“The reason I'm suddenly bringing up superintelligences — or AGI (artificial general intelligence) — is because throughout every defense of generative AI is a deliberate attempt to get around the problem that generative AI doesn't really automate many tasks.”
Our reading, the sentence that was checked:
Generative AI does not really automate many tasks.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, Pop Culture, paragraph 22.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-pop-culture-9b0953-a2
We recorded 27 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 2024-07-08. 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 doesn't really automate many tasks
Our interpretation
- Assess the task-automation proposition, not the surrounding allegation about advocates' intentions.
- The plain reading is that generative AI completes only a limited range of work tasks without people performing or repairing substantial parts of them.
- Task automation is distinct from assistance, adoption, and replacement of whole jobs. High volume on one task does not establish a broad range of automated tasks.
- Use April–June 2024, the latest completed quarter before publication, for this ongoing-state observation.
Counts as supported if: Representative deployment evidence for the target period shows that reliable, end-to-end automation remained confined to a limited range of work tasks.
Counts as contradicted if: Representative evidence for the target period establishes reliable, end-to-end automation across a broad range of work tasks.
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
On February 27, 2024, Klarna reported that its assistant handled two-thirds of customer-service chats in its first month, before the April–June 2024 assessment period [e1]. A retrospective preprint revised July 23, 2026 reports roughly 60% success for Q2 2024 models on evaluated text tasks taking humans about 1.5 hours [e2]. These results do not establish whether dependable workplace automation covered a limited or broad range of tasks during April–June 2024 [e1, e2].
Limitations
- Klarna's company disclosure does not independently establish successful end-to-end completion, human intervention rates or representative task coverage [e1].
- The retrospective study evaluates text-task capability rather than deployed workplace automation [e2].
- The claim supplies no denominator or threshold for 'many' tasks.
Evidence (2 items)
Context only: Klarna, distributed by PR Newswire,
700 full-time agents
Why it matters here: Klarna reported substantial customer-service automation, which challenges an impression of negligible practical capability. This earlier, company-reported workload equivalent neither establishes the breadth of automated tasks nor counts eliminated jobs.
- Publisher
- Klarna, distributed by PR Newswire
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: arXiv; Mertens and colleagues,
across many tasks.
Why it matters here: The retrospective study reports roughly 60% success for 2024-Q2 models on text tasks taking humans about 1.5 hours. This is material evidence of broad capability, but selected text-task evaluations do not establish reliable automation in actual workplaces.
- Where
- Abstract
- Publisher
- arXiv; Mertens and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
Representative deployment evidence for April–June 2024 measuring task variety, successful end-to-end completion, human intervention and deployment coverage, assessed against an explicit denominator and threshold for limited versus broad automation.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, Pop Culture
The reason I'm suddenly bringing up superintelligences — or AGI (artificial general intelligence) — is because throughout every defense of generative AI is a deliberate attempt to get around the problem that generative AI doesn't really automate many tasks.
- Where
- paragraph 22
- Published
- 2024-07-08
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-pop-culture-9b0953-a2: 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.
HeldBy this site
The final-text check failed: rationale_supported_by_evidence; overclaims: The evidence establishes meaningful automation and broad text-task capability, but it does not measure how many kinds of work were reliably automated in April–June 2024..
About assessment wyea-pop-culture-9b0953-a1, which was not published.