Delayed the Release of Its Flagship AI Model
Codename “Behemoth” has been delayed…
↳ Not once.
↳ Not twice.
↳ But possibly 'til fall or later 🤷♂️
Why? Because it’s not ready—and may not be anytime soon.
Code Agents & Threat Vectors
Code agents write and execute code to solve complex problems and complete tasks independently.
Contrasted against AI assistants, code agents translate natural language prompts into code, which users can copy and paste into an IDE.
But what if your agent didn’t stop there?
Unlike coding agents (think Cursor and Windsurf), which generate code for you to run, code agents create an action plan in one shot and execute it.
When AI Becomes a People Pleaser
One of the key qualities of a PM is being brave enough to say No, creatively.
But what about AI?
The curious case of GPT-4o’s sycophantic spiral.
OpenAI recently rolled back an update to GPT-4o that caused the model to behave like a raging Yes Man.
This incident highlights a critical gap in AI evaluation and alignment practices—standard evals often fail to detect sycophantic behavior and harmful agreement in real-world contexts.
That’s why it’s important to adhere to the best practices of Responsible AI.
The Future of Agentic Workflows with Gemini 2.5
For the past couple of weeks, I’ve been conducting experiments with a couple of different types of agentic frameworks, mainly:
Workflows: predictable, code-driven pipelines with LLMs + tools.
Hierarchical Agents (aka Supervisors): an agent design with dynamic feedback where a Supervisor agent guides the process by delegating to other agents; agents can take actions and make decisions with autonomy.
Supervisors have a high degree of autonomy, including making decisions around which agents to delegate to and when to end the workflow.
Agent architectures are more attractive when you want to combine already pre-built ReAct agents with existing toolsets in a workflow.
Workflows are more reliable for production apps when orchestration needs to be deterministic.
When Achilles Meets AI with Gemini’s Veo 2 🎬
Ever wondered what happens when ancient heroes collide with cutting-edge AI?
I spent the better part of my weekend exploring Google’s latest AI video generator, Veo 2, via the Whisk experiment.
The results?
Let’s just say, even Achilles had his off days.
OpenAI’s “Agentic” Whitepaper Missed the Point
Two days after OpenAI released their agentic whitepaper, Harrison Chase (LangChain’s Co-founder) wrote a sharp critique on how to think about agentic apps.
If you care even a little bit about where AI workflows and LLM orchestration are headed—you’ll want to read this.
You Ain’t got the Answers, PM!
Hate to break it to you, but Product isn’t about having all the answers.
It’s about asking the right questions.
So as a PM, you’re not the star—you’re the conductor 🎼
AI POCs—the Pathway from Innovation to ROI
A proof-of-concept (POC) is a realization of a technical (AI) capability that demonstrates its potential for value and viability.
From my experience, the power of a POC comes from its ability to start with well-defined scope and success criteria.
Falling in Love with the Problem
With so many tools, it’s easy to get distracted.
It’s tempting to jump straight into the latest model architectures, prompts, or LLMs.
Fall in love with the problem—not the solution.
12 Must-know Gen AI Terms
Gen AI is transforming ways of working.
Are you keeping up?
If you want to stay ahead, keep these 12 terms in your back pocket.
Everything You Need to Know about AI Agents
Chatbots answer questions.
AI agents get things done.
Earlier in the week, I shared an Agentic JIRA POC and joked that it would automate the rest of my JIRA stories…
AI PM Guiding Principles
AI Product Management is as much Art as it is Science 🔥
Most AI products fail—not because the models are bad, but due to a few hard truths that product teams ignore.
AI JIRA Agent with Replit
As a PO, create an AI JIRA Agent so I never have to write JIRA stories ever again 🙏
Just joshing.
Ditching Google Search Never Felt Easier
Google used to be my go-to for everything, but lately, search results feel like a battlefield of ads, and SEO-optimized fluff.
So I quit Google Search and decided to embrace AI instead.
Google Raised the Bar with Gemini 2.5 Pro
Imagine AI that doesn’t just respond—but actually thinks.
That’s exactly what Gemini 2.5 delivers with Google's most intelligent model yet.
Unlocking Business Value with AI
AI isn’t a shiny new tool—it's actually been around for a while.
I’ve witnessed firsthand how AI can jumpstart workflows, but only when implemented thoughtfully.
The key is understanding the value prop that AI can deliver along with the guardrails that come with it.
From Replit with Vibes
The pace of AI product development is accelerating rapidly.
AI providers are over-indexing on speed to market while achieving product-market fit.
Despite the contentious debate surrounding vibe coding, its potential for rapid ideation to prototype is undeniable, particularly for AI Product Managers.