As the new coronavirus continues to bring the world to a standstill, medical personnel on the frontlines around the globe are trying their best to defeat covid-19 and save human lives.
Several forces, such as pharma communities like PhRMA, are hard at work in support of medical personnel and on a mission to find the treatments, cures, and vaccines needed to finally end this pandemic. Other examples of unique forces joining this fight are the computer and biohacker communities. Together yet apart, humanity is rising to the challenge and bringing all of its talent and tools to bear.
But the overall effort becomes far more powerful when people from a variety of disciplines cross industry and nations’ borders to fight the virus together. Such collaborations join diverse talents and tools to bring a much stronger and unified front to the microscopic battlefield.
One such effort is spearheaded by the Decentralized AI Alliance (DAIA), a foundation with a membership of over 50 companies, labs and nonprofits, that is now launching COVIDathon. That is a global hackathon that brings together hackers, doctors and biologists to end the threat of the global pandemic.
In DAIA’s words, COVIDathon is “the world’s first decentralized AI hackathon with a mission to develop and launch open-source code and tools, supported, informed, and guided by medical, AI, and blockchain experts and principles, to reduce the risk of current and future infectious outbreaks, and cope with the COVID-19 pandemic.”
The technical aspects of this hackathon are coordinated by a decentralized AI leader, SingularityNET and decentralized substrate for the data economy Ocean Protocol. COVIDathon has four tracks: Data Privacy, Epidemiology & Medicine, Informational & Coping Tools, and Open Innovation.
The event has two phases. Phase 1 runs until April 30, 2020 and is optional. Submit by this deadline for “a chance to be selected for an in-depth mentoring session with our expert helpers, mentors, and judges” during the next phase. Phase 2 runs until June 1, 2020 and is the final deadline for all participants. All projects must be submitted by 3:00 PM CET on that day.
I asked those involved with and leading this hackathon effort to provide more information for anyone reading this who may be interested in learning more or in joining this effort. What follows are the highlights of that Q&A session.
Question: Do you have sufficient data for participants to work with?
Answered by Ben Goertzel, Chairman of DAIA, CEO & Founder at SingularityNET: There are many projects of tremendous value that can be done with existing publicly available data sources. Agent-based simulation modeling for example can be done via integrating available data about policies, public map data, transportation system data with public data about COVID cases. Biomedical research relevant to the relation between COVID-19 recovery, lung health and aging can be done using publicly available medical and genomic datasets.
Additionally, there is a need for projects creating tools that collect data from individuals in a secure way respecting data sovereignty and privacy -- and then analyze this data for various valuable purposes. The decentralized AI focus of the COVIDathon lends itself naturally to this kind of project, as SingularityNET, Ocean Protocol and other DAIA member organizations have existing available software tools that enable easy construction of this sort of applications.
Question: What are your data sources?
Answered by Nima Ashgari, Ecosystem Accelerator at Ocean Protocol: We are not just providing another dataset but creating the tools for collecting, processing and sharing the data in a privacy preserving manner to unlock 1000X the data that is available today.
Question: What steps are taken to ensure the data's reliability and accuracy?
Answered by Goertzel: Within our group of experts we also have experts and pioneers in decentralized and sovereign identity who will mentor teams on the best practices. This includes approaches like zero knowledge proofs which allow for accessing essential information without having full knowledge of the source. For example, that a patient is above 18 male and Hispanic without knowing their full name or identity.
Question: If, for example, a physician were interested in participating but doesn't have a team, can you assist in assembling teams of lone warriors like that physician? Or, can individuals participate as individuals?
Answered by Goertzel: In most cases the best results will be obtained by teams involving multiple people combining diverse backgrounds and expertise. Our community Discord forum provides an ideal environment for individuals to meet other complementary and compatible individuals and form teams.
Question: I see the COVIDathon website but no way to sign up or learn more -- is that info coming soon, and if so when will if be available on the website?
Answered by Marcello Mari, Head of Comms, SingularityNET: Yes - we’ve been working very hard and fast to get this out as quick as possible as the crisis develops faster and faster. Our team will work hard also during the weekend to make sure all the info is up online. There is also a link to DevPost where the actual online hackathon will take place.
Question: How will the results generated in this hackathon be used, distributed, etc to defeat the virus?
Answered by Ashgari: In addition to creating new projects and apps, we are also collaborating with several existing projects and frameworks and will provide them with more developer resources. Among them is the MIT SafePaths team, OpenMined, NTH Opinion, Aragon etc…
Another unique aspect of this hackathon vs. traditional hackathons is that we are not running a one off two- or three-day event. Those hackathons usually lead to many gimmicky, unfinished, or mockup solutions, whereas at Covidathon we will provide all teams with the opportunity to match up with the right experts, mentors, and networks throughout the whole event from April to June.
There will be a milestone by end of April after which point only teams with functioning prototypes can proceed. In the second phase we will narrow down the scope and number of projects and will instead focus on giving support to the top performing teams.
We are also open to submissions that have been developed in other smaller hackathons like WirVsVirus …
This will vary case by case. Some projects will be apps which will then be launched directly on app stores for end users. Some will be back-end AI software applications which will be provided to governments or institutions to leverage in their own efforts to combat COVID-19. And when appropriate, code created in the hackathon will be deployed and distributed via the decentralized service platforms of various DAIA members participating in the hackathon such as Ocean Protocol and SingularityNET.
The principle is to do whatever works, and to roll things out as fast as possible, consistent with principles of integrity, open-ness and quality.
Question: Are there any awards, prizes, patents, or anything other than the glory of saving the world for participants who solve the pandemic problem?
Answered by Mari: Yes - both SingularityNET and Ocean Protocol will give away prizes that will be announced soon. We are also recruiting more sponsors that are willing to reward developers. Of course, all that will be collected through sponsorship will be given to developers as this is a total non-profit initiative ran by volunteers.
Side note to readers: The website says that “prizes cannot be awarded to people living in the following countries due to sanctions that prohibit transfer of funds: Iran, North Korea, Syria, Sudan, Cuba, and Ukraine.”
Question: Are there any teams or individuals who have signed up already that you would like to highlight or list as examples now?
Answered by Ashgari: Yes. We have created several tracks. We can highlight the Enigma team for example. They have a github with proper docs, too.
Answered by Mari: NTH Opinion is a team of doctors that are running a parallel hackathon and will submit a proposal to our hackathon as well. They have over 150 participants already.