The people and the money are in different buildings.

All scores are modeled estimates. Every claim is tagged by confidence level. Full methodology below.

What could break this

Faster packets ≠ faster decisions. A compressed borrowing-base package still waits for the VP's calendar, the lender committee, and the bank's own reserve engineers. Maybe 30% of cycle time is work product. The rest is organizational friction AI doesn't touch.

Bad source data breaks everything. Well logs from the 1970s were hand-transcribed. Conflicting spreadsheets, stale type curves, vintage assumptions. AI on bad data produces bad results faster.

Weakly trusted outputs increase review burden. If the model's work isn't trusted, reviewers spend more time checking than they saved. The net effect can be negative until confidence builds.

How robust is the ~90% figure?

Below: what happens when we stress field-layer headcount and compensation — the two inputs most likely to be underestimated. Non-field employment estimates and compensation proxies are held constant; stressing those would require a broader sensitivity test. The pattern holds across the range tested because the compensation gap between layers dominates the math.

Rows: field employment multiplier (how much larger or smaller the physical layer headcount might actually be). Columns: field compensation adjustment. The base case uses the current model assumptions. Even at the extremes — 2× field headcount at 1.5× field pay — the above-field share still shows the majority of exposed dollars sitting outside physical operations.

Extended sensitivity: above-field parameter stress

What happens when we vary the parameters the above-field layers use? Rows: multiplier on all non-Physical layer bases. Columns: multiplier on all non-Physical compensation proxies. Both directions weaken the result by making above-field layers less dominant.

What drives the ~90% — and what survives stress

Three factors build the above-field concentration. The decomposition below shows how much each contributes (under the score-neutral employment model), and the stress tests show what happens when we deliberately weaken the most attackable assumptions.

Stress tests

The above-field share ranges from 74–96% across all stress tests (parametric Monte Carlo: 75.5–96.1%; anchor-row-only test: 74.2%). Even under the harshest tested specification (BLS-anchored rows only, excluding all formulaic estimates), roughly three of every four exposed wage dollars still sit above the field. The directional finding is not an artifact of the employment model, the compensation proxies, or the scoring method.

What could break this

344 of 404 employment estimates use a deterministic formula (layer base × tier multiplier) rather than observed data. Three tier bands (0.75×, 1.0×, 1.25×) assigned by stable hash provide structural spread, not an economic model. The compensation proxies are layer-level averages, not role-specific wages. If physical-layer headcount is substantially larger than modeled, or if above-field compensation proxies are too high, the ~90% figure compresses — but the stress grid shows it stays above 74% even when physical headcount doubles and physical comp increases 50%. The anchor-row test (74.2%, using 60 BLS/industry-sourced rows only) provides the hardest floor. The direction holds across all specifications. The exact percentage is scaffold-dependent — treat it as an interval (74–96%), not a point estimate.

What this does not claim

(a.k.a. the section most research puts in an appendix nobody reads)

The Jevons trap — and why G&A is the wrong denominator

Here's what most energy companies do first with AI: email drafts. Meeting summaries. Formatting. Which makes sense — nobody gets fired when a model hallucinates a meeting recap. But it misses the point by about 40:1.

A mid-size E&P spends $50–100M on G&ADerivationRange based on public filings. Pioneer Natural Resources reported ~$200M G&A in 2023 (including merger costs; ~$100M excluding). Diamondback Energy runs ~$0.69/boe cash G&A. Mid-market operators with 50–100K boe/d production typically fall in the $50–100M range depending on accounting treatment.. It spends $2–6B on development capital. That's a 30-to-1 ratio. The entire AI-for-energy pitch is aimed at the smaller number. It's like hiring a personal shopper to optimize your grocery bill while your investment portfolio picks itself.

The Jevons analogy is a hypothesis about demand: as analysis becomes cheaper, firms may choose to ask more questions. The workflow scenarios below illustrate that possibility. They do not demonstrate the size of the response or its effect on investment returns.

Which is why this is not an efficiency story. Saving 20% of G&A on $75M is $15M. A 2% improvement in capital allocation on a $3B development program is $60M — four times the G&A savings the industry's writing press releases about. The industry is optimizing the small number and ignoring the big one.

That's the leverage structure of this industry. A handful of knowledge workers control a capital budget that dwarfs their salaries by orders of magnitude. Dominator illustrates the cost of a development decision going wrong. This report does not establish that insufficient scenario coverage caused the result or that AI would have prevented it. Make analysis cheap, and the question changes from "can we afford to run more cases?" to "can we afford not to?"

The same allocation problem appears across the AI supply chain. The IEA estimates that capital expenditure by five large technology companies could rise roughly 75% from more than $400B in 2025. This is total company capex, not a clean measure of AI-only investment. The analytical task is to identify which sites, contracts, and infrastructure choices turn spending into productive capacity. IEA, April 2026.

The ownership problem

If the value is in the capital allocation decision — not the meeting summary, not the email draft, not the reformatted slide deck — then the person who owns that decision has to own the AI deployment. Not "approve it." Not "sponsor it." Own it. Be the user. Personally.

Right now, most energy companies hand AI to the IT department or to a "digital innovation" team two reporting layers removed from anyone who touches a capital decision. That's how you get chatbots for the help desk and summarizers for the weekly all-hands. Useful. Worth maybe $15M a year. And completely disconnected from the $3B development program where the real leverage sits. Everyone writes a LinkedIn post about digital transformation. Meanwhile the analysts are still manually building variance tables for a billion-dollar credit facility.

The CFO who reviews four scenarios that someone else built is a different person than the CFO who directs a system that runs forty and flags the three that change the decision. Same title. Completely different leverage on the capital budget.

The Concho example illustrates how consequential development choices can be. Public outcomes alone do not establish what analysis the team lacked. The organizational question is more useful: does the owner of the capital decision also control the quality, timing, and review of the analysis that supports it?

The practical version: energy companies that succeed with AI need someone in the room where capital gets allocated who also controls the AI deployment that feeds that decision. Call it what you want — Chief Decision Officer, VP of Decision Intelligence, the CFO who actually uses the tools. The title doesn't matter. What matters is that the person directing AI and the person signing the AFE are the same person, or sit next to each other. When they're three org layers apart, AI gets aimed at the G&A line. When they're the same person, it gets aimed at the capital budget.

A large capital-to-overhead ratio makes decision support worth examining. It does not establish a return on AI. Give a named decision owner responsibility for the pilot, an explicit quality bar, and a budget for independent review.

DRI case · Treasury and lender readiness

A mid-market E&P with a $1.2B reserve-based lending facility runs two borrowing-base redeterminations per year. Each cycle: 3 analysts, 3 weeks, rebuilding the same variance tables and lender Q&A packets from scratch. The lender asks the same fifteen questions every cycle. Most teams haven't structured last cycle's work product for reuse.

Illustrative pilot design. Start with the recurring packet: source comparison, variance tables, covenant extraction, and lender Q&A. Measure preparation and reviewer hours on completed cases, then evaluate held-out cases with the same acceptance criteria. Price model usage, software, integration, and setup. Any benefit to facility utilization or capital allocation needs separate evidence; it is not included in the labor business case.

What stays human: negotiation posture, lender relationship, representation of downside scenarios, signoff. What the system composes: source comparison, variance tables, covenant extraction, Q&A draft generation, package assembly — delivered before the cycle starts, not rebuilt from scratch. Full beachhead breakdown →

Role	Tasks AI compresses	Tasks AI amplifies	Tasks that stay human
Reservoir engineer	Decline curve fitting, type-curve generation, reserve report assembly, data gathering from production databases, variance commentary drafts	Scenario comparison (can now run 40 instead of 4), sensitivity analysis across price/decline/spacing, pattern recognition across analogue wells	Subsurface judgment calls, well spacing decisions, reserve certification signoff, risk framing for the board
Treasury analyst	Borrowing-base packet assembly, covenant compliance checks, lender Q&A drafts, amendment redlining, data reconciliation across systems	Exception detection (catches covenant breaches earlier), cross-cycle comparison (persistent memory across redeterminations), scenario stress testing on covenants	Lender negotiation posture, downside framing, final representations, signoff authority, relationship management
Land / title analyst	Division order calculation, lease abstraction, JIB exception screening, curative document comparison, ownership chain reconciliation	Cross-asset title pattern matching (flag similar defects across properties), historical exception memory (recalls curative outcomes from prior cycles)	Curative negotiation, title opinion judgment calls, counterparty relationship management, legal liability decisions
Regulatory analyst (utility)	Exhibit assembly, discovery response drafting, data request compilation, testimony support document preparation, precedent citation lookup	Cross-docket pattern analysis (identifies commissioner tendencies), IRP scenario modeling (more alternatives evaluated), consistency checks across multi-year filing history	Regulatory strategy, testimony delivery, commissioner relationship management, settlement negotiation, policy judgment
Lineman	Paperwork: daily job briefing forms, time entry, incident reporting templates, outage documentation	Predictive routing (AI optimizes storm response dispatching), outage pattern recognition (learns from historical restoration sequences)	Climbing, switching, grounding, live-line work, safety assessment, crew leadership, storm response decisions

The role does not vanish. The economics of the role change.

For a reservoir engineer, faster preparation may release time for interpretation and alternative designs. Whether that makes the work more valuable depends on the quality of those alternatives and the decisions they inform. A compressibility score alone does not establish a percentage of hours saved.

For a treasury analyst, the test is whether a reconciled, traceable lender package takes less total human time to produce. Negotiation and downside framing remain part of the role. Track how released capacity is used before assigning it economic value.

For a lineman, document preparation and dispatch support are a smaller part of a job centered on physical execution and safety. That contrast motivates the benchmark’s layer comparison; it does not measure actual time allocation.

Failure mode to monitor

The "stays human" column is only durable if organizations actually invest the freed-up time in judgment and analysis rather than simply reducing headcount. The risk: a company compresses its five-person treasury team to two people, but keeps the same workflow volume — so each remaining person rubber-stamps AI outputs at 3× speed instead of reviewing them at appropriate depth. The amplification column only works if people have the time budget to fill it. If compression translates to headcount cuts rather than role restructuring, the centaur model degrades into an automation model with a human-shaped rubber stamp at the end. This is an organizational design choice, not a technology constraint.

Across these examples, preparation, decision support, and human accountability are distinct activities. A pilot should measure them separately so faster drafting does not hide extra review or weaken the decision.

Don't ask whether AI replaces the job

Ask which part of the seat was the job.

Each tile is a task inside one role. Some flow to AI. Some stay with the human. A few recombine into new functions. The seat is being unbundled — not eliminated.

AI doesn't delete the org chart

It redraws the center.

The old pyramid thins in the middle. The decision layer stays. The field layer stays or grows. A new thin control layer appears around the machine.

The hours don't disappear

They move to higher-consequence work.

Hours leave drafting, reconciliation, and packet assembly. They arrive in exception handling, scenario exploration, and AI governance. The hours don't vanish — they migrate upward.

You can automate the training ground out of existence

Then who makes the decisions in 2035?

Junior manual work builds pattern recognition, which builds judgment, which earns signoff authority. AI slices out the junior layer. The firm now has a problem: how does it produce future decision-makers?

New work functions, not job forecasts

We estimate new work, not new jobs. Functions appear before titles do. The demand model:

These functions deliver business value, not soft benefits. An evidence architect brings compounding memory. An exception manager brings downside protection. An eval lead brings the trust that lets you actually deploy. A workflow owner brings adoption and budgeted ROI. An apprenticeship steward brings future judgment supply.

The net employment math

A mid-size operator creates 3–7 FTEs across emerging functions while restructuring 20–50 above-field positions. Net headcount likely declines. Per-person value rises. Total payroll may stay roughly flat. The IEA's 1.7-to-1 retirement-to-entrant ratio means in many cases AI-driven compression doesn't cause layoffs — it prevents the capability gap from widening. A 50-person technical team that loses 8 to retirement and replaces 5 (with AI assistance) maintains roughly the same output. HR calls this "transformation." Everyone else calls it Tuesday.

2025–2026: Pilot phase — emerging functions handled as side responsibilities. 2027–2028: Production phase — workflow architect and eval roles become distinct positions. 2029+: Regulatory pressure formalizes governance and provenance roles. The firms that benefit most treat this as workforce restructuring, not simple automation.

Confidence: scenario

These work functions are projected from benchmark structure and cross-industry labor evidence, not observed in operating companies today. The pattern — technology shifts create control work around the seams of the technology — shows up every time (database administrators before databases; DevOps before cloud; compliance officers before Sarbanes-Oxley). We are confident the functions will exist. We are not confident about timing or headcount.

Go deeper

Discuss this research with AI

Opens a new conversation with a prompt that links back to this report. The model can read the page and discuss the findings, methodology, or implications for your company.

A note on the data. 60 of 404 positions use BLS/industry employment anchors, including analogs rather than exact occupational matches. The remaining 344 use a formulaic estimate: layer base × tier multiplier (three bands: 0.75×, 1.0×, 1.25×). These are structural scaffolding for directional wage-pool analysis, not census-quality headcounts. Compensation uses six layer-level proxies ($58K–$180K), not role-specific wages. The directional finding — that roughly nine of every ten AI-exposed wage dollars sit above the field — holds across a wide range of employment assumptions. Individual role numbers should not be cited as precise. Full methodology: methodology FAQ.

How to reproduce every number in this benchmark

The core benchmark — scores, employment, compensation, and the ~90% headline — can be reconstructed from the published CSV and the three formulas below. Some interactive exhibits (role explorer, workflow economics, demand scenarios) use additional embedded data structures visible in the page source.

Formula 1 — Exposure-weighted wage bill

This exposure metric drives the wage-bill ranking, slope chart, and ~90% headline. Heatmap and treemap area use employment, not exposure-weighted wages. The core exhibits trace back to this formula applied row-by-row across the 404-position CSV.

Formula 2 — Employment estimation (344 formulaic rows)

Where:

Methodology note (v32.1): Employment estimation is decoupled from compressibility scoring. An earlier version used max(0.15, 1.3 − score/8) as a score factor, which gave low-scoring roles more people by construction. The continuous jitter 0.7 + (hash mod 600)/1000 was also replaced with three explicit tier bands to eliminate false precision. All employment figures for formulaic rows are displayed with rounding and a "~" prefix to signal estimation uncertainty.

The 60 anchor rows override this formula entirely — their employment figures come from BLS Occupational Employment and Wage Statistics or industry-derived analogs, flagged in the employment_source column. Note: two of these 60 rows are not people-roles (Board pack production, A&D diligence support) — they are workflow/artifact rows with industry-estimated volume, not BLS occupational matches.

Formula 3 — Compensation proxies

These are BLS OEWS-derived median annual wages for the layer's representative occupations, not role-specific compensation. Every row in the same layer uses the same proxy. Figures reflect wages only; employer-paid benefits are not included.

Compressibility score — what it measures

Each position's compressibility score (1–10) represents how plausibly current AI can compress the preparation stack around that role or workflow. It is built from five sub-dimensions:

The score is a calibrated composite, not a simple average. Roles with high task exposure but low adoption feasibility (e.g., safety-critical field roles) score lower than roles where both align (e.g., document assembly and reconciliation). The full scoring rationale is published per-row in the CSV's rationale column and in the scoring method note.