Change the work before the job title
AI adoption gets interesting where work gets stuck: unwritten rules, awkward handoffs, changing roles, and the people who can build something useful.
April 29, 2026

We keep coming back to the space between tasks: the request someone has to explain again, the unwritten rule, the report waiting on another team. This issue’s stories suggest that’s fertile ground for AI adoption. Before rewriting the org chart, get curious about where work gets stuck. A grand new title won’t fix a lousy handoff.
Amazon’s “builder” experiment leaves a career question
Reuters reports in USA Today that Amazon is replacing product workers’ titles at Ring and Blink with “builder” roles, prompting employee concerns about advancement even as Amazon says compensation and promotion paths remain unchanged.
Our take: A title can be a status trophy, but it can also tell people what progress looks like. If you flatten the labels, make the career path more visible: show what greater scope means and how it earns a promotion. Otherwise, “everyone’s a builder” leaves employees guessing which contributions count.
Game studios find the limits of handing everyone a chatbot
Wharton’s study of 20 game-studio interviews describes workflow automation stalling on unwritten knowledge, with progress coming when people used AI to handle adjacent tasks.
Our take: Buying everyone a chatbot leaves a lot of organizational nonsense intact. We would pay attention to the person who knows exactly why a report takes three departments to finish, then give them help improving that handoff; they have context a generic automation demo doesn’t.
“Reshaped” jobs still require a plan
BCG’s microeconomic model estimates that AI could reshape 50–55% of US jobs over two to three years, while explicitly distinguishing its analysis from an unemployment forecast.
Our take: “Your job will still exist” is a pretty thin development plan. If routine work disappears and judgment becomes a bigger part of the job, people need practice, support, and a realistic workload before their performance expectations change.
The AI builder needs to understand the business
Joe Galvin argues in Inc. for AI builders who combine technical capability with business context to create tools for individuals and small teams.
Our take: We like the role; we’re less sold on his ladder ending with autonomous agents as the peak of AI skill. A builder who fixes a recurring reporting headache and knows when to stop automating is more useful than someone determined to put a bot in charge of everything.
Stop giving people an “AI resistant” identity
Rebecca Hinds writes in Inc. that leaders’ narratives and labels can constrain how employees approach AI and skill development.
Our take: “AI resistant” is a convenient label when you haven’t asked what’s getting in someone’s way. Ask which part of the work they want help with and which part they care about doing themselves; that gives you a much better starting point than another pep talk about embracing change.
One useful move
Have a request you need to hand to someone else? Paste your rough notes into your usual AI tool and ask:
My notes: {{request_notes}} Turn these notes into a clear request I can send to a colleague. State what I need, why I need it, and what a finished result should include. Keep it short. Use only the details in my notes, and flag anything missing instead of filling it in.
Fill in any missing detail that matters, check that it asks for the right thing, and use it.
Something fun
Ask your chatbot to pitch a cozy video game set in the least exciting room in your office:
Design a cozy video game set entirely in an office supply cupboard. Give it a title, one oddly lovable character, and three tiny quests. The stakes should be absurdly low, but the trailer narrator should sound convinced this is the adventure of a lifetime.
We nominate a stapler who dreams of becoming a hole punch. Please respect the career pivot.
Built in the past two weeks
We built a small publishing helper that converts a Markdown guide to HTML and puts the formatted content on the clipboard for our learning platform. The aim is to bring headings, lists, and links into the editor in one paste, with a final formatting check before publication.
