Agents are going to change how we work and how we get compensated. Every time we speak to a leader at the forefront of working with agents, we hear the same thing: power users are getting massive productivity boosts, and many others are struggling to even break even with the time they put into the tools. Those that do figure out how to turn agents into an infinite workforce are going to enter a scenario much like the Hunger Games — walk out infinitely successful, or fall back to normality. Coinbase restructured its entire organization around people who can operate this way, cutting 14% of staff to rebuild around agent-capable teams. ClickUp cut 22% and announced million-dollar salary bands for whoever could perform at that level.
Charlie is Chief Design Officer at Atlassian, where he leads design across Jira, Confluence, Rovo, and the newly announced Dia browser.
Why 783 tab interactions a day means even tiny friction changes produce outsized aggregate gains — and where to look first
The 25-year holy grail of adaptive interfaces is technically solved — what remains is the design question of how much is right for teams
Why structured objects (goals, strategy, people) beat expensive inference — and why most vendors are paying more for worse results
How Atlassian’s design system serves agents and humans from the same object with 10% variation — and what the 10% tells you
Why vibe coding raised the floor so everyone can build, which is exactly why the ceiling on what design must deliver also rose
Why video captures intent that text never can — and how Atlassian is encoding it into the Teamwork Graph
“It’s no longer a technology limitation — it’s a design question. Where is the balance point? And that’s a good place to be.” — Charlie, Chief Design Officer, Atlassian
Listen now: Spotify | Apple Podcasts | YouTube
Jonathan Su is Chief Product & Technology Officer at Procurify, the platform thousands of finance teams use to approve, pay, and audit every dollar that moves through the business. He is leading Procurify’s full shift to AI-native product development.
Why 35% of finance leaders say trust — not model capability — is the biggest factor in whether agents actually ship
The operating model most companies skip: governance, audit trail, single source of truth — before the agent touches work
What AI ROI actually looks like — 63% time savings, 60% better data accuracy, plus the business KPIs that prove it
Why value is moving up the stack as frontier models commoditize generic intelligence — workflow, context, data, distribution
How procurement teams redesign workflows around agents instead of tacking AI on top of an already broken process
The hire that beats 20 years of experience: grit, taste, judgment, and the ability to learn in 4-month cycles
“Managing an agent is more than just using a tool. It’s sort of like supervising labor.” — Jonathan Su, Chief Product & Technology Officer, Procurify
Listen now: Spotify | Apple Podcasts | YouTube | Episode page
Thanks for reading Product Impact Reports | Improve Your Success in AI Era! Subscribe for free to receive new posts and support my work.
Every CEO Will Post a Layoff Notice Like This. Here Is Why.
Brian Armstrong’s May 2026 Coinbase memo reduced headcount by 14% and announced the company was “rebuilding Coinbase as an intelligence, with humans around the edge aligning it.” Most coverage treated it as a headcount story. The three structural principles inside the same letter — five management layers maximum, no pure management roles, AI-native pods — trace directly to Andreessen’s coordination tax thesis, Paul Graham’s Founder Mode, and Sam Altman’s Intelligence Age argument. The companies moving to this model measure employees against leverage — the ratio of organizational impact to human judgment applied — not against output volume.
The jobs being eliminated are not coming back when the market turns. Most organizations are not telling their people this. The ones who find out too late are the ones who didn’t read the memo.
Read: Every CEO Will Post a Layoff Notice Like This. Here Is Why.
The AI Job Apocalypse Won’t Happen. Here’s What Will.
Ezra Klein made the case that mass AI unemployment is unlikely — and the research largely agrees. Technological unemployment of that scale has been predicted repeatedly since the Industrial Revolution and has never arrived. Brittany’s analysis goes one step further: that argument is true and completely beside the point for the generation whose career path just became structurally cheaper. The job count won’t collapse. What disappears is the entry-level path and the implicit promise that time in a role translates to seniority.
The real damage is happening one level below the headlines: the entry-level path is gone, the apprenticeship model is gone, and time in a role no longer buys you what it used to. The cohort that entered the workforce in the last three years is the first to feel this fully.
Read: The AI Job Apocalypse Won’t Happen. Here’s What Will.
Playbook for Knowledge Workers to Survive the AI Jobpocalypse
Your expertise was not commoditized. Your deliverables were. That distinction carries a 17-minute playbook. The Stanford AI Index 2026 reports a 20% employment decline for software developers ages 22–25 since 2024. One in three employers expect workforce reductions in the next year. The Coinbase memo gave the restructuring template. What nobody is publishing is what to do about it at the individual level — not “learn AI skills” and not “everything will be fine,” but an honest capability matrix and four operational paths through a tough market.
The four paths — navigate the reorg, move to a better-positioned org, pivot to an adjacent domain, or go independent — each matched to a real read of your capability profile and organizational reality. For UX researchers, digital marketers, and project managers specifically, the article names the pivot that makes the most sense in each role rather than offering generic advice that applies to none.
Read: Playbook for Knowledge Workers to Survive the AI Jobpocalypse
How Apple’s WWDC 2026 Siri Announcements Change How We Build for iOS
Apple spent two years turning a promise into an embarrassment. WWDC 2026 was the repair — and buried inside the design language fixes and the Siri revamp is a developer unlock that changes the surface area for building AI-powered iOS experiences. Siri AI, now powered by Google Gemini and built on a “data is only used to execute your request” architecture, opens the Apple ecosystem — messages, emails, photos, health data, calendar — to approved developers for the first time. The personal context layer that ChatGPT and Claude have been trying to earn through permission is now available through App Intents, on-device, discarded after each request.
By choosing Google over OpenAI, Apple delivered a structural blow to Sam Altman’s consumer AI ambitions that no product announcement could match. For iOS builders, the five integration paths are specific and buildable today — and the window before iOS 27 ships is the time to register your App Intents.
Read: How Apple’s WWDC 2026 Siri Announcements Change How We Build for iOS
Eight AI ROI Frameworks, Explained: Who Should Use Each One and When
The Forrester TEI number in your vendor pitch — 116% three-year ROI for Copilot, 416% for Google Workspace with Gemini — is built on a composite organization that doesn’t exist. It is a capital defense instrument, not an internal measurement tool. Most teams are applying it to the wrong problem.
AI Value Acceleration published the AI ROI Framework Atlas this month — the first document to profile all eight distinct published AI measurement systems against the same six-field grid. The finding that reframes every AI investment case: the right frame isn’t ROI. It’s a VC portfolio. Safe bets that fund the machine. Big bets that force structural change. Moonshots that generate the organizational lessons safe bets can’t.
Read: Eight AI ROI Frameworks, Explained: Who Should Use Each One and When | Get the Atlas: aivalueacceleration.com
AI Strategy Resources
Four signals from the research and practitioner layer this week.
Why Employees Aren’t Transparent About Their AI Usage — HBR, June 2026. Workers are hiding their AI use from managers out of fear of being seen as less competent or replaceable. Organizations that don’t fix this will misread their own adoption numbers.
Afraid of Your Own Culture? Romano Roth on the Cybernetic Enterprise — Zühlke Group, May 2026. Roth’s analysis: 98.4% of enterprise AI systems are execution harness, 1.6% decision logic. Without a feedback loop protecting human judgment, AI-generated work degrades in quality over time — what he calls “Slop Creep.”
Microsoft 2026 Work Trend Index — Microsoft, 2026. 80% of “Frontier professionals” are producing work they couldn’t a year ago — but this group is a minority and the gap between them and median performers is widening, not narrowing.
Stanford AI Index 2026 — Economy Section — Stanford HAI, 2026. Software developer employment for ages 22–25 is down 20% since 2024. One in three employers expects further reductions this year. The pipeline below mid-career is thinning structurally, not cyclically.
If you’re an AI leader, adoption is no longer the goal. Value is. Most organizations are still measuring the wrong thing — tracking who’s using the tools instead of what the tools are actually delivering. AI Value Acceleration works with leaders to close that gap. If your AI program is producing activity but not outcomes, let’s talk.
Thank you for your support. As always feel free to send any questions to arpy@ph1.ca or comment on this post. They’re very helpful for us to know what topics and problems matter most.
Browse all two seasons of episodes at productimpactpod.com.
Thanks for reading Product Impact Reports | Improve Your Success in AI Era! Subscribe for free to receive new posts and support my work.