“Generative AI can’t do the job of a software engineer, and it fails while also costing abominable amounts of money.”
Our reading, the sentence that was checked:
Generative AI cannot do a software engineer's job, and it fails while costing huge amounts.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, The Case Against Generative AI, paragraph 261.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-the-case-against-generative-ai-7c2756-a1
We recorded 39 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2025-09-29. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
From the speaker
- Generative AI can’t do the job of a software engineer
- it fails
- costing abominable amounts of money
Our interpretation
- The surrounding passage distinguishes generating code from an engineer's broader responsibilities. Read the claim as inability to perform a sustained engineering role reliably with ordinary workplace oversight.
- Do not require flawless performance: human engineers also make mistakes.
- No reporting period is specified. Use September 1–29, 2025, the latest calendar month observable before publication.
- The expense description supplies no company, amount, cost category or comparison. It is a value judgment that cannot receive a separate numerical verdict.
- Metric scope is null because this is not a claim about an identified company's finances.
Counts as supported if: Representative end-to-end evaluations show contemporary systems unable to carry a sustained engineering workload at acceptable quality with ordinary oversight. The expense component additionally needs a stated cost basis and meaningful comparator.
Counts as contradicted if: A verified sustained deployment or representative trial demonstrates that a contemporary system performs an engineer's substantive responsibilities at acceptable quality and supervision levels.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
SWE-Bench Pro supports the claim of substantial failure on complex tasks: its leading models resolved roughly a quarter of the public problems. It also records successful engineering work. Neither result establishes whether AI could sustain an entire engineering role under ordinary oversight. The expense claim has no defined measure or comparator, so the compound assertion cannot receive a supported or contradicted verdict.
Limitations
- A benchmark of individual problems does not reproduce an engineer's full responsibilities.
- Success depends on the tools and instructions provided to the model.
- The cost language cannot be converted fairly into an invented financial threshold.
Evidence (1 item)
Supports the claim: Deng et al., arXiv,
23.3% and 22.7%
Why it matters here: Low resolution rates support the narrower claim that contemporary systems fail frequently on complex engineering tasks. Successful resolutions also demonstrate partial capability; this evaluation does not establish inability to perform every engineering role or excessive total costs.
- Publisher
- Deng et al., arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
A sustained, independently assessed engineering deployment with documented human supervision, quality and total costs; a defined comparator is also needed to assess the expense component.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, The Case Against Generative AI
Generative AI can’t do the job of a software engineer, and it fails while also costing abominable amounts of money.
- Where
- paragraph 261
- Published
- 2025-09-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-the-case-against-generative-ai-7c2756-a1: Current. Outcome: Unresolved. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.
About assessment wyea-the-case-against-generative-ai-7c2756-a1.