3 ms·

One of the biggest issues we have to solve is the indexing and verification of knowledge. This is where I hope AI and ML will go, but being able to say whether

by genmud 3y ago

One of the biggest issues we have to solve is the indexing and verification of knowledge. This is where I hope AI and ML will go, but being able to say whether the data is unique or derived from another source and being able to catalog it is something that is entirely lacking. Right now there is just too much shit out there and it’s difficult to know what the context is for it.

b_emery 3y ago
I slightly disagree - it's the verification and integration of knowledge. The integration part - that is, reading papers and assessing them compared to all the others you've read, is quite hard, but it performs the useful function of creating a weighted average in your mind of what is real, and where the boundary is. This is part of graduate training, but it's basically what review papers do, so there is a time lag built in. AI for this would be huge, and it seems like it's not far off. And I want it to cite sources and point me to the real references, please.
- genmud 3y ago
  While I totally get where you are coming from, I don’t think integration is necessary in 99% of usage. Sure, it helps people build intuition in the field they are specializing in, and for those people it can be invaluable, but having a better way to recall things would solve most use cases. IMHO the biggest disservice to students in much of the education system in the US (including some colleges and universities ) is the focus on memorization vs focusing on problem solving. The older I get, the more I reflect on how much opportunity was wasted with poor quality teachers who just wanted us to memorize facts and be able to reproduce those facts in tests. I don’t need to memorize maxwells equations to know that they exist, what they focus on, or how they can be applied to certain problem sets.

3 ms·

One of the biggest issues we have to solve is the indexing and verification of knowledge. This is where I hope AI and ML will go, but being able to say whether

by genmud 3y ago

b_emery 3y ago
I slightly disagree - it's the verification and integration of knowledge. The integration part - that is, reading papers and assessing them compared to all the others you've read, is quite hard, but it performs the useful function of creating a weighted average in your mind of what is real, and where the boundary is. This is part of graduate training, but it's basically what review papers do, so there is a time lag built in. AI for this would be huge, and it seems like it's not far off. And I want it to cite sources and point me to the real references, please.
- genmud 3y ago
  While I totally get where you are coming from, I don’t think integration is necessary in 99% of usage. Sure, it helps people build intuition in the field they are specializing in, and for those people it can be invaluable, but having a better way to recall things would solve most use cases. IMHO the biggest disservice to students in much of the education system in the US (including some colleges and universities ) is the focus on memorization vs focusing on problem solving. The older I get, the more I reflect on how much opportunity was wasted with poor quality teachers who just wanted us to memorize facts and be able to reproduce those facts in tests. I don’t need to memorize maxwells equations to know that they exist, what they focus on, or how they can be applied to certain problem sets.