With increased regulation, the overwhelming chance of failure and lawsuits looming for each new treatment, it's little surprise that the private sector is abandoning medical research - or at least wanting to share the costs.

One of the four founding tenets of Science 2.0 since its inception, along with publication, communication and public participation, has been collaboration. In medicine, for example, the Science 2.0 vision for collaboration would drug companies and government regulators from an early stage.

But the culture is not really ready yet - a team of rivals may work in politics but rarely in business - and that means the technology can't be created. Any number of collaboration projects have started off with fanfare and reality has quickly set in. The numbers are scary. The European Union has already spent billions on its Innovative Medicines Initiative (IMI) and though it has created more than 40 consortia, they are now saying they need even more money and collaborations to do anything meaningful. If the goal of Science 2.0 is to create more committees, that's $7 billion well-spent, but the U.S. President’s Council of Advisors on Science and Technology (PCAST) wants a U.S. counterpart to IMI, even though it has accomplished nothing.

Writing in Science Translational Medicine, Magdalini Papadaki, a research associate at MIT, and Gigi Hirsch, executive director of the MIT Center for Biomedical Innovation, are at least calling for an investigative approach to figure out what works before spending too much money. Currently, these efforts are all top-down. Industry and government officials decide to do it and then poll some of their constituents about what will work; then they include all of those in a giant framework. That's standard for government, but in successful product development, you instead come up with a feature set, then you pare it down, and then you pare it down again, until you get a core thing that has real value. After that, the users tell you what is really needed. Every unsuccessful company knows you don't take product development advice from people who are not customers but government and academic bureaucracies use naysayers to ask for more time and funding and new committees. And they are not even creating a product, they are only trying to create a framework.

Streamlining drug discovery, and therefore making it less costly and less of a time sink, is easy enough: The bulk of the approval work could be done in parallel with clinical trials, rather than a linear step-by-step process.

"One of the interesting paradoxes of biomedical innovation is increasingly going to be that even though we have the scientific knowledge required to provide potentially better treatments for patients—or even to prevent disease in those who are at high risk—we may be unable to help patients benefit from them anytime soon," said Gigi Hirsch in their statement about their new effort. They don't have an answer, they are simply laying out the problem we all know exists.

Papadaki and Hirsch say that in the last decade more than 100 multi-stakeholder collaborations have been created to try and solve shared problems in cures or diagnosis. No wonder there is 'consortium fatigue' out there. Have these consortia accomplished anything meaningful?

But, they rightly note, that means there is still a lot of optimism. Companies would not keep participating in new committees if they didn't believe this next one would do something meaningful. "A willingness to share proprietary data among industry competitors represents a dramatic shift in the culture of the historically highly competitive pharmaceutical industry," they write.

And share some cost. Creating small molecule medications across disease areas is, on average, a 15 year endeavor that costs over $2 billion and fails more than 95% of the time. And that's without the lawsuits after a product is in use.

The authors note that people experienced in these government/academic/corporate consortia have seen the pitfalls - less efficiency rather than more. 'Consortium fatigue' has set in after 100 of these things.

But clearly some parts do work. An evidence-based approach by outsiders - a Science 2.0 approach to collaboration - could at least help future consortia avoid what has clearly not worked - and perhaps save the US taxpayers $5 billion duplicating what has shown not to work in Europe. The downside is the only way to know is to ask the people who are parts of consortia and their answers will be subjective, which is not all that evidence-based.

Hirsch and colleagues have received a start-up grant to look at case studies (they highlight complex adaptive systems, network mapping, collective intelligence and network-centric warfare but there are a lot more) and try to see what works. That is a good (and far less expensive) start to bring science collaboration into the 21st century.