Assumed Audience: My future self to reflect on what was discussed during this webinar
The AI for Good webinar discussed topics on ways we can adapt machine learning tools for social good.
AI is a tool, and like any other tool, AI can be used for positive social value, as well as to increase one’s wealth and power. We therefore need to figure out ways to incentivize research on AI for social good.
Examples of the need for International coordination.
International organization to facilitate AI for social good applications, to act as a hub for connecting problem owners, machine learners, developers, start-ups and non profits the end goal is to create a “clearing house” to bring all bits of the puzzle together to achieve an end goal.
We need a better alignment between incentives and public good. We need to avoid the many issues with current pharma (see books: Big Pharma, Bad Pharma). AI needs data, drug data is expensive and currently secret. There is a need for open discovery or open-source. Unfortunately this is far from the reality, especially since pharma companies are profit driven. Although there is some exchange of data.
What can we do?
Labs (academic or private) get incentivized by receiving funds or recognition for the knowledge they are providing.
R&D is fully open and transparent
Results are free for poor countries, cheap for rich countries
One step: Fund a public data collection pipeline (assays) and provide benchmarks for ML researchers to join drug discovery R&D.
Watch the social dilemma.
- Psychological manipulation: for who’s perspective?
We need to find a different social and economic platform that is a social network as well as creates economic value.
Everyone must benefit from technology or else.