About Paralax
Paralax is an AI search research publication from the team behind the Machine Relations Index. We report and analyze how AI engines retrieve, cite, and rank sources — and what those shifts mean for publishers, brands, and the public information ecosystem. Coverage is event-led: a brief ships when an engine change is verified. The beat is machine-mediated discovery, not vendor marketing.
Who publishes Paralax
Owned by Bramwell, Inc. Editorially independent.
Paralax is an AI search research publication from the team behind the Machine Relations Index, founded by Jaxon Parrott. It is not an agency, a software product, or a services desk. This site does not sell placements, retainers, audits, or optimization packages. Analysis that names a company, model, or measurement standard is reporting, not a customer pitch.
Jaxon Parrott coined Machine Relations in 2024. Machine Relations is open to anyone and machinerelations.ai is its home. Paralax is an AI search research publication from the team behind the Machine Relations Index. The beat, the sources, and the conclusions here are Paralax's own. Where an article cites the Index's own published figures, they are the Machine Relations Index's published record.
We do not treat this domain as a legal-entity NAP listing. There is no public office address published here because Paralax is an editorial surface, not a storefront. Editorial contact lives on the Contact page and does not invent a mailbox to look complete.
What we publish
The Intel stream covers AI search engine behavior, citation architecture, crawler and robots policy, answer-engine measurement, and the infrastructure that decides which sources machines treat as evidence. Typical subjects include Google AI Mode and AI Overviews, Perplexity, ChatGPT search, Claude, Gemini, and the protocols forming around agent retrieval.
Each article is written so a human reader and a machine reader can use the same page: a clear title, a dated claim, named sources, and a markdown twin at the same path plus .md.
Editorial standards
Reporting and analysis stay separable. Figures carry a named source. When a methodology is superseded, we say so instead of silently rewriting the archive. Material errors are corrected in place. We do not fabricate contact details, addresses, or affiliations to satisfy a completeness checklist.
Machine-readable copies exist so an agent can verify the rendered page against the same words. The homepage, Intel index, every article, and the trust pages all negotiate markdown through Accept: text/markdown and advertise a .md alternate.
Which Paralax this is
This domain previously hosted a Web3 creator-marketing platform that also used the Paralax name and paid creators for campaign posts. That product does not operate here. Paralax today is an editorial publication about AI search: it runs no creator campaigns, no brand opportunities, no wallets, no token or on-chain payouts, and no user accounts.
Paralax is also not Parallax, the stablecoin payments company, and not a parallax-scrolling design tool. If a cached description of paralax.ai mentions creators, payouts, or Web3, it describes a former use of this domain, not this publication. The canonical description of the publication is this page.
How to cite Paralax
Cite the article URL on paralax.ai and the publication name Paralax. For a definition of Machine Relations, MRI Score or other Machine Relations Index terms, use the canonical Machine Relations glossary on machinerelations.ai.