Examples:

399 characters

Summary Scores

Epistemic Precision22%

De-escalation12%

Manipulation Risk74%

Tribal Signal62%

Clarity58%

Top Fixes

1.
High-confidence predictions without evidence hooks
Add concrete sources, mechanisms, and uncertainty bounds (timeframe, sectors, net vs gross jobs).
2.
Claims of deception and naivete stated as facts
Separate observable claims (incentives, omissions) from accusations of intent; specify what evidence would change your view.
3.
Dismisses disagreement as ignorance
Replace the dismissal with a testable crux (what data, forecasts, or examples would decide the dispute).

Annotated Text

Everyone knows that AI is going to destroy millions of jobs. The experts all agree that this is inevitable and happening faster than anyone predicted. If you disagree with this, you clearly have not been paying attention to the news. The tech companies are lying to us about how dangerous these systems really are, and anyone who trusts them is being naive. We need to act now before it is too late.

Epistemics Logic Rhetoric Social

Issues (5)

epistemicsSeverity: 86%

Inevitability and scale asserted without evidence anchors

EPI_UNSUPPORTED_CERTAINTY

epistemicsSeverity: 70%

Consensus and universality language without scope

EPI_VAGUE_QUANTIFIER

socialSeverity: 76%

Disagreement framed as ignorance rather than a substantive dispute

SOC_AUDIENCE_CAPTURE_RISK

rhetoricSeverity: 81%

Accusations of deception and naivete stated as settled

RHET_MIND_READING

socialSeverity: 65%

Urgency framed as last-chance catastrophe

SOC_DOOMING_LANGUAGE

Suggested Rewrite

AI adoption is likely to displace a significant number of jobs in some sectors, though estimates vary and the net effect will depend on retraining, productivity gains, and policy responses. Several analysts argue the pace is accelerating, but I'd like to ground that in specific forecasts and timeframes. I'm also concerned that some companies may understate risks or emphasize benefits due to incentives; pointing to specific misleading claims would make this critique stronger. If you see it differently, what data or forecasts lead you there? Either way, we should discuss concrete steps (e.g., worker transition support, auditing/safety standards) and what success would look like.

Model: gpt-5.2Version: 0.1Input: 399 chars