The gains are real; the review bill decides who keeps them
EM outlook · –
In 2027–2028, we expect lower effort per accepted result in repeatable digital work with clear outcomes and affordable verification. Customer-support results already beat the blanket skepticism; ’s later developer evidence points toward improvement. The opportunity is to redesign around those conditions. The rival case is that even these gains disappear when the full review bill arrives.
What this means
The prize is more room for difficult, worthwhile work. Compare accepted results with the full human effort they took, then use the capacity you recover to improve the service, tackle a neglected problem or give people time back.
Why we take this view
Two years for gains already observed in some assisted tasks to spread through workflow redesign. The forecast is about suitable workflows, with hard-to-verify work expected to lag; it is not a uniform productivity percentage.
What has to hold
Before deployment, the target work is repeatable and digital, has a clear acceptance test, and can be checked affordably. Quality standards stay fixed and delayed defects count.
Teams can select or redesign tasks instead of forcing every task through the same AI process.
The strongest rival
Even in the defined target workflows, selection and omitted review costs explain most apparent gains; lower effort does not survive a complete comparison.
Models become better at checking their own output, reducing the review burden even on complex work.
What would change our mind
Comparable later studies measure accepted outcomes, total human effort and defects after handoff, with credible treatment of selection.
The gain disappears when review and downstream repair are counted, or persists when they are included.
Compare total effort and accepted quality in workflows selected before deployment, including downstream review and repair.
Where this could lead
Checkable digital work takes less effort to finish
Base case · –
Teams lower total effort in repeatable digital workflows with clear acceptance tests and affordable verification. They define those conditions before deployment, then count the work all the way through handoff and repair.
What would make it happen
Existing assisted-work gains and improving developer experience favor wider gains over two years, with task selection and redesign driving the spread.
Conditions
The target workflow is repeatable and digital, with a clear acceptance test and affordable verification established before the comparison.
They count verification and later defects, then concentrate use where the full result improves.
Watch for
Comparable task cohorts show lower total effort at stable quality.
Benefits persist after the novelty of the rollout and include downstream repair.
The gain disappears even in repeatable digital work selected for clear outcomes and affordable checking. Hidden defects and integration work prove more costly than expected; faster drafts fail to become cheaper accepted results.
What would make it happen
This remains a live alternative throughout the adoption window, especially where correctness is expensive to establish.
Conditions
Verification and repair cost more than the team expected before rollout.
Selection of enthusiastic users or favorable tasks explains apparent gains that do not survive comparable full-work measurements.
Watch for
Blind quality review finds more defects in apparently faster work.
Time saved at generation returns as integration, correction or customer support.
Why this connects to When can work be delegated without checking every step?
EM interpretation
A result can be useful because a person caught its mistakes. Removing that review changes the job: the agent must detect failure, recover or ask for help. That is why productivity gains can arrive before dependable autonomy.
Track the human interventions behind a successful result before applying it to unsupervised work.
Why this connects to When is an AI deployment worth its total cost?
EM interpretation
Accepted work is where capability becomes economic value. A slower process can still pay if it produces something better or previously impossible; a faster draft can lose money if nobody needs it or finishing it costs too much.
The buyer’s gain and the supplier’s investment return are separate calculations.
Review and training work may be redistributed across experience levels.
Perspectives
Joel Becker, Nate Rush, Beth Barnes and David Rein
authors of the early-2025 controlled developer study · 2025-07-10 · From the author’s own work
In their randomized study of 16 experienced open-source developers completing 246 issues, AI access increased completion time by 19%. They explicitly restrict what this says about other developers, settings or later tools.
Sources and passages
Publication date of this study report.
The study used early-2025 tools in developers’ familiar repositories; it is not an estimate for all work or present-day AI.
Joel Becker, Nate Rush, Tom Cunningham, David Rein and Khalid Mahamud
authors of the follow-up methods update · 2026-02-24 · From the author’s own work
judges that developers likely benefited more from AI in early 2026 than in early 2025, while warning that participation, task selection and time measurement make the size of that improvement hard to estimate.
Sources and passages
Publication date of the authored source.
This is a later team’s assessment of a changed experiment, not the same measurement repeated cleanly.
author reporting technical-worker self-assessments · 2026-05-11 · From the author’s own work
The survey reports substantial perceived value gains, while distinguishing value from task speed and cautioning that respondents can misestimate what AI contributes.
Sources and passages
Publication date of the authored source.
Self-report in a technical-worker sample, not audited customer economics or a measured effect on national productivity.
Publisher publication date, checked on original page.
Original publisher page and exact retained source reviewed. Source retention does not establish every downstream reuse or training right.
Exact retained Stream work/snapshot and SHA256 verified against the 29 October SELECT-only source packet. That lookup found no joined public Stream or bridge reading revision; no reading identity is invented.
This acceleration makes human attention the bottleneck for agent output.
Retained extracted text, character offsets [767, 839). Context (EM paraphrase): The authors report rising internal output alongside agent use and interpret human attention as a constraint. Adoption is not randomized; the report does not establish a general causal productivity effect.
Publication date displayed by the original publisher.
Original publisher page and exact retained source reviewed. Source retention does not establish every downstream reuse or training right.
Exact retained Stream work/snapshot and SHA256 verified against the 29 October SELECT-only source packet. That lookup found no joined public Stream or bridge reading revision; no reading identity is invented.
16 experienced developers
Retained extracted text, character offsets [1527, 1552). Context (EM paraphrase): Methods: developers worked on 246 real issues in familiar open-source repositories using early-2025 tools. This population does not represent all developers or later systems.
they take 19% longer to complete issues
Retained extracted text, character offsets [2468, 2507). Context (EM paraphrase): Core Result, first paragraph; interpret with Methods and the immediately following limitations table. Reported result for the 2025 study, not a current universal estimate.
Publication date displayed by the original publisher.
Original publisher page and exact retained source reviewed. Source retention does not establish every downstream reuse or training right.
Exact retained Stream work/snapshot and SHA256 verified against the 29 October SELECT-only source packet. That lookup found no joined public Stream or bridge reading revision; no reading identity is invented.
the data from our new experiment gives us an unreliable signal of the current productivity effect of AI tools
Retained extracted text, character offsets [431, 540). Context (EM paraphrase): Opening discussion, paragraph beginning Unfortunately. Subsequent paragraphs describe participation, task selection and time measurement.
likely biases downwards
Retained extracted text, character offsets [715, 738). Context (EM paraphrase): METR says nonparticipation by developers unwilling to work without AI likely understates speedup; the authors believe early-2026 tools speed developers more than early-2025 tools. Lower study pay, task selection and concurrent-agent time accounting limit the size estimate.
Publication date displayed by the original publisher.
Original publisher page and exact retained source reviewed. Source retention does not establish every downstream reuse or training right.
Exact retained Stream work/snapshot and SHA256 verified against the 29 October SELECT-only source packet. That lookup found no joined public Stream or bridge reading revision; no reading identity is invented.
Participants self-reported a median 1.4–2x change in the value in their work due to AI tools.
Retained extracted text, character offsets [1308, 1401). Context (EM paraphrase): METR surveyed 349 technical workers in February–April 2026. The range reflects self-reported value under different question formulations, not a confidence interval or observed causal return. METR gives reasons to doubt the magnitude and distinguishes value from speed.
Version 2 submission date displayed by ; first submitted 23 April 2023. The versioned abstract is the cited object.
Versioned author abstract checked directly. The HTML rendering displayed a conflicting 2026 manuscript date, so this edition cites only the clearly versioned abstract and does not assign that later date to the study.
Original public source read on 30 October 2026. An exact retained Stream identity has not been resolved for this edition.
by 15\% on average
Version 2 abstract, second sentence. EM context: staggered introduction among 5,172 customer-support agents; the outcome is issues resolved per hour, not company profit.
Less experienced and lower-skilled workers improve both the speed and quality of their output
Version 2 abstract, third sentence. EM context: benefits differ by experience; the most experienced workers show small speed gains and small quality declines.
Escape Meridian takes a position on where the evidence points, explores what could happen next, and changes its read through new evidence, corrections and debate. The outlooks are our judgments; the sources retain their own voices.
About this edition
EM outlooks authored 30 October 2026 from the retained 29 October source review, supplemented by five original public sources checked on 30 October. Employment evidence centers on US workers and selected workplaces; capability evidence centers on software tasks. Outlook dates express EM judgments about the stated outcomes, not dates forecast by the cited authors. Source publication, access and evidence-review dates remain distinct.
