Details: An article is published announcing multiplex editing of ≥5 genomic loci (one loci affecting multiple genes - i.e. via frame shift - doesn't count, has to be simultaneous edits at different places) in at least 30% of the cells in at least one tissue in a living human organism - not counting embryo/pre-implantation, fetuses count if they survive.

The article is considered credible if there is a prediction market (on manifold or another platform with equal or stronger empirical results for market calibration) with >200 participants running at >80% "YES" on "Will this method be reproducible/replicable?" (or a question to the same effect. As in: "Will an independent biopsy reveal this person's genes actually got edited?")

Or, it is considered credible if it is published by a lab with a >80% track record on successful replications, or it is peer reviewed and published in a journal with a strong record for replications in that field (within 5% of the successful replication rate of Nature or Cell for their peer reviewed gene editing publications.)

I realize this second criteria may be a weaker test of credibility than the first. Yet, that is the civilization we live in. Thus, if someone just makes something up and peer reviewers pass it on and Nature publishes it, this market will pay out, and the winners will have a small celebration to go along with their bitter disappointment.