“They don’t even speed up engineers — they actually make them slower!”
Our reading, the sentence that was checked:
Coding LLMs do not speed up engineers and actually make them slower.
Assessment: Mixed Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, The Case Against Generative AI, paragraph 262.
- 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-64900a-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
- Causal claim
- 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
- They don’t even speed up engineers
- they actually make them slower
Our interpretation
- They refers to coding LLM tools in the preceding sentence.
- The plain reading is a general claim about engineers' work, not an explicitly limited claim about one study population.
- Speed means elapsed working time to complete comparable engineering tasks, including interaction with the tool and correction of its output.
- Assess causal evidence available for tools and work conducted before the post. No particular trial or reporting period is named.
- Opposite effects in credible controlled settings warrant a mixed verdict, while rejecting an unconditional generalization.
Counts as supported if: Randomized or comparably credible studies show that access to coding LLMs increases time to complete comparable tasks.
Counts as contradicted if: Randomized or comparably credible studies show that access to coding LLMs decreases time to complete comparable tasks.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Does the evidence support the cause?
Outcome: Mixed
Part of the claim holds and part does not, and the record says which part is which.
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
METR found a 19% slowdown for experienced developers working in familiar repositories, supporting that part of the claim. Google's randomized trial found a speedup on a complex enterprise task, contradicting the claim that coding LLMs do not speed engineers up. The evidence supports different effects in different settings, not a general slowdown.
Limitations
- The trials should not be pooled into an industry-wide estimate without accounting for their different settings.
- Google evaluated its own internal tools.
- Neither study establishes the effect of later tools.
Evidence (2 items)
Supports the claim: Becker et al., METR; arXiv,
increases completion time
Why it matters here: The randomized trial covered 16 experienced developers completing 246 tasks in familiar mature projects using February–June 2025 tools. It directly supports a slowdown in this setting.
- Where
- Abstract, results
- Publisher
- Becker et al., METR; arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Goes against the claim: Paradis et al., Google; arXiv,
significantly shortened the time
Why it matters here: A randomized trial with 96 Google engineers on a complex enterprise task estimated about a 21% reduction in completion time. This contradicts the unconditional assertion that coding AI does not speed engineers up.
- Where
- Abstract, results
- Publisher
- Paradis et al., Google; arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, The Case Against Generative AI
They don’t even speed up engineers — they actually make them slower!
- Where
- paragraph 262
- Published
- 2025-09-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-the-case-against-generative-ai-64900a-a1: Current. Outcome: Mixed. 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-64900a-a1.