Let me introduce myself. My name is Kristian Rönn and I work for the Future of Humanity Institute (FHI) as a project manager. This is my first LessWrong post.
tl;tr:
Wouldn’t it be useful to have a directory of researchers (dependent or independent) that do research with high consequentialist significance?
Solution: sign up at mirivolunteers.org and help us crowdsource such a list!
The best contributor will win a 30 minute skype conversation with one of the FHI researchers!
The long version:
We all know that some scientific research has extremely high consequentialist significance (good or bad). But much research with high consequentialist significance also falls squarely within the scope of philosophy (see William MacAskill’s blog post on high-impact philosophy). Such important areas of philosophical research include ethics, epistemology and philosophy of mind (especially applied to existential risk research). Yet professional philosophy is a diseased discipline.
First, professional philosophers very often work on trivial questions (e.g., “how is knowledge best analyzed?”).
Secondly, when they do address questions of importance, they are seldom motivated by a concern to find the truth, and answers are instead assessed on poor criteria, such as originality, elegance, “cleverness”, and tendency to vindicate cherished beliefs.
Fortunately, there are a few philosophers addressing many of the relevant questions, and from whose work we can learn a great deal. It also seems like the problem of building a friendly AI forces us to think about the philosophical questions mentioned in a more formal/rigorous way. Because in order to design a friendly AI I need to figure out:
How to make correct inferences in an optimal way (epistemology). E.g. Gödel machines and AIXI.
What set of things/concept exists in the agents ontology (metaphysics); and more specifically.
How to distinguish between those things that are a part of the set of conscious things and those who are not (philosophy of mind).
How to track the identity of things in time and space (personal identity over time).
(Author’s note: I believe some of the problems listed above are not solvable, since they are just a quibble over semantics)
This means that a lot of the most important philosophy is done by computer scientists, physicists and mathematicians (sometimes without even knowing it). All non empirical research could (very broadly speaking) be classified as philosophy.
-Wouldn’t it be useful to have a directory of researchers (both in science and philosophy) that are doing research of high consequentialist significance? -The answer is: yes! -But how can we make such a list? -By letting everyone on LessWrong and elsewhere crowdsource the list, by filling out this Google Spreadsheet (you are encouraged to add independent researchers such as LW contributors on the list).
If you also sign up on mirivolunteers.org you will be able to collect points for your contributions. The volunteer who collect most points to this challenge over the next 30 days will be rewarded with the opportunity to video chat with an FHI researcher—your choice of either Stuart Armstrong, or Nick Beckstead!
(I would like to thank Pablo Stafforini for inspiring and writing an earlier document on which parts of this post are based)
Help FHI crowdsource a directory of researchers with high consequentialist significance and win a prize
Let me introduce myself. My name is Kristian Rönn and I work for the Future of Humanity Institute (FHI) as a project manager. This is my first LessWrong post.
tl;tr:
Wouldn’t it be useful to have a directory of researchers (dependent or independent) that do research with high consequentialist significance?
Solution: sign up at mirivolunteers.org and help us crowdsource such a list!
The best contributor will win a 30 minute skype conversation with one of the FHI researchers!
The long version:
We all know that some scientific research has extremely high consequentialist significance (good or bad). But much research with high consequentialist significance also falls squarely within the scope of philosophy (see William MacAskill’s blog post on high-impact philosophy). Such important areas of philosophical research include ethics, epistemology and philosophy of mind (especially applied to existential risk research). Yet professional philosophy is a diseased discipline.
First, professional philosophers very often work on trivial questions (e.g., “how is knowledge best analyzed?”).
Secondly, when they do address questions of importance, they are seldom motivated by a concern to find the truth, and answers are instead assessed on poor criteria, such as originality, elegance, “cleverness”, and tendency to vindicate cherished beliefs.
Fortunately, there are a few philosophers addressing many of the relevant questions, and from whose work we can learn a great deal. It also seems like the problem of building a friendly AI forces us to think about the philosophical questions mentioned in a more formal/rigorous way. Because in order to design a friendly AI I need to figure out:
What utility function it ought to have (ethics). E.g. Coherent Extrapolated Volition.
How to make correct inferences in an optimal way (epistemology). E.g. Gödel machines and AIXI.
What set of things/concept exists in the agents ontology (metaphysics); and more specifically.
How to distinguish between those things that are a part of the set of conscious things and those who are not (philosophy of mind).
How to track the identity of things in time and space (personal identity over time).
(Author’s note: I believe some of the problems listed above are not solvable, since they are just a quibble over semantics)
This means that a lot of the most important philosophy is done by computer scientists, physicists and mathematicians (sometimes without even knowing it). All non empirical research could (very broadly speaking) be classified as philosophy.
-Wouldn’t it be useful to have a directory of researchers (both in science and philosophy) that are doing research of high consequentialist significance?
-The answer is: yes!
-But how can we make such a list?
-By letting everyone on LessWrong and elsewhere crowdsource the list, by filling out this Google Spreadsheet (you are encouraged to add independent researchers such as LW contributors on the list).
If you also sign up on mirivolunteers.org you will be able to collect points for your contributions. The volunteer who collect most points to this challenge over the next 30 days will be rewarded with the opportunity to video chat with an FHI researcher—your choice of either Stuart Armstrong, or Nick Beckstead!
(I would like to thank Pablo Stafforini for inspiring and writing an earlier document on which parts of this post are based)