“LLMs write code wicked fast, but some coders are starting to report burnout, and only relatively modest gains relative to that burnout, as a new story from Connie Loizos at Tech Crunch illustrates”
Our reading, the sentence that was checked:
LLMs write code very fast, but some coders report burnout and only modest gains compared with that burnout.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, About that Matt Shumer post that has nearly 50 million views, paragraph 8.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-about-that-matt-shumer-post-that-89d9ae-a2
We recorded 4 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 2026-02-11. 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)
Conditions
- some coders (qualifier, stated by the speaker)
From the speaker
- LLMs write code wicked fast
- some coders are starting to report burnout
- only relatively modest gains relative to that burnout
- as a new story from Connie Loizos at Tech Crunch illustrates
Our interpretation
- The claim is about reports, not a population rate. It holds if credible recent reporting, especially the cited TechCrunch story, documents some coders or engineers describing burnout alongside modest gains.
- 'Wicked fast' is uncontroversial framing and is treated as secondary.
Counts as supported if: The TechCrunch story by Connie Loizos, or similar reporting from the period, documents engineers or coders using AI who report burnout and gains that are modest compared with the extra strain.
Counts as contradicted if: The TechCrunch story does not exist or does not describe burnout among AI-using coders, and no other reporting from the period shows it.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
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
The TechCrunch story Marcus cites exists. It is by Connie Loizos and was published on 9 February 2026 (e1). It describes in-progress UC Berkeley research at a 200-person tech company. In that research, an engineer said that with AI 'you don't work less. You just work the same amount or even more' (e1). The story also quotes a Hacker News commenter who said stress had tripled while productivity rose 'maybe 10%' (e2). The researchers link the extra work to 'fatigue, burnout' (e3). This meets the claim that some coders are starting to report burnout with modest gains. One caveat: the researchers say AI's boost to what workers can do is real (e3). The 'modest gains' part therefore rests mainly on the forum comment.
Limitations
- The evidence is a small set of reports: one interviewed engineer and one anonymous forum commenter. It does not show how common burnout is among coders.
- The Berkeley study covers tech workers at one company, not only programmers, and the story does not quantify their gains.
- The METR item on coding speed (e4) is dated after the post and serves only as context for 'wicked fast'.
Evidence (4 items)
Supports the claim: TechCrunch,
As one engineer told them, “You had thought that maybe, oh, because you could be more productive with AI, then you save some time, you can work less. But then really, you don’t work less. You just work the same amount or even more.”
Why it matters here: The Loizos story exists and reports an engineer at an AI-embracing tech company describing more work, not less.
- Where
- paragraph 5
- Publisher
- TechCrunch
- Kind of source
- News report
- Describes events of
- Public from
Supports the claim: TechCrunch,
expectations have tripled, stress has tripled and actual productivity has only gone up by maybe 10%.
Why it matters here: The story quotes a tech worker reporting stress that has tripled against a gain of about 10%. That matches 'modest gains relative to that burnout'.
- Where
- paragraph 6
- Publisher
- TechCrunch
- Kind of source
- News report
- Describes events of
- Public from
Supports the claim: TechCrunch,
It confirms it, then shows where all that augmentation actually leads, which is “fatigue, burnout, and a growing sense that work is harder to step away from, especially as organizational expectations for speed and responsiveness rise,” according to the researchers.
Why it matters here: The underlying UC Berkeley/HBR research reports burnout among AI adopters. It also says the augmentation is real, which tempers the 'modest gains' part.
- Where
- paragraph 9
- Publisher
- TechCrunch
- Kind of source
- News report
- Describes events of
- Public from
Context only: METR,
This is also partly because many AI agents code much faster than human software engineers.
Why it matters here: This backs the uncontroversial 'wicked fast' framing. The page version read is dated after the post, so it is context only.
- Publisher
- METR
- Kind of source
- Benchmark result
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, About that Matt Shumer post that has nearly 50 million views
LLMs write code wicked fast, but some coders are starting to report burnout, and only relatively modest gains relative to that burnout, as a new story from Connie Loizos at Tech Crunch illustrates
- Where
- paragraph 8
- Published
- 2026-02-11
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-about-that-matt-shumer-post-that-89d9ae-a2: Current. Outcome: Supported. 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 marcus-about-that-matt-shumer-post-that-89d9ae-a2.