ViewCred is Not a Follower Count
Most crypto news platforms have the same problem: reach is determined by who paid to boost it, not who's actually right.
Anonymous shills move markets on follower counts. Sponsored content masquerades as journalism. And nobody tracks whether the people making calls are actually accurate over time.
We're building ViewFT, an AI-powered financial markets intelligence platform, and from day one we knew aggregating news wasn't enough. So we built ViewCred into the core of the product.
What ViewCred actually is
It's a credibility score, inspired by the Greek concept of Ethos: trust built through demonstrated character. Think of it as a credit score for information accuracy.
You can't buy it
You can't transfer it
You earn it through consistent, accurate contributions over time
You lose it fast if your content gets confirmed as misleading
The score runs from 0 to 1,000 across five tiers: New, Contributor, Trusted Voice, Authority, and Oracle. Oracle status takes a minimum of 24 months of sustained accurate contribution. No shortcut.
How it connects to the platform
ViewFT is the game: creators publishing, viewers engaging, content competing on quality. ViewCred is the rulebook. AI is the infrastructure running underneath everything, indexing content on publish, distributing based on credibility signals, and pre-screening flags before they reach human reviewers.
Every piece of content that gets disputed goes through a three-layer verification process: AI automated fact-checking (30% weight), an elected editorial board (30% weight), and a community vote weighted by each voter's own ViewCred score (40% weight). Voters have their own credibility on the line, which prevents coordinated manipulation.
If content is confirmed misleading, the creator gets slashed. Penalties scale by tier. Oracle-level violations carry 10x the penalty of a New account, because higher trust means higher responsibility.
Why this was hard to build
The temptation with credibility systems is to make them gameable: let money accelerate the score, let follower counts influence it. We deliberately blocked every one of those shortcuts. What actually moves ViewCred: prediction accuracy (25%), content quality (20%), community trust (20%), platform activity (20%), verification history (15%).
None of those can be faked at scale.
Where we are
Currently in beta. Would love feedback from anyone who's thought about trust infrastructure, information markets, or credibility systems. Happy to go deep on the design decisions.
