One Sentence Takeaway
Stop assigning AI tasks and start working with it as a reasoning participant, then lead the whole system through orchestration: track capabilities, evaluate outcomes, reallocate responsibility, and retune interactions until Clock Drift turns into clock drive.
Summary
At the IT Forum CIO Summit in Praia do Forte, Neil Redding names what broke: AI can now execute faster than most organizations can decide, a gap he calls Clock Drift. Samsung’s early ChatGPT ban was a governance story; the newer signals are two non-technical reporters building a working project-management tool in about an hour for $15, and a lawyer at Anthropic shipping a legal review plugin his whole team runs. The unsanctioned version of that plugin already exists inside most companies: shadow execution, with roughly half of employees using unapproved AI tools. If an organization cannot decide at the speed of execution, someone else executes its strategy first, which makes the operating model itself an attack surface. The underlying shift is from tool to participant. Reasoning changed the relationship: tools do tasks, participants co-create outcomes, and the ceiling on results is now the quality of human intent, framing, and context. Interactions compound into dynamics such as emergence and reciprocity, shown live as Neil builds a context layer for his own business with Claude Code, which questions his design and proposes a better one. Proprietary context is the durable moat, so maintaining a shared context layer is a 2026 priority. The leadership model that follows is orchestration: track evolving capabilities, evaluate outcomes, reallocate responsibility between humans and agents, and retune the interaction patterns. Done well, deliberate decisions ride on AI-speed execution, and Clock Drift becomes clock drive.
Key Ideas
Clock drift is an operating-model problem
Work inside organizations still moves through specs, reviews, and committees, while AI-enabled work on everyone’s laptop runs at a different tempo. Neil defines Clock Drift as AI executing faster than the company can decide, and stresses that the drift comes from organizations failing to move with AI rather than from AI moving too fast. Samsung’s early ChatGPT ban, the CNBC reporters’ $15 vibe-coded tool, and the roughly 30% software-sector drop that followed show the gap at market scale.
Shadow execution is the natural response
Mark Pike, associate general counsel at Anthropic and not an engineer, spent an afternoon with Claude building a legal review plugin his whole team now uses, covering marketing copy review, contract redlining, and privacy impact assessments, cutting turnaround from days to hours. His version is sanctioned. Recent studies put about 50% of employees on unapproved AI tools, which is what happens when the decision clock runs a quarter or a year behind the execution clock.
Tools do tasks; participants co-create outcomes
Last year, prompting skills were enough. Reasoning changed that: the moment AI reasons rather than just responding to commands, it stops being a tool and becomes a participant. A task prompt asks for a known deliverable; a participant prompt supplies intent, constraints, and what to watch for, and returns judgment. Anthropic’s Economic Index found prompt sophistication and response quality nearly perfectly correlated, so the ceiling is the quality of human intent, framing, and context rather than model capability.
Interactions compound into dynamics
The nature of the interactions determines the nature of the outcomes. A traditional loop of brief, report, committee, and escalation loses signals over weeks; an evolved pattern of briefed intent, continuous analysis, and live options converges in days with less rework. As patterns compound, the system shifts from discrete to cumulative, prompted to reciprocal, planned to emergent. Matt Schumer’s viral post about no longer being needed for the technical work of his job marks that emergence, alongside GPT-5.3 Codex helping create itself.
Context is the moat and the bottleneck
Larger context windows and sustained reasoning are behind the recent jump in long-horizon autonomous work, visible in METR’s evaluation data. What differentiates a business is its proprietary context: customer relationships, internal decisions, institutional knowledge, much of it poorly shared, including voice notes and chat insights. Neil builds a context layer for his own business live with Claude Code, then reframes context engineering as relationship engineering across four challenges: coherence, trust and delegation, guardrails and governance, and optimal human review.
Leadership is the orchestration of participant dynamics
Humans are moving from mostly doers to mostly deciders, yet Deloitte finds 84% of companies have not redesigned a single job around new AI capabilities. The practice Neil prescribes is delegate first, then a continuous cycle: track evolving capabilities, evaluate outcomes, reallocate responsibility between humans and agents, and retune interaction patterns. His orchestration dashboard shows the cycle live as an upgraded research agent takes two financial tasks from analyst Casey Kim, with a human confirming and Casey moving to oversight.
Notable moments:
A show of hands on two prompts for the same Q2 sales data: the room recognizes both the task version and the participant version, and more hands go up for the participant framing.
Claude, working inside PowerPoint, reviews Neil’s own Clock Drift slide for an audience of top Brazilian CIOs and advises him to substantiate the concept first; Neil tells the room, “keep me honest.”
During the live Claude Code build, Claude pushes back on Neil’s tile-based design (“Good challenge”) and proposes an aggregated landscape with drill-down layers that becomes round two of the context layer.
A callback to the event’s own survey: 34% of the room named orchestrating digital transformation among their top concerns for 2026. Neil’s recurring “tamo juntos” lands with the Brazilian room.
In Neil’s words:
“If your organization can’t decide at the speed of execution, someone else is going to execute your strategy before you can even approve it.”
“Leadership is the orchestration of participant dynamics.”
“Leadership is orchestration. And orchestration turns Clock Drift into clock drive.”
