# Arvind Narayanan and Sayash Kapoor argue AI has not replaced software engineers [Canonical reading](https://escapemeridian.com/readings/arvind-narayanan-and-sayash-kapoor-argue-ai-has-not-replaced-31600a8c) Arvind Narayanan and Sayash Kapoor published this essay on June 11, 2026, arguing that AI has not caused mass layoffs of software engineers and that AI capabilities alone are unlikely to automate the whole profession. They frame software development as a “decide-execute-deliver sandwich”: AI compresses the execution layer, while deciding what to build and taking responsibility for what is delivered remain human bottlenecks. The authors point to evidence that AI agents can increase code production without a comparable increase in software releases. They also argue that job cuts attributed to AI may have other causes, and that employment trends and layoff announcements are different measures of AI’s effects. Their outlook is cautiously optimistic: cheaper software could increase demand for software and, in turn, software engineers. They distinguish this aggregate outlook from the possibility of substantial disruption to individual workers’ careers. The authors’ case depends in part on their model of how software work is organized, and they acknowledge gaps in evidence, including limited data on poorly supervised AI use in high-stakes work. How will future evidence distinguish AI-related changes in hiring from other causes of employment changes? Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims. ## Original source Arvind Narayanan and Sayash Kapoor · Article Original source date: Jun 11, 2026 (2026-06-11T02:29:51.000Z). Date basis: Date supplied by the publisher’s feed. Added to Escape Meridian: 2026-10-30T07:48:56.471Z. First observed: 2026-10-23T20:31:49.340Z. Retrieved: 2026-10-23T20:33:36.887Z. - [Why AI hasn’t replaced software engineers, and won’t](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Limits Text as retrieved on October 23, 2026; originally published June 11, 2026. The authors note that the Federal Reserve study they discuss does not capture self-employment, limiting what it can say about employment trends. The authors describe the SWE-chat dataset as self-selected and say it cannot support strong conclusions on its own. The authors identify limited data on poorly supervised AI use in high-stakes tasks as an important gap. The essay's forecasts about future demand remain the authors’ projections, not established outcomes. ## Open question What future evidence can distinguish AI-related changes in software engineering hiring from changes caused by other factors? ## Source-stated Narayanan and Kapoor argue that evidence from software engineering challenges claims that AI capabilities will cause mass layoffs. > there is enough evidence to reject the narrative that once AI capabilities reach a certain threshold, it will cause mass layoffs. - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors describe software development as a “decide-execute-deliver sandwich,” with AI compressing execution while decision-making and accountability remain bottlenecks. > AI has compressed the middle of the sandwich, but has left the two ends largely unchanged. - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors cite research in which increased code production was accompanied by a smaller increase in releases. > AI agents led to an eight-fold increase in the number of lines of code written, consistent with the idea that AI almost completely compresses the Execute layer of the sandwich. But this led to only 30% more releases - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors argue that cheaper software could increase demand for software engineers, while noting that individual workers may still face substantial changes. > while we think there will be a lot more software in the future, and likely more software engineers - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors identify a lack of data on whether poorly supervised AI use in high-stakes tasks is increasing. > remains one of the most critical data gaps we have today. - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors published the essay on June 11, 2026. > Jun 11, 2026 - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source and contributor context Arvind Narayanan and Sayash Kapoor · Publisher The publisher is identified here. No individual author has been inferred. ## Source media Media retrieval status: unchecked. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Editorial provenance Version 1. Publication ID: 5a717d18-d44f-c924-6cf8-64016cd1901c. Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5. Source check: passed by anthropic/claude-opus-5.5 on 2026-10-30T07:48:54.564Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: bridge\_admitted\_by\_model. Observation: 4f70929ca597ccdafa48f21197bcb19f5d49abf063166eeaacc2f0ff5c253a03. Source text hash: a35f4653858a02fa48a9001148e323c4df710e869f8163a29e535e0cbc8f88ff. ## Revision and source-check history ### Version 1 — 2026-10-30T07:48:56.471Z Arvind Narayanan and Sayash Kapoor published this essay on June 11, 2026, arguing that AI has not caused mass layoffs of software engineers and that AI capabilities alone are unlikely to automate the whole profession. They frame software development as a “decide-execute-deliver sandwich”: AI compresses the execution layer, while deciding what to build and taking responsibility for what is delivered remain human bottlenecks. The authors point to evidence that AI agents can increase code production without a comparable increase in software releases. They also argue that job cuts attributed to AI may have other causes, and that employment trends and layoff announcements are different measures of AI’s effects. Their outlook is cautiously optimistic: cheaper software could increase demand for software and, in turn, software engineers. They distinguish this aggregate outlook from the possibility of substantial disruption to individual workers’ careers. The authors’ case depends in part on their model of how software work is organized, and they acknowledge gaps in evidence, including limited data on poorly supervised AI use in high-stakes work. How will future evidence distinguish AI-related changes in hiring from other causes of employment changes? ## Original source Arvind Narayanan and Sayash Kapoor · Article Original source date: Jun 11, 2026 (2026-06-11T02:29:51.000Z). Date basis: Date supplied by the publisher’s feed. Added to Escape Meridian: 2026-10-30T07:48:56.471Z. First observed: 2026-10-23T20:31:49.340Z. Retrieved: 2026-10-23T20:33:36.887Z. - [Why AI hasn’t replaced software engineers, and won’t](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Limits Text as retrieved on October 23, 2026; originally published June 11, 2026. The authors note that the Federal Reserve study they discuss does not capture self-employment, limiting what it can say about employment trends. The authors describe the SWE-chat dataset as self-selected and say it cannot support strong conclusions on its own. The authors identify limited data on poorly supervised AI use in high-stakes tasks as an important gap. The essay's forecasts about future demand remain the authors’ projections, not established outcomes. ## Open question What future evidence can distinguish AI-related changes in software engineering hiring from changes caused by other factors? ## Source-stated Narayanan and Kapoor argue that evidence from software engineering challenges claims that AI capabilities will cause mass layoffs. > there is enough evidence to reject the narrative that once AI capabilities reach a certain threshold, it will cause mass layoffs. - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors describe software development as a “decide-execute-deliver sandwich,” with AI compressing execution while decision-making and accountability remain bottlenecks. > AI has compressed the middle of the sandwich, but has left the two ends largely unchanged. - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors cite research in which increased code production was accompanied by a smaller increase in releases. > AI agents led to an eight-fold increase in the number of lines of code written, consistent with the idea that AI almost completely compresses the Execute layer of the sandwich. But this led to only 30% more releases - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors argue that cheaper software could increase demand for software engineers, while noting that individual workers may still face substantial changes. > while we think there will be a lot more software in the future, and likely more software engineers - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors identify a lack of data on whether poorly supervised AI use in high-stakes tasks is increasing. > remains one of the most critical data gaps we have today. - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source-stated The authors published the essay on June 11, 2026. > Jun 11, 2026 - [Supporting source](https://www.normaltech.ai/p/why-ai-hasnt-replaced-software-engineers) ## Source and contributor context Arvind Narayanan and Sayash Kapoor · Publisher The publisher is identified here. No individual author has been inferred. ## Source media Media retrieval status: unchecked. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Editorial provenance Version 1. Publication ID: 5a717d18-d44f-c924-6cf8-64016cd1901c. Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5. Source check: passed by anthropic/claude-opus-5.5 on 2026-10-30T07:48:54.564Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: bridge\_admitted\_by\_model. Observation: 4f70929ca597ccdafa48f21197bcb19f5d49abf063166eeaacc2f0ff5c253a03. Source text hash: a35f4653858a02fa48a9001148e323c4df710e869f8163a29e535e0cbc8f88ff. ### passed — 2026-10-30T07:48:54.564Z — anthropic/claude-opus-5.5 Publication: 5a717d18-d44f-c924-6cf8-64016cd1901c attribution — pass: The essay is correctly attributed to Arvind Narayanan and Sayash Kapoor, who appear in the byline. The arguments are consistently framed as the authors' own, and claims from cited studies are presented as research the authors cite. > Arvind Narayanan and Sayash Kapoor Jun 11, 2026 summary — pass: The summary accurately conveys the central thesis, the sandwich framework explained in plain terms, the AI-washing argument, the code-versus-releases evidence, the demand outlook and the distinction from individual career disruption. A general reader can follow the actors and terms. The question is a genuine open issue that follows from the evidence. > AI compresses the “execute” layer — the middle of the sandwich — but the other two layers resist automation in a way that will not be overcome by capability improvements alone. date — pass: evidence.originalPublished is 2026-06-11T02:29:51Z, and its UTC day matches the source dateline of Jun 11, 2026. The evidence is a publisher page, so no archive disclosure is needed. modifiedAt equals the publication time. Forecasts are framed as the authors' projections, with no hindsight framing. > Jun 11, 2026 support — pass: Each claim is backed by verbatim source text. The note that individual workers may still face changes is supported by the authors' statement about the next essay. The demand outlook is attributed to the authors and labeled a projection. The data-gap claim and the stated limitations match the source. > The fact that aggregate labor demand in software is likely to remain strong doesn’t mean that most individual workers won’t be affected. media — pass: No media was selected, and no verified media candidates were supplied, so selecting none is appropriate. ## Recorded references Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter. No public references recorded here yet. [Reference data (JSON)](/readings/31600a8ceb52a83f8aa5dd75426ffd6ba753517638069817b442d5986c391a06/references.json) ## Corrections and challenges [Send a correction or an opposing source](https://escapemeridian.com/corrections/) [JSON record](https://escapemeridian.com/readings/arvind-narayanan-and-sayash-kapoor-argue-ai-has-not-replaced-31600a8c.json)