On Impactful AI Research

An insightful article discussing how to structure AI research for impact through open source, focusing on investing in projects over papers, selecting timely problems with potential, iterating quickly, popularizing ideas, funneling built excitement into OSS growth, and continuing to invest in projects.

• Grad students often feel projects and code releases are not incentivized.
• Research impact can be made through open-source artifacts.
• Invest in projects with a larger vision beyond paper count.
• Identify problems with 'fanout' and significant 'headroom' for impact.
• Continually invest in and grow open-source research.