How an agentic AI course changed my mind about AI adoption - Shots.net

TEXT START: Rather than seeing artificial intelligence as a threat to her job, FIN Studios Head of Studio, Tabitha Onofri, chose to understand it.

The Dissection

This is an adoption manual wrapped in career reassurance. It correctly identifies immediate efficiencies: connecting legacy systems, removing administrative repetition, and concentrating creative workers on higher-value tasks. But it treats the first layer of automation as the endpoint.

The article frames AI as an assistant that liberates creatives from “filler.” Under the Discontinuity Thesis, that filler is the entry point. Once agents coordinate contracts, approvals, logs, briefs, treatments, revisions, and production workflows, they accumulate operational context. The system does not merely save human time; it learns enough of the workflow to reduce how many humans are required to execute it.

The Core Fallacy

The central error is confusing task preservation with job preservation.

Automating repetitive work does not guarantee that the same number of directors, editors, artists, or studio heads will remain economically necessary. It increases the output controlled by each surviving worker, compresses production timelines, lowers client costs, and intensifies competition. The “extra time” created for creatives is also leverage for employers to reduce headcount, expand workloads, or demand cheaper output.

The article assumes human judgment is a permanent moat. Under P1, cognitive systems improve across precisely the judgment, coordination, and iteration layers that creative industries use to justify employment. Under P2, the industry cannot maintain a stable human-only domain once AI-enabled firms can deliver comparable work faster and cheaper. The result is P3: fewer people retain access to economically necessary labor, even if the remaining work still contains human taste or oversight.

Hidden Assumptions

• Productivity gains will be returned to workers as creative freedom rather than captured by owners as margin, speed, or headcount reduction.
• “Human judgment” is scarce, durable, and resistant to automation.
• Creative quality will remain expensive enough to support existing staffing levels.
• AI will automate only low-status administrative work and stop before reaching core creative production.
• Firms will adopt AI to improve jobs rather than to redesign and eliminate them.
• Human-centred implementation determines the economic outcome. It does not; ownership, competition, and cost superiority do.
• Training workers to use AI gives them control over the capital. It usually makes them more effective operators of someone else’s system.
• A gradual crawl-walk-run rollout changes the destination. It only changes the speed of arrival.

Social Function

Primary classification: transition management and ideological anesthetic, with a substantial partial truth.

The partial truth is operationally useful: skepticism, staged implementation, system integration, and verification do matter. Bad automation can scale bad processes. But the article uses this truth to smuggle in a false conclusion—that careful adoption preserves the existing occupational structure.

Its social function is to help professional workers accept AI without confronting the ownership question. It converts a displacement problem into a learning problem: if workers are threatened, they supposedly need more knowledge, flexibility, and willingness to adapt. That shifts responsibility for structural dispossession onto the people being dispossessed.

The Verdict

This is a competent workflow memo and a weak economic diagnosis. Agentic AI may not immediately erase the author’s title or eliminate every creative role. It will erode the staffing, bargaining power, junior training ladder, and price structure that make those roles broadly available.

The article mistakes the first visible effect of AI—removing drudgery—for its terminal effect: removing the need for as many humans to produce the same commercial output. It is not evidence that AI preserves creative work. It is evidence that the industry has begun building the machinery that will make creative labor cheaper, thinner, and progressively less necessary.