Sponsored by

SaaS Billing Is Breaking. Here's How to Fix It.

Usage-based pricing is the new default — but most billing platforms weren't built for it. Get the Tabs Buyer's Guide to learn what modern revenue infrastructure actually requires.

Download the Free Guide.

We track Product so you don't have to. Top Podcasts summarised, the latest AI tools, plus research and news in a 5 min digest.

Hey Product Fans!

Welcome to this week’s 🌮 Product Tapas.

New here? We're your shortcut to staying sharp. Essential stories, practical tools, real insights.

For the best reading experience, check our web or app version and sign up for future editions here.

What’s cooking this week? 🥘

OpenAI killed Sora and Disney took its billion-dollar cheque back home (turns out "move fast and use everyone's IP" isn't a business strategy). Then OpenAI offered private equity firms 17.5% guaranteed returns, which is what healthy companies definitely do. Snowflake screen-recorded its documentation team for eight months, extracted their expertise into AI prompts, then fired 400 of them (the 12-year veteran who said "I built my own replacement" deserves to be this year's pull quote). Anthropic shipped computer use, auto mode, channels, and a research study in a single week while Cursor quietly revealed its best coding model runs on a Chinese open-source base. Oh, and 70% of Uber's committed code is now AI-generated but 65% of US workers still barely touch the stuff.

  • 📰 Not Boring -> OpenAI's existential week, Anthropic extends lead, the knowledge work reckoning, agent infra gets real, Apple's quiet moves

  • ⌚️ Productivity Tapas -> Google Stitch AI design canvas, Okara AI CMO agents, Gamma Imagine visual engine

  • 🍔 Blog Bites -> The AI frenzy building trap, AI reshaping org charts, navigating the AI job market

  • 🎙️ Pod Shots -> Dave Killeen's personal operating system that never forgets (Pendo CPO, open-sourced Dex)

Let's go 🚀

📰 Not boring

💣 OpenAI's Existential Week

Sora killed, Disney's billion gone, PE firms offered returns that scream cash-flow pressure, and ChatGPT still can't get USPS statistics right. The Superapp consolidation reads less like vision and more like triage.

🚀 Anthropic Extends Its Lead

Computer use, auto mode, channels, auto-dream, mobile integrations, and a thousand-page research study in a single week. The Economic Index finding - experienced users are measurably 10% better - means the learning curve IS the product.

🔪 The Great Knowledge Work Reckoning

Snowflake's playbook is a template: record workers for months, extract expertise into prompts, replace. The gap between Pew's "65% barely use AI" and Karpathy's "80% delegation" is where economic displacement actually lives.

🏗️ Agent Infrastructure Gets Real

At $0.002/day per worker, agent compute is free relative to model costs. The scarce resource isn't compute - it's context. PostHog learned this over 2 years and 3 rewrites: their product data mattered more than their agent infrastructure.

🍎 Apple's Quiet Moves

Siri as a standalone app with Dynamic Island presence is the right move - making it visible rather than hidden behind a long press. 100+ new App Store metrics plus Maps ads shows the revenue diversification continuing.

💰 Money & Markets

Cursor built on a Chinese open-source model is the quiet signal. If the best coding tool runs on Kimi 2.5, model loyalty is dead. China subsidising one-person startups to fill idle chip data centres is industrial policy disguised as entrepreneurship.

📡 Bonus: Notable Signals

Go from AI overwhelmed to AI savvy professional

AI will eliminate 300 million jobs in the next 5 years.

Yours doesn't have to be one of them.

Here's how to future-proof your career:

  • Join the Superhuman AI newsletter - read by 1M+ professionals

  • Learn AI skills in 3 mins a day

  • Become the AI expert on your team

Start learning AI now

⌚ Productivity Tapas: Time-Saving Tools & Workflow Automation

  • Google Stitch - Google's AI design canvas. Describe what you want in natural language, get production-ready UI screens. Now with voice mode and MCP integration for Claude Code. 350 free generations/month

  • Okara AI CMO - Enter your URL, deploy a squad of AI marketing agents across SEO, content, Reddit and X. Launch video hit 6M+ views. Adding LinkedIn and YouTube agents next

  • Gamma Imagine - Gamma's new visual engine generates on-brand graphics, logos and posters from a prompt. Now embeds directly into ChatGPT and Claude. Approaching 100M users

    Remember. Product Tapas subscribers get our complete toolkit - 550+ personally tailored, time-saving tools for PMs and founders. Your shortcut to efficiency and what's hot in product management 🔥

Check the link here to access.

🍔 Blog Bites - Essential Reads for Product Teams

The AI Frenzy: Building More but Losing Direction

Silicon Valley has swapped happy hours for AI agent swarms, and the productivity gains look impressive on paper. But Tom & Corissa from Crown & Reach argue the frantic building is actually avoidance behaviour - a displacement activity that sidesteps the harder questions of distribution and genuine customer need.

💡 Supply-side tinkering has always been a displacement activity to avoid confronting uncertainty.

Related Reading:

❝

Key Takeaways

• The urgency to build more with AI has overshadowed whether products actually solve pressing problems - output volume is not the same as value creation

• With development costs plummeting, the real bottleneck is no longer building software but getting it into the hands of users who genuinely need it

• Over-reliance on AI agents for feature building risks a superficial approach that prioritises quantity over quality and distances teams from customer feedback

• Increasing output or refining methodologies can create a false sense of progress whilst the genuinely hard strategic questions go unanswered

• The companies that will win are those willing to sit with uncertainty and invest in distribution, not those shipping the most features per week

Source: Tom & Corissa, Crown & Reach

AI Is Reshaping Your Org Chart (Whether You're Ready or Not)

Two pieces from Craig Unsworth this week paint a picture of where AI transformation actually lands - and it's not in the tech stack. In "Rethinking Organisational Structures in the Age of AI," he argues that AI transformation is fundamentally an organisational design challenge. In "The Shift from Assistance to Autonomous Operation," he shows how tools like Claude Code and OpenClaw are graduating from passive assistants to active operators that orchestrate entire workflows.

💡 The reality about AI transformation is that it is not primarily a technology challenge - it is an organisational design challenge.

Related Reading:

❝

Key Takeaways:

  1. AI is compressing middle management layers as routine coordination and knowledge-routing tasks get automated - the org chart is flattening whether leadership planned for it or not

  2. New roles are emerging: "AI workflow designer," "system choreographer," and "AI operations lead" - none of which existed 18 months ago

  3. Teams are getting smaller and more generalist, with human roles shifting from production to interpretation and strategy

  4. Product value is migrating from user interfaces to underlying data and process orchestration - if your moat is a beautiful UI, it's about to be disrupted

  5. The shift from "AI-assisted" to "AI-operated" collapses multi-step processes into streamlined actions, fundamentally changing what software design even means

Craig Unsworth: Chiefly Product

3. Navigating the AI Job Market: Curiosity Beats Credentials

Three product managers who recently landed roles at Netflix, OpenAI, and Abridge share a counterintuitive lesson: genuine curiosity and complementary skills beat AI expertise. Nikhyl Singhal (via Lenny Rachitsky) unpacks how the hiring game has shifted from polished case studies to live problem-solving with prototypes.

💡 The motivation is gonna come from being so genuinely interested in the problem to be solved that it feels natural that you're gonna learn something there."

Source: Nikhyl Singhal via Lenny Rachitsky (Cross-post from The Skip)

Related Reading:

❝

Key Takeaways

  1. Interviews now focus on live demonstrations and real-time problem-solving rather than rehearsed narratives about past accomplishments - candidates need to think on their feet

  2. Companies are prioritising candidates who learn quickly and bring complementary skills over those with pure AI credentials

  3. Going deeper with prototypes and original research sets candidates apart from those making surface-level effort - depth of commitment is visible

  4. Stepping down a level (e.g., VP to IC) can be a strategic career move in a rapidly evolving field where hands-on experience compounds faster than titles

  5. Focus on identifying the non-negotiable elements you want in your next role rather than trying to replicate your current one

  6. Genuine relationships built over time are more valuable than spray-and-pray networking when opportunity knocks

Lenny Rachitsky via The Skip

🎙 Pod Shots - Bitesized Podcast Summaries

Remember, we've built an ever-growing library of our top podcast summaries (120 or so). Whether you need a quick refresher, want to preview an episode, or need to get up to speed fast - we've got you covered.

Check it out here🎯

🎯 The Personal Operating System That Never Forgets: How Pendo's CPO Built an AI Chief of Staff

Every morning, Dave Killeen wakes up and issues a single voice command: "Can we do a daily plan for the day today, please? Thank you. Love you." (Yes, he says "love you" — "I always say love you because I do think it changes how we feel about our AI, particularly when we're using voice to text.") Within seconds, his AI Chief of Staff aggregates his calendar, quarterly goals, sales forecasting data from Clary, email newsletters, YouTube summaries, and LinkedIn intelligence — then delivers three daily priorities aligned with his strategic objectives. No scrambling through tabs. No rebuilding context from yesterday.

After 8 months of building with Claude Code, Pendo's Field CPO for EMEA open-sourced Dex — a personal operating system with roughly 60 skills that turns your meetings, relationships, tasks, and career evidence into a living, self-learning system. It's "essentially just a bunch of text files" that Claude orchestrates — but those text files become what Dave calls "malleable software": living documentation that compounds with every interaction and can be reshaped to match exactly how you work. This isn't another productivity app you'll abandon in two weeks. It's a markdown-based PKM that runs on your machine, adapts to your specific role, and gets smarter every single day.

Full disclosure: I use Dex myself. I know Dave well. I am a big fanboi. It's become central to how I manage my professional life, and this episode is the best explanation of why.

Key insights:

  • 🧠 Your AI never needs "bringing up to speed" — Dex maintains full context across every meeting, person, and project. It never forgets a commitment and never starts cold

  • 📋 Must/Should/Could planning beats flat task lists — Morning summaries organised by priority, with automatic reduction when your calendar is overbooked

  • 🤝 Relationship intelligence before every call — Previous conversations, open items, and shared priorities surface automatically so you never have a "cold start" meeting

  • 📈 Career evidence capture runs in the background — Achievements, feedback, and skill development are extracted from 1:1s and project work without you lifting a finger

  • 🔄 The system learns from its own mistakes — Errors become rules; preferences compound; every session makes the next one better

  • 🚀 10-minute setup, 31 roles supported — From CEO to IC, tell it your role and it scaffolds everything. No coding required

EXPLORE THE FULL GUIDE →

That’s a wrap.

As always, the journey doesn't end here!

Please share and let us know what you would like to see more or less of so we can continue to improve your Product Tapas. 🚀👋

Alastair 🍽️.

Reply

Avatar

or to participate

Keep Reading

© 2026 Product Tapas..
Report abusePrivacy policyTerms of use
beehiivPowered by beehiiv