Updated 6 Nov

AI exposes data problems instead of solving them. Metrics nobody agreed on make it give confident wrong answers

Metrics & ExperimentationAI in Product

Reader question: "Our numbers never agree. Is that an AI problem?" Small and mattering: the reader arrives asking why the assistant is wrong and the answer is that nothing it was handed was ever agreed. The three tools returning three revenue numbers is the sentence that lands it. Not a news cycle: vendor posts and practitioner essays with no shared trigger.

sources

last 30d

First seen 10 Feb 2026

Sources saying it

17mentions, since 14 Aug

Month Quarter All time

What kind of source is saying it

Amplitude, Thoughtworks Insights and 1 more writing

Who said it

Artificial intelligence – MIT Technology Reviewtechnologyreview.com · 2 November 2026
Raising the quality of insights in agentic analytics
…When analytics agents deliver insights of substandard quality, the main reason is a lack of business context, as cited by 55% of surveyed executives. Fragmented or inaccessible data is the other major factor, according to 45%. …
Thoughtworks Insights

AI exposes data problems instead of solving them. Metrics nobody agreed on make it give confident wrong answers

Sources saying it

What kind of source is saying it

Who said it

More trends

Work got cheaper to produce, but review took longer, so the organization didn't get faster

Most tasks don't need the strongest model. Match the model to the task to save money