Hi all, mechanical + electrical engineering undergraduate who enjoys research and hopes to pursue a career in it. While I’m not especially interested in the main EA cause areas (AI safety, biosecurity, cybersecurity, etc.), I’d still like my research to have as much impact as possible.
Does anyone know of any websites that compile neglected scientific problems beyond those discussed by 80,000 Hours? Not necessarily looking for problems ranked by overall importance, just collections of underexplored problems where additional work could have (somewhat) decent marginal value.
If anybody has any insight, would really appreciate it.
do you have any criteria for impact? Programmable matter? Fusion? microfission? batteries? novel wireless? new materials in general? droneswarms? autoresearchers?
If nobody is working on a problem, there’s usually a reason:
It’s not a real problem
It’s already solved—you just don’t know it yet
There’s some other reason not to work on it (e.g. nobody will fund it)
Heuristics for finding good problems others will miss
You could realistically be the best in the world at it, even with 100+ people working on it. Rewards (money, status, influence) go disproportionately to the best people in a space. You can channel those rewards into impact.
It’s high-risk, high-reward. Most people are overly risk-averse.
It takes a long time (people overly discount the future) and has a strong feedback loop, so you know you’re actually making progress.
Solving it would cost the incumbents money. Permanent solutions threaten expensive, recurring short-term ones e.g. if Ozempic’s makers had a one-shot vaccine with the same effect, they wouldn’t sell it.
You can otherwise establish that people avoid it for non-impact reasons (e.g. it’s low status, boring)
Only powerless entities have the problem e.g. animals, people in extreme poverty).
Existing solutions are terrible but still widely used. strong validation that it’s a big problem.
[Question] Resources for finding neglected scientific problems (beyond EA)?
Hi all, mechanical + electrical engineering undergraduate who enjoys research and hopes to pursue a career in it. While I’m not especially interested in the main EA cause areas (AI safety, biosecurity, cybersecurity, etc.), I’d still like my research to have as much impact as possible.
Does anyone know of any websites that compile neglected scientific problems beyond those discussed by 80,000 Hours? Not necessarily looking for problems ranked by overall importance, just collections of underexplored problems where additional work could have (somewhat) decent marginal value.
If anybody has any insight, would really appreciate it.
do you have any criteria for impact? Programmable matter? Fusion? microfission? batteries? novel wireless? new materials in general? droneswarms? autoresearchers?
No real criteria, am open to anything (being an undergrad, I think I have sometime to explore).
Neglectedness is often a trap.
If nobody is working on a problem, there’s usually a reason:
It’s not a real problem
It’s already solved—you just don’t know it yet
There’s some other reason not to work on it (e.g. nobody will fund it)
Heuristics for finding good problems others will miss
You could realistically be the best in the world at it, even with 100+ people working on it. Rewards (money, status, influence) go disproportionately to the best people in a space. You can channel those rewards into impact.
It’s high-risk, high-reward. Most people are overly risk-averse.
It takes a long time (people overly discount the future) and has a strong feedback loop, so you know you’re actually making progress.
Solving it would cost the incumbents money. Permanent solutions threaten expensive, recurring short-term ones e.g. if Ozempic’s makers had a one-shot vaccine with the same effect, they wouldn’t sell it.
You can otherwise establish that people avoid it for non-impact reasons (e.g. it’s low status, boring)
Only powerless entities have the problem e.g. animals, people in extreme poverty).
Existing solutions are terrible but still widely used. strong validation that it’s a big problem.
I am not aware of such a list, but I will propose a scientific problem neglected for cultural reasons:
Are UAPs advanced technology? We’ve received funding to investigate this question with distributed passive radar nodes.