· The Board Briefing
One decision.
How fast skills depreciate now.
Aggregated forecasts, 2024 to 2026.
Frontier lab leadership on record statements. Cross referenced with industry skills decay studies and L&D vendor data.
Quarterly is the new floor.
Any reskilling budget approved annually is already out of date when signed. Move budget to quarterly draw down with continuous topic refresh.
The largest infrastructure build in history.
62 percent compute · 24 percent energy · 14 percent talent.
The talent slice is the smallest line on the balance sheet, yet it is the binding constraint on every other allocation.
Talent will bottleneck, not silicon.
Companies that underinvest in workforce relative to compute will find their data centers idle. Sequencing matters more than scale.
Task economics, restructured.
−18% at the top. −85% at the bottom.
Strategic narratives compress least. Repeatable knowledge work compresses most. The economic gravity points downward, not horizontally.
Reallocate to judgement work.
If 68 percent of cost for seven major categories evaporates, the remaining workforce reorients around tasks that do not compress: trust, taste, orchestration, governance.
How much of the org chart moves to bot.
Leading firms hit 52 percent.
Frontier companies report bot migration ratios above 50 percent of task volume by mid 2027. The gap with median firms widens, not closes.
Headcount stays. Composition changes.
Most boards budget for headcount cuts. The data points the other way: same headcount, different skill mix, higher output. The reset is in role design, not severance.
Why most strategies never stick.
57 of 500 have a named AI workforce lead.
Often buried two levels below the CEO. Only nine companies have an executive level CDO equivalent reporting to the CEO directly.
Naming the role is half the work.
The decision is not whether to install the role but who reports to whom. Under CHRO drifts to L&D. Under COO drifts to automation. Under CEO compounds.
Where the evidence points.
Taken together, the five figures converge on three implications for organisations. They are described here as observations drawn from the source material, not as instructions.
Accountability tends to be undefined.
The episodes repeatedly note that workforce AI rarely has a single named owner. Some organisations describe a Chief Digital Orchestrator, a senior role that owns AI agent operations and the operating rhythm. Where the role sits in the structure appears to shape outcomes.
Roles are mapped, not just counted.
A recurring practice in the dialogue is plotting roles against options such as build, buy, borrow, or automate. The output discussed is a capability map and a view of gaps, rather than a headcount target.
Adoption is incremental and measured.
Speakers describe migrating a small number of high volume workflows first, running them under human oversight, and measuring cost per task and quality before extending. Adoption is presented as staged rather than wholesale.
One body of evidence.
Skills are depreciating faster.
The source material points to shorter skills half lives and rising cost compression on definable knowledge work. The pattern is consistent across episodes rather than tied to a single claim.
Task volume migrates to agents.
Speakers forecast a growing share of task volume moving to agents by 2027, with leading firms further ahead. These figures are projections, not settled outcomes, and are presented as such.
Ownership is mostly unassigned.
Across the dialogue, few organisations have a single named owner for workforce AI. The evidence raises the question of accountability rather than resolving it.
A synthesis, not a recommendation.
Five briefings distilled from over 250 episodes of Moonshots with Peter Diamandis and the Moonshot Mates. Source noted below. Compiled by TALENT on demand.