Two of the predicted trends by Stephen H. Friend on the near future of science have a true ‘p2p’ flavor:

(the trends in bold will disappear and be replaced by their more distributed counterparts)

“4. Indications for drugs will be determined by clinical trials performed by the biotech/pharma company developing the therapy. Most drugs today get approved by the FDA even if they only work in a small fraction of patients. This practice is going to end, because once a drug is approved with regards to its safety profile it may well be that the definition of who should get the drug is modified continuously by large trials organized by payers and patients. In real time, these groups will evolve the criteria of who should get which drug by participating in ongoing trials even after the drug has “been approved.” This is something to look forward to, as it will take much of the trial-and-error nature out of prescription medicine. It will be a world of real evidence based therapies

5. Hunter-gatherer approaches where large groups collect massive clinical and genomic information and expect that they as the data generator will be the data analyzer. Funding of large cohort studies like the famous “Framingham Heart Study” that has been following the health of patients in Framingham, MA for more than 60 years have been extremely valuable. The old methods used for these studies assumed that the analysis would be done by the small group of primary collectors of the samples and data. This model, too, will be fading as distributed groups of scientists evolve the knowledge faster and more efficiently than those who generated the data. Remember, this is already how physicists work today. Also remember Jim Gray of Microsoft Research, and his ideas on “The Fourth Paradigm,” which says that scientific theory, experimentation, and large-scale computational simulation will begin to interact and reinforce each other in ways that will speed up scientific progress.”