Innovation is Disproportionately More Effective at Change
The EA community (correctly) identifies a lot of problems, both recently emerged and longstanding problems that we should tackle. However, the focus is usually based on utilizing money or time to either donate money or volunteer time to these causes. This post is about how these strategies might, in some cases, be ineffective at solving these issues, and how innovation and high-level research often drives the solutions to these problems, rather than any of these things, and how even donating to these causes might not be the best way to stimulate these efforts.
The focus on AI safety, and the large amounts of human rights abuses, poverty, and preventable suffering that exists are focuses that should exist. AI misalignment is a problem that funding and time might help, but whether or not this occurs is likely in the hands of a few AI alignment researchers of a certain degree of skill at a certain few companies. A few key breakthroughs are the difference between potential doom and AI alignment.
Another avenue where a few key discoveries may radically reduce suffering is in agricultural sciences. Research in this field might allow billions of people in the longer term to have significantly better lives. Large jumps in automation may speed up the access to a lack of housing some people have, water purification breakthroughs may drastically reduce specific illnesses worldwide.
Historically speaking, institutions with more funding and more personnel do not always yield better results. In the field of science, breakthroughs often arise from individuals who are both talented and highly passionate in their specific field. Money and time provided by people who are not both will quite likely not help in these specific fields.
So, what’s the conclusion? It’s simple: if you’re wondering what to do about a few specific problems, and you’re wondering what to do, donating money to high-yield charities and other sources is better, but, if you can find a higher-yield way in which you can improve your own skills and directly contribute to research in these areas, the potential innovation you might assist in will be worth far more than most other things you can do. Scientific and analytical discoveries often affect a disproportionately larger amount of people than even multimillion or even, in some cases, multibillion dollar donations.
I see that this is your first Forum post. Good work!
Did you know that Open Philanthropy, the largest grantmaker guided by EA principles, funds a lot of scientific research? Their page says:
Besides funding scientific research directly, Open Phil is exploring ways to accelerate scientific and technological progress through better institutions and policies (while trying to avoid increasing catastrophic risks from frontier technologies).
I didn’t know that, thanks for pointing that out!
Hey, welcome to the forum!
As a scientist, I definitely think the value of research is often undervalued by society at large. But it’s important to remember that for every life-changing discovery that revolutionises society at large, there are hundreds of small discoveries that only make things a little better on the margins, and tens of thousands of very promising research paths that end up leading nowhere. When accounting for the impact of the revolutionary change, you also have to account for the cost of all the failed attempts and dead-end research paths, and also for the possibility that a successful, practical approach to a proposed revolutionary tech just doesn’t exist.
It’s still worth it, of course, just be prepared for disappointment along the way.
That’s absolutely true. Incremental discoveries are definitively the norm and practical approaches to revolutionary tech don’t exist. However, it is possible to intentionally focus research efforts on underlooked fields of research that have a large impact. Hype cycles often push research into specific fields and identifying areas where improvements have a high impact can indirectly do a lot of good.