You’re reading The Steady Beat, a weekly pulse of must-reads for anyone orchestrating teams, people, and agents across the modern digital workplace – whether you’re managing sprints, driving roadmaps, leading departments, or just making sure the right work gets done. Curated by the team at Steady.
Own the output
Dave Kellogg nails the awkward dance we’re all doing with AI at work: everyone’s using it, nobody quite knows what to disclose, and “the AI wrote that part” has become the new “the dog ate my homework.” His framework: you own whatever you present, period. AI is a calculator, not a co-author. Just as nobody credits Casio when they nail a financial model, you shouldn’t hide behind Claude when your board summary lands well, or blame it when your analysis falls flat. The real danger is the cognitive atrophy that comes from letting AI do your thinking. Loading a board deck into Claude to generate discussion questions? Smart use of the tool. Presenting those questions without reading the deck yourself? That’s abdication. Kellogg draws the line between scaffolding work (agendas, formatting) and substantive analysis, where AI can handle the former while you focus on the latter. But the moment you stop challenging and genuinely engaging with the output, you’ve crossed into intellectual outsourcing.
— Kellblog, 6m, #ai, #leadership, #strategy
Death of the middle
Meta and Block are placing matching bets that middle management as we know it is obsolete, just using different euphemisms for the funeral. Meta’s Reality Labs has reorganized everyone into “AI Builders,” “Pod Leads,” and “Org Leads,” with AI now handling performance reviews and promotion decisions that used to justify a manager’s existence. Block went further, killing the manager title entirely in favor of “player-coaches” who ship work alongside their teams instead of directing from above. Jack Dorsey wants a structure flat enough that he can connect directly with all 6,000 employees, a claim that should make anyone who’s studied org design raise an eyebrow. The shared thesis is that fewer layers mean faster decisions and less bureaucratic friction. And there’s real logic to it, since coordination overhead is the silent tax on every growing company. But neither announcement addresses who handles the messy human work that good managers actually do, like resolving conflicts, developing careers, creating psychological safety, and translating strategy into context. These aren’t tasks you can hand to an AI or bolt onto an IC’s already-full plate. The real risk is that they’re confusing the title with the function, and they’ll discover the difference the hard way when their best people start leaving because nobody’s watching out for them.
— Business Insider, 5m, #management, #ai, #leadership
The quiet refusal
Remember when the big worry was employees sneaking ChatGPT past IT? The script has flipped. A new WalkMe survey reveals that 80% of enterprise workers are now avoiding or outright rejecting the AI tools their companies deployed for them. 54% bypassed them in the last month alone, choosing to do the work manually instead. The trust gap is wild: only 9% of workers trust AI for complex decisions versus 61% of executives, a 52-point chasm that should alarm anyone responsible for rolling these tools out. Multiple leaders independently reached for the same metaphor: it’s like handing everyone a Ferrari without teaching them to drive or giving them fuel. And the productivity math is brutal. Workers lose 51 days per year to technology friction, almost perfectly canceling out the 40-60 minutes daily that skilled AI users save. Workers aren’t rejecting AI ideologically, but struggling with inadequate training, unclear use cases, and missing support. 34% don’t even know which tools their employer has approved. The companies that’ll actually get somewhere with this are the ones figuring out the human-AI handoff, specifically when the person should act, when the agent should act, and how trust gets built between the two.
— Fortune, 8m, #ai, #transformation, #leadership
Curious but wary
The generation that was supposed to be AI-native is turning skeptical, fast. A new Gallup survey of over 1,500 Americans ages 14 to 29 reveals that while half of Gen Z uses AI daily or weekly, their feelings about it are curdling. Hopefulness dropped from 27% to 18% in a single year. Nearly a third now say the technology makes them feel angry. And close to half of working Gen Z-ers believe AI’s workplace risks outweigh its benefits, an 11-point jump from last year. They’re worried about entry-level jobs disappearing before they can land them, about their creativity and critical thinking atrophying, and about outsourcing the very skills they haven’t finished developing yet. Adoption hasn’t grown either; it’s flatlined despite broader access. For managers, this is a signal worth taking seriously. Curiosity was still the most common emotional response to AI among this group. But they’re arriving with real anxiety about what AI means for their careers and their humanity. If your onboarding strategy assumes younger workers will enthusiastically adopt every AI tool you throw at them, you’re about to learn otherwise. Meet them with honest conversations about where AI actually fits, because “just use the tool” isn’t a strategy when trust is this thin.
— The New York Times, 5m, #ai, #leadership, #management
Bad code is a choice
Bram Cohen, the inventor of BitTorrent and not exactly a Luddite, wants you to know that “vibe coding” has jumped the shark. The trend where developers deliberately refuse to look at their own code, treating AI as a black box that magically produces working software, is a philosophy that guarantees technical debt at scale. Cohen watched it happen firsthand: massive code duplication between agents and tools went completely unnoticed because developers treated examining the codebase as a violation of vibe coding principles. And pure vibe coding is a myth anyway. You’re still building infrastructure, writing plan files, writing system rules. You’re just pretending you aren’t making technical decisions. Cohen’s alternative is more honest: conversational code review where you discuss specific problems with AI, examine examples together, and provide architectural guidance while letting the AI handle implementation. AI is terrible at spontaneously noticing problems but excellent when given direction and context. Bad software, Cohen argues, is a decision you make. And refusing to look at the code doesn’t make it someone else’s decision, it just makes it a worse one.
— Bram Cohen, 5m, #engineering, #ai, #productivity
Echo of the week
Echoes are AI agents in Steady that automatically gather and deliver work context to teams on a schedule, answering recurring questions about progress, capacity, and coordination so you stop burning hours assembling the same information manually.
Big-picture intentions – Stay up to date on the latest goals across your organization. This Echo automatically generates monthly summaries of newly created goals, giving department heads, product leaders, and executives visibility into emerging initiatives before teams advance too far. Spot overlaps, coordinate resources, and catch misaligned efforts before they gain momentum.
The lightweight teamwork OS
Teams rely on two coordination loops to function: a big-picture loop connecting plans to progress, and a ground-level loop keeping teammates in sync.
Problem is, status quo approaches to running those loops are an incomplete, inconsistent, and inefficient tangle of meetings, emails, chat threads, dashboards, and manual toil.
Steady is the teamwork OS that runs both loops for you. Purpose-built agents continuously distill updates and activity into personalized intelligence that keeps everyone aligned and informed automatically.
The outcome: high-performing teams that deliver better work, 3X faster.
Learn more at runsteady.com.
