What does it check?
One job, different surfaces.
The tools look at different kinds of evidence, but they are not separate missions. Sale Checker asks whether the price claim holds up. AI Trust asks what commercial and privacy rules sit behind an assistant. Political Money follows documented federal spending. Media Check looks for disclosed provenance. Last Good Model asks what changed after a product people liked. Human Web applies the same idea while you browse.
The common question is simple: what is the important part somebody might not put in the headline?
Why the name?
Buying and owning ordinary stuff should not require a research project. LessBother exists to do more of that annoying checking for you.
Who pays for a good result?
Nobody. A manufacturer, retailer or AI company cannot pay to improve a LessBother conclusion. If the project ever accepts donations, they do not buy placement or editorial control.
How does LessBother check things?
There is no secret scoring sauce. The rules are deliberately boring:
- Use the closest source to the fact. Manuals, support pages, terms, price pages, regulatory records and service documentation beat somebody repeating a claim.
- Do not make the claim bigger than the evidence. If one model has a restriction, LessBother does not quietly turn that into “the whole brand is bad.”
- Old evidence can expire. Prices, subscriptions, firmware policies, privacy rules and features change, so current claims need rechecking.
- Keep uncertainty visible. “We do not know yet” is a valid result.
Yes, AI helps build LessBother.
That is not hidden. AI-assisted coding and research organization make this project possible. But AI output is not treated as source evidence. A claim still needs something independently checkable behind it.