Netflix Chief Product and Technology Officer Elizabeth Stone uses the phrase "one zoom out": before solving a request, step back and examine how it affects the surrounding systems, teams, customer experience, and business result.
That idea frames this week's essay. AI can help people take on work outside their specialty, but companies still need experts to define how the work should be done, handle exceptions, and understand how a change in one system affects the rest.
AI Changes What Specialists Are For - The main essay. Netflix is hiring more systems thinkers who can decide which technologies, data definitions, and operating rules should be shared across teams. As agents act across several business systems, repeated local solutions become harder to sustain. If each team invents its own integrations, access paths, and interpretations of shared data, the pieces become difficult to operate together.
This is not the end of Salesforce engineers, NetSuite developers, Workday experts, or other application specialists. AI changes their contribution. The repeatable parts of specialist knowledge can become templates, shared components, supported integration patterns, and rules that other teams and AI tools can reuse. Specialists can then focus more of their time on exceptions and decisions that require experience and judgment.
Netflix's designers illustrate the shift. As AI lets more non-designers build product experiences, specialists spend more time on design systems, templates, and shared product language. Deep design still matters on the most important work, but specialists also define what good looks like and encode part of that judgment for others.
Business applications need the same move. A customer account change may begin in Salesforce but also affect billing and support. The specialist still owns Salesforce's data model, permissions, and release risk. Teams must agree on customer identifiers, which fields move between systems, and what happens when systems disagree. Once those decisions become a supported pattern, teams can reuse the logic instead of rebuilding it for every project.
AI also helps specialists stretch into nearby work, especially when they already have relevant experience. In an experiment at IG Group, marketers using AI nearly matched experienced web analysts when producing investment articles, while developers and data scientists farther from the assignment scored lower. AI helped everyone organize the work, but adjacent expertise still mattered to the finished result.
For CIOs, that means expanding specialists into connected workflows without expecting them to master unrelated fields. Bring the Salesforce expert into the earlier conversation about how a customer change moves through sales, billing, support, contracts, and data. Over time, the role grows beyond maintaining one application and toward helping the company decide how several parts of the business should work together.
Also in this issue:
- Patterns & Signals - High-risk agents need isolated environments and shutdown plans, while total AI spending can rise even as individual services get cheaper.
- Signals to Watch - AI time savings are being absorbed by more work and review, automated interviews are moving into late-night hours, familiar signs of AI writing are fading, and one-time training demos are not enough to change behavior.
- Meanwhile... - Rice University researchers can print working electronic circuits directly onto delicate materials, including living tissue.
- What I'm Consuming - Stanford's virtual biotech of 37,000 agents, organizational change in procurement, agentic development loops, BCG's leadership playbook, and product quality after coding gets easier.
- After Hours - Christopher Nolan's The Odyssey in 70mm IMAX, a practical-effects spectacle that feels like a defense of filmmaking as a human art.
Before solving the request, take one zoom out. Ask how it affects the customer, which systems and teams it touches, and whether it improves the business result. Then decide what can follow a common pattern and which decisions still need an expert.