Posts tagged "Agentic Ai"

25 posts

← Back to all posts

The era of the augmented architect

3 min read

Product managers are evolving from process coordinators to augmented architects: AI-powered builders who own strategy and commercial outcomes.

The product management landscape of this year is defined by convergence: pressure to build faster (Vibe Coding), generate cash (Revenue Accountability), and prove value (Founder Mode). The "Product Management is dead" narrative is hyperbolic, but the role is changing. The administrative, process-heavy layer is dying: automated by agents or eliminated by Founder Mode flattening. The traffic cop model doesn't survive when AI generates working prototypes in minutes. What survives is the core function: value creation....
Read more →

Why AI Agents Fail Today Despite the Hype

3 min read

AI agents excel at coding but fail at business tasks. The gap reveals what's missing: learning from experience and error recovery.

The agentic AI hype promises autonomous decision-makers that replace employees or dramatically boost efficiency. The reality, according to Maria Sukhareva (Principal AI Expert at Siemens) in "Why AI Agents Disappoint," is that general-purpose AI agents don't work for most real-world business use cases. The WebArena benchmark proves the gap quantitatively. Researchers created realistic web environments (e-commerce sites, forums, development platforms) and asked GPT-4-based agents to complete end-to-end tasks like "Find the cheapest phone case and...
Read more →

World Models Teach AI to See

3 min read

LLMs excel at language. World models learn by watching: understanding space, time, and physics. I'm tracking why this matters for product builders.

On Lenny's recent podcast, Fei-Fei Li called LLMs "wordsmiths in the dark": eloquent but ungrounded in physical reality. The phrase resonated because it captures exactly what language models can't do: understand space, navigate environments, predict physics, or reason about the 3D world we inhabit. I've been following world models with growing curiosity. The contrast with LLMs is stark. Where language models learn statistical patterns from text, world models learn by watching: absorbing spatial relationships, temporal...
Read more →

AI Agents Multiply Work and Eliminate Jobs Simultaneously

3 min read

AI agents boost output but multiply review work while threatening entry-level jobs. Two patterns reshaping knowledge work simultaneously.

Traditional automation follows a script. You map the steps, define the logic, and the system executes. If-then-else at scale. AI agents are different. They have decision-making authority. You give them a goal, and they figure out the path, making choices on the fly based on context. That shift from scripted execution to delegated judgment changes what happens to your workload. What the data shows A recent study from Faros AI analyzed over 10,000 developers across...
Read more →

Context Engineering Turns AI Agents From Goldfish Into Assistants

7 min read

Google's whitepaper shows why stateful AI requires engineering context with Sessions and Memory, not just better prompts.

Your AI agent is brilliant. It can write code, analyze documents, and answer complex questions with remarkable sophistication. It is also a goldfish. Every conversation starts from scratch. Every user is a stranger. Every context is new. Google just released a whitepaper on context engineering that tackles this fundamental problem. The paper introduces a systematic framework for making LLM agents stateful using two core primitives: Sessions and Memory. The framework formalizes the architectural patterns that...
Read more →

The Gap Between AI Adoption and Enterprise Value

4 min read

Two-thirds of organizations are stuck in AI pilot phase. The gap between adoption and enterprise value isn't technology—it's redesigning workflows.

Two-thirds of organizations are stuck in the pilot phase with AI. They run experiments, they test use cases, they see promising results—then nothing scales. McKinsey's latest State of AI report (November 2025) reveals the pattern: 90% of organizations regularly use AI, but only 39% report enterprise-level EBIT impact. The gap between adoption and value isn't a mystery. It's a choice. What the data shows The survey of organizations reveals three distinct clusters: Most organizations (nearly...
Read more →

You're Not an AI User, You're an AI Manager

4 min read

AI agents transform knowledge work from execution to management. ICs need allocation and judgment skills, not just execution speed.

A year ago, an engineer typed code into an IDE. Maybe GitHub Copilot suggested lines. Maybe they asked ChatGPT for help. Today, that same engineer prompts an agent to write substantial chunks of code, then reviews what comes back. The work that used to take days now takes hours. The job didn't disappear. It became something fundamentally different. Aaron Levie, Box CEO, puts it directly: "The job of an individual contributor really begins to change...
Read more →

Early Experience: A Different Approach to Agent Training

4 min read

Meta's Early Experience research explores agents learning from their own rollouts. Early results look promising—here's what changes if it scales.

AI agents are currently in use, handling customer service interactions, automating research workflows, and navigating complex software environments. But training them remains resource-intensive: you either need comprehensive expert demonstrations or the ability to define clear rewards at every decision point. Meta's recent research explores a third path. Agent Learning via Early Experience proposes agents that learn from their own rollouts—without exhaustive expert coverage or explicit reward functions. It's early, but the direction is worth understanding....
Read more →

Agentic AI: It's the Readiness and Access Story

3 min read

The gap isn't whether AI agents work—it's who can deploy them. Infrastructure inequality is creating two types of organizations. Act now.

The gap between hype and reality isn't the story everyone's missing about agentic AI. The gap between who's positioned to deploy it and who's stuck waiting for infrastructure—that's the story. And that gap is widening every quarter. The technology is proven—access to it is not Nearly every senior enterprise developer is experimenting with AI agents right now. One in four enterprises is deploying them across teams this year. The question isn't whether autonomous AI systems...
Read more →

Anthropic’s Quiet Advantage in the AI Race

3 min read

Anthropic’s enterprise and life sciences focus shows how AI companies are testing different business paths: one broad, one deep toward sustainable growth.

“Anthropic’s growth path is a lot easier to understand than OpenAI’s. Corporate customers are devising a plethora of money-saving uses for AI in areas like coding, drafting legal documents, and expediting billing.” — The Wall Street Journal That line captures an important dynamic in today’s AI market: two companies building similar technology, but betting on very different ways to make it sustainable. OpenAI is chasing scale — hundreds of millions of users, a consumer-facing brand,...
Read more →

It's not a Search Problem, it's a Distribution Problem

4 min read

AI search consolidates power across three layers: intent capture, routing, and monetization. The market structure implications are explained.

Yesterday, I discussed how Atlas builds OpenAI's interface-to-platform flywheel through continuity of context. Today, it's an extension: the market structure implications of a company controlling all three layers of distribution. The conversation about AI search misses the real shift. This isn't about which tool delivers better results. It's about who controls the starting point for online activity and whether that control consolidates or fragments. Google has held that position for two decades. Chrome captures 65%...
Read more →

From Interface to Platform: What OpenAI’s Atlas Browser Might Really Signal

6 min read

OpenAI’s Atlas browser isn’t a surprise, but it might be a signal. This post explores how Atlas fits into OpenAI’s evolving ecosystem and why the interface-to-platform shift matters.

Everyone saw this one coming: Open AI's Atlas browser. Rumors of an OpenAI browser had been circulating for months, alongside steady hints in partnerships, SDK updates, and app integrations. When Atlas finally arrived, it didn’t feel like a shock. The surprise isn’t that OpenAI built a browser. It’s why they built one, and what that might unlock. Because Atlas isn’t just another entry in the growing list of “AI browsers.” It’s the latest move in...
Read more →

Claude Skills Might Be Anthropic’s Most Exciting Update Yet

3 min read

Claude Skills let users package reusable workflows that make Claude adaptable and modular: a practical leap in how AI assistants work.

I just tested Claude Skills, and it’s awesome. Anthropic is on a roll here with disruptive innovation. If you haven’t tried it yet, here’s a quick rundown of why it matters, and why it could reshape how we work with AI assistants. What Claude Skills are Skills are like snap-on capabilities for Claude. Instead of rewriting prompts or uploading instructions every time, you can package reusable logic, scripts, and guidelines into a small folder. Claude...
Read more →

AI Isn’t Replacing Human Help, It’s Redefining It

3 min read

AI isn’t replacing human help—it’s redefining it. History shows that every new technology first sparks fear, then expands what humans can do.

“AI will be the greatest source of empowerment for all.” - OpenAI’s Fidji Simo It’s a bold vision, one that suggests anyone, anywhere, could find help for whatever they need: a business idea, a mental block, or even emotional support. The promise sounds inspiring, but it also sparks a familiar unease. We’ve heard versions of this before. Every major wave of technology begins with the same fear: that something deeply human will be lost. When...
Read more →

AI Agents Grow Work Instead of Replacing It

3 min read

AI agents don’t replace jobs—they expand what teams can achieve. Learn how product leaders can turn automation gains into growth opportunities.

When a new technology shows up, most people ask, “Whose job will this replace?” A better question is, “What new work will this create?” In a recent interview with Every, Box CEO Aaron Levie shared a useful way to think about AI. He said AI agents don’t shrink human work. They expand it. By taking care of repetitive coordination, AI gives teams more room for creative thinking and faster experimentation. This idea matters for any...
Read more →

Planning Better with Claude Sonnet 4.5

2 min read

Claude Sonnet 4.5 introduces usage tracking and thinking mode enhancements that streamline AI planning and session management.

I’ve been using Claude Sonnet 4.5 (in Claude Code) for a couple of weeks, and two small updates changed how I plan my sessions. The first is usage tracking. There’s now a simple command (/usage) that shows how much I’ve used Claude Code—both per session and across the week. It sounds minor, but it’s a big deal if you use Claude as a daily coding or research companion. Before this, I had no sense of...
Read more →

Thinking Through Agentic Loops

3 min read

Exploring how agentic loops extend feedback loops by adding autonomy, iteration, and goal-directed action in systems and AI.

I’ve long been fond of feedback loops. Systems thinking taught me to look for them everywhere: how a fitness tracker nudges you to walk more, how customer signals shape a product roadmap, how our habits form through repeated cues and responses. Feedback loops are elegant in their simplicity: an action produces an effect, which feeds back to influence the next action. Recently, I came across the phrase agentic loops. At first, it sounded like another...
Read more →

Sora 2 Changes the Video Play

3 min read

Sora 2 pushes AI video into mainstream use. Here’s what it enables now, who gets disrupted, and why B2B teams should pay attention.

OpenAI’s Sora 2 is not just a model upgrade. It’s text-to-video with sound, physics that make sense, and a social app where anyone can remix clips. That shifts AI video from a lab demo to something that can spread in the wild. The following screenshot is from the video generated realistically with this prompt (shared by the Sora team): "A person is standing on 2 horses with legs spread. make it not slowmo also realistic....
Read more →

AI Platforms as the New Distribution Layer

4 min read

OpenAI’s Instant Checkout turns ChatGPT into a commerce channel. Here’s what product managers need to know about AI-native distribution.

Seven hundred million people use ChatGPT every week. That’s not just a user base, that’s a distribution channel that makes traditional retail look small. With its new Instant Checkout feature, OpenAI isn’t just adding payments. It’s signaling that AI platforms are on their way to becoming full-blown storefronts. For product strategists, this marks a shift as significant as the arrival of the App Store. Distribution itself is being rebuilt inside AI platforms. From Infrastructure to...
Read more →

What a Gigawatt of AI Really Means

3 min read

What a gigawatt of AI really means, and how abundant intelligence could reshape technology, healthcare, and society.

Sam Altman wrote yesterday about a future of abundant intelligence, imagining a world where we add a gigawatt of new AI infrastructure every week. This week, we already saw their partnerships with Nvidia and Oracle. He teased about more partnerships and details coming soon: >"If AI stays on the trajectory that we think it will, then amazing things will be possible. Maybe with 10 gigawatts of compute, AI can figure out how to cure cancer....
Read more →

From Product-Led to Product-Agentic Growth in B2B

7 min read

Product-Agentic Growth reframes how B2B software must evolve: AI agents doing work for users, reshaping PLG with autonomous workflows, predictive expansion, and real ROI.

Picture this: A procurement manager signs up for your B2B marketplace. Within 30 minutes, your product has analyzed their company's spending patterns, identified $2M in potential savings, and pre-vetted 15 suppliers that match their compliance requirements. It drafted three RFPs based on their historical templates and scheduled demos with the right stakeholders. The procurement manager didn't do any of this. The product did. This isn't just good onboarding. It's not even personalization in the traditional...
Read more →

APIs are the Strategic Foundation for Agentic AI and Beyond

7 min read

APIs are not small. They are the backbone of digital growth and the foundation of agentic AI and MCP orchestration.

(Expanding on my earlier quick-thought piece on APIs) APIs Hidden in Plain Sight APIs are often dismissed as “technical plumbing,” invisible to most business leaders. Yet they quietly power nearly every digital interaction, from mobile payments to streaming recommendations. Some of the most valuable companies in the world—Amazon, Stripe, Twilio—built their fortunes by turning APIs into products. Now, APIs are entering an even more strategic chapter. They are becoming the backbone of agentic AI and...
Read more →

Agentic Browsers Meet Their Hardest Test: Security

2 min read

Agentic browsers face significant security risks, such as prompt injection, but early defenses demonstrate why security will be the true differentiator.

Claude for Chrome (now in pilot), Perplexity’s Comet, and Dia are all pushing the idea of a browser that doesn’t just display pages but acts within them. But as soon as you let an AI click, type, and execute, the hardest problem comes into view: security. The quiet threat of prompt injection Anthropic deserves credit for going deep on vulnerabilities in its Claude for Chrome pilot. “Some vulnerabilities remain to be fixed before we can...
Read more →

Agentic AI Needs APIs to Act

2 min read

Agentic AI can reason, but it needs APIs to act. APIs are the execution layer that makes AI autonomy real.

APIs are often seen as back-office plumbing, but in the emerging world of agentic AI, they are the execution layer that makes autonomy possible. Without APIs, AI remains stranded in theory—able to reason, but unable to act. From Copilots to Agents The last wave of AI adoption has been copilots—tools that help users write emails, summarize documents, or draft code. These copilots assist, but they don’t take initiative. Agentic AI is different. Agents can plan,...
Read more →

The Winner's Curse: Rhyming History in the AI Era

3 min read

Why AI disruption differs from past paradigm shifts: faster cycles, probabilistic computing, and why today's tech winners face an accelerating curse.

Ben Thompson's latest piece hits on something crucial: when computing paradigms shift, yesterday's winners often become tomorrow's strugglers. >The risk both companies are taking is the implicit assumption that AI is not a paradigm shift like mobile was. In Apple’s case, they assume that users want an iPhone first, and will ultimately be satisfied with good-enough local AI; in AWS’s case, they assume that AI is just another primitive like compute or storage that enterprises...
Read more →