Observed Signal · Jul 1, 2026 · Analysis · Source: The Business Engineer · Impact: 4/5 · Sentiment: Positive
The Harness Shift: Agentic Surfaces Replacing Chat
This analysis synthesizes usage data published by OpenAI (Codex report) and Anthropic (Economic Index) to argue a phase change: AI is shifting from conversational assistants to agentic surfaces or “harnesses” that execute delegated workflows. Both labs’ measurement systems show the same pattern — conversation-based metrics are breaking down as users increasingly deploy multi-step agents. Key empirical signals include OpenAI employees routing 99.8% of internal work through Codex, organizations showing 17.3% of users touching agentic surfaces but 63.3% of output flowing through them, and individuals at ~0.7% active but generating 16.5% of agentic output. The piece frames this as a platform war (consolidated universal harness vs. embedded proliferated harnesses), highlights SKILL.md as a primitive, and warns of risks from training methods that reduce model diversity.
Uses empirical usage data from two leading AI labs to argue a structural industry shift: value and pricing power migrate from models to orchestration/harness layers, increasing compute demand and creating a platform-standard battle; implications affect product design, enterprise adoption and infrastructure.
Track OpenAI Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- OpenAI’s Codex accounts for 99.8% of work-related output inside OpenAI across Codex and ChatGPT combined (per the article).
- OpenAI populations active on the agentic surface (trailing 28 days): OpenAI workforce ~98%, organizations 17.3%, individuals 0.7%.
- Share of output routed through the agentic surface: OpenAI workforce 99.8%, organizations 63.3%, individuals 16.5%.
- External skill use rose from 5.4% to 26.6% in ~three months; inside OpenAI skill use is 96.2%.
- OpenAI internal adoption timeline: engineers surpassed 50% agentic usage by January 2026 and 90% by March 2026; some non-technical functions jumped from ~20% to 75% within a month (early 2026).
Connected Companies & Entities
3 Entities mapped“OpenAI’s _The Shift to Agentic AI: Evidence from Codex_ (report cited) and multiple internal adoption statistics (e.g., Codex accounts for 9...”
“Anthropic’s latest Economic Index is cited, including the statement that 'a chat transcript no longer fully captures how people are using AI...”
“Published on the Business Engineer site and signed by author Gennaro Cuofano ('With massive ♥️ Gennaro Cuofano, The Business Engineer')....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Why I Ended Up in the AI Harness
Gennaro Cuofano's essay describes a multi-year sequence of AI inflection points that forced a shift from operating inside chat interfaces to directing autonomous, multi-agent systems — a "harness." He outlines four scaling eras since 2020 (pre-training, test‑time reasoning, agency, orchestration/swarms), cites key technical developments (ChatGPT's 2022 release, OpenAI's o1 model, Anthropic's Model Context Protocol/MCP) and argues value is migrating outward from models to orchestration, operations and outcome-based services ("AGaaS"). Cuofano frames authorship — wanting outcomes, choosing tradeoffs, and taking responsibility — as the only durable human role as capabilities commoditize. The piece situates the orchestration/swarms era as current (June 2026) and links the change to new business models, form factors, and faster inflection-point compression.
Agent Harness Evolution and the Attention-Interface
The article analyzes how AI agents improved around Christmas 2025 due to co-evolution of large models and the surrounding "agent harness" (environment, tools, context, and guardrails). It traces stages from prompting-based loops (ReAct) through premature autonomy (AutoGPT/BabyAGI), retreats to human-in-the-loop (IDEs/Copilot), and the crossover where models outpace harnesses (Claude Code, Feb 2025). Empirical results (Harness-Bench, OpenAI ARC-AGI-3) show harness design can materially change agent performance. The author argues models gradually absorb harness capabilities, leaving a remaining harness focused on human-centric concerns (permissions, trust, attention). The piece predicts companies will ship explicit human attention policy surfaces as the next standard harness component.
The Harness Society: AI as an Economic Substrate
An analytical essay arguing that AI has moved from a standalone industry into a pervasive economic substrate — a "harness society" — which reorganizes prices, infrastructure, talent markets and organizational design whether firms or workers opt in or not. The author synthesizes multiple data points: an Anthropic study of 400,000 Claude sessions (showing humans do ~70% of planning while models perform ~80% of execution and a 5× expert/novice output gap), Gallup findings that non‑adopting tech workers face roughly 3× the layoff risk, Microsoft’s shift toward per‑task AGaaS pricing (Copilot Cowork at $0.01/task), and company case studies (Block as an "amplifier," State Farm as a "shrinker"). Core claims: the transition to agentic AI creates an outcome‑pricing problem (AGaaS requires precise outcome definitions), hollowing of mid‑career roles, a broken apprenticeship pipeline, and governance as a scarce political resource.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.