“we keep running into the problem that these things are interesting, surprising, but not particularly useful for anything.”
Our reading:
Generative AI products such as ChatGPT and DALL-E are interesting but not particularly useful for anything.
Assessment: Contradicted
What happened
Was it true for the period described?
Outcome: Contradicted
The evidence goes against what they said.
Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.
Evidence up to . Checked .
Why
Zitron said that just over a year after ChatGPT and DALL-E appeared, these tools were 'not particularly useful for anything.' Controlled studies from that window show clear practical gains. A peer-reviewed randomized experiment in Science (July 2023) found ChatGPT cut time on professional writing tasks by 40% and raised quality by 18%. An NBER field study of 5,179 customer-support agents (revised November 2023) found a generative AI assistant raised issues resolved per hour by 14%, and by 34% for novice workers. These are sizeable, measured benefits on real work tasks. They contradict the plain reading that the tools had no meaningful use. His wider worries about reliability and economic value are separate claims and are not judged here.
Caveats
- The evidence covers text and chat assistants, not image generators like DALL-E.
- The task studies are narrow. They do not show broad economic returns.
- The Science abstract was read through a Stanford repository listing because science.org blocked access.
Evidence (3 items)
Goes against the claim: Science (Noy & Zhang), abstract listed by Stanford SCALE,
Our results show that ChatGPT substantially raised productivity: The average time taken decreased by 40% and output quality rose by 18%.
Why it matters: A randomized experiment found ChatGPT cut time on professional writing tasks by 40% and raised quality by 18%. That is a large, practical benefit, and it contradicts 'not particularly useful for anything'.
- Publisher
- Science (Noy & Zhang), abstract listed by Stanford SCALE
- Kind of source
- Peer-reviewed paper
- About events on
- Published
Goes against the claim: National Bureau of Economic Research (Brynjolfsson, Li, Raymond),
Access to the tool increases productivity, as measured by issues resolved per hour, by 14% on average, including a 34% improvement for novice and low-skilled workers
Why it matters: A field study of 5,179 real customer-support agents found a generative AI assistant raised issues resolved per hour by 14%. This is real-world usefulness inside the claim's window.
- Where
- abstract
- Publisher
- National Bureau of Economic Research (Brynjolfsson, Li, Raymond)
- Kind of source
- Preprint, not peer reviewed
- About events on
- Published
Context only: TechCrunch,
ChatGPT now has 100 million weekly active users, OpenAI CEO Sam Altman announced on Monday
Why it matters: Shows large repeat use by late 2023. This is context only, because use does not by itself show usefulness.
- Where
- paragraph 1
- Publisher
- TechCrunch
- Kind of source
- News report
- About events on
- Published
How we judged it
- Kind of claim
- Observation
- Deadline
- None given. We don’t add one.
Where we had to interpret
- Plain reading: in the roughly 15 months after ChatGPT launched, generative AI products such as ChatGPT and DALL-E had no meaningful practical use for real tasks.
- Well-designed studies showing clear, measured gains on real work tasks from these tools during that window would contradict the claim. A broad usefulness claim needs only one solid counterexample domain to fail.
- 'Not particularly useful' is softer than 'useless', so small or doubtful gains would not count against him. Large, statistically clear gains would.
Right if: Credible studies or data from late 2022 to February 2024 showing that ChatGPT, DALL-E or similar tools gave no meaningful practical benefit on real tasks.
Wrong if: Credible controlled studies from late 2022 to February 2024 showing large, measurable productivity or quality gains from generative AI tools on real work tasks.
We wrote this test after the outcome was known.
We found 31 claims in this post. See them all.
Full record
Source and dates
Where's Your Ed At, Subprime Intelligence, paragraph 4.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-28. The post may have been edited after 2024-02-19.
Where it was said (1)
Said once in the posts we read.
: Where's Your Ed At, Subprime Intelligence
we keep running into the problem that these things are interesting, surprising, but not particularly useful for anything.
- Where
- paragraph 4
- Published
- 2024-02-19
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Contradicted. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.