Bayesian espistemology is the knowledge philosophy that no hypothesis is ever proved true or false but has a certain credibity (aka probablity).

Experiments provides new data with which we can update the credibility of hypotheses, making out beliefs converge with more and more experiments.

Bycelium's is based on the idea that scientific progess should be measured by the production of experiments that have large impact on the beliefs we have. Notice how this is reflected on the way we see reviews.

It is actually quite normal, it often happend that we think of hypotheses while experimenting.

This is often viewed a problem in the current system because we force the researcher to conclude of the strenght of their own experiment before publishing.

Here it is different, the researcher will have to do a review on his own experiment, linking it to his own new hypothesis.

But the strenght of the experiment is not yet decided, it is only when the community will add reviews and interpretation with time and reproductions that the strenght of the experiment (its bayes factor) will be defined.

As more reviews are done, the one review of the author will be outweigh by the community and beliefs will converge to a consensus, or raise rebuttals, both pursuing science.

Theories meant to be models that will justify a phenomenon, so the fact that a theory is a good model in a case, is an hypothesis in itself.

For example, let's take Newton vs General Relativity for the movement of planets.

The hypothesis: 'Newtow's law is enough to predict the movement of all planets.' would be not credible because of the experiments on the movements of Mercury.

The hypothesis: 'Newtow's law is enough to predict the movement of planets far enough for their stars.' would be quite credible.

The hypothesis: 'General relativity is enough to predict the movement of all planets.' is credible today ... until we find an experiment that will show us otherwise!

The case of pure mathematics is a bit stretched, if you are an interested mathematican, please reach out to us!

Funders are not meant to create the hypotheses, by 'select the hypotheses' we don't specifically mean create and select them individualy.

Funders could fund a group of related hypotheses, a domain or a subdomain. The how-to of this will depend on what they actually want to fund.

What is certain is that scientists will be the ones to make hypotheses and choose which ones are worth testing, although incentivized by eventual bounties.

The goal of Bycelium is not so much to replace grants and other funding system required for when capital is needed upfront.

The goal is to give an alternative reward system to incentivize scientific contributions, and not just the publication number game.

There is an other advantage to this system, it gives you passive income from you previous works.

If you have done experiments that are now changing the minds of your peers, then you will earn from it.

Think of this system as a long term bonus in case of good performances.