Blockchain and Science, Not Ready for Prime Time

James Evans discusses blockchain at AAAS conference (A. Kotok)

15 Feb. 2019. A panel of data and behavioral scientists envision great potential for blockchain in science, but show few cases of blockchain in use today among researchers. The discussion took place at the annual meeting of American Association for Advancement of Science, or AAAS, in Washington, D.C.

Blockchain is a system for capturing data about a transaction in a network, but with the data distributed among the various parties to the transaction. Data about the transaction are broken up into blocks, with each block connected in a chain. Each block is also time-stamped and encrypted with an algorithm giving it a unique identifier or fingerprint, also linked mathematically to the previous block in the chain. This linking of uniquely identified and encrypted blocks in the chain ensures the integrity of the data, as well as protects the data from hacking.

James Evans, a University of Chicago sociologist outlined serious problems of reproducibility in science that blockchain could address. Evans says the lack of reproducibility is leading to a “crisis of confidence” in science, with a lack of robustness reinforced by known communities of researchers, rather than seeking out lesser-known, disruptive scientists. Blockchain, says Evans, can make possible registration of hypotheses by researchers, with subsequent data and analysis linked together by disparate groups. This type of process could open up the conduct of science, adding more sources of ideas to the mix.

Krishna Ratakonda, chief technology officer for IBM’s blockchain solutions in Yorktown Heights, New York, says blockchain offers a technology to build trust among disparate groups, enabling more parties to conduct data gathering, curation, and analysis rather than concentrating it all in one lab. “Why have one group do everything?” asks Ratakonda. With blockchain, research can be done in more of an ad hoc manner, leveling the playing field between established labs and more innovative groups.

Kweku Opoku-Agyemang, an economist with Center for Effective Global Action at University of California in Berkeley, says blockchain makes possible a “tamper-proof, indelible time-stamped record of research hypotheses and other decisions.” Opoku-Agyemang discussed pre-analysis plans, a step that requires stating and documenting a hypothesis before data gathering and analysis. These pre-analysis plans can be time-stamped with blockchain and connected to data collection and analysis in a more robust and trusted process.

While the idea of blockchain has promise, the panel members could offer few examples of its use in scientific practice. Opoku-Agyemang mentioned one test of blockchain with peer review. Even more concerning, Ratakonda told of a blockchain application in the food industry, to better track food items through the supply chain. One spice manufacturer, says Ratakonda, took 5 years before the company could furnish data for the project, suggesting more than technology needs to be considered.

More from Science & Enterprise:

• Blockchain Manufacturing Model Described
• Genomics-Blockchain Company Launches, Raises $4.3 Million
• Analysis Reveals Ransomware Payment, Support Networks
• MIT Building Statewide Health Innovation System
• Tracking Algorithms Devised to ID Human Traffickers

Disclosure: The author owns shares in IBM.

*     *     *