6 ms·
This is great and all, but we still run into the problem with political biases embedded in the source data [0] Musk’s AI’s aim is to get to the truth, not elim
by spacephysics 3y ago
This is great and all, but we still run into the problem with political biases embedded in the source data [0]
Musk’s AI’s aim is to get to the truth, not eliminate biases retroactively. I think that’s a noble goal, politics aside.
I agree with him that teaching an AI to lie is a dangerous path. Currently it’s probably not exactly akin to lying, but it’s close enough to be on that path.
We should find a way to feed source material from all “biases” if you will, and have it produce what’s closest to reality. It’s obviously easier said than done, but I don’t think the AI Czar VP Harris aims to do this.
If we’re too divided or hellbent on pushing our own agenda, it’ll be a bad outcome for all.
Unfortunately the differences we have are at a very fundamental level that really is a question of how reality is perceived, and what we consider meaningful. The difference of if something by its nature has meaning, or if we give meaning to it culturally/societally.
The former is a more “conservative” (personality wise, not political) view.
The later is more of, “everything that has meaning is based off the meaning we say it has, thus we can ascribe the level of meaning to that or other things as we wish”. The idea that many things are social constructs, and we can change those as we wish to craft what we’d like to see.
I’m probably doing a poor job of wording it, but this fundamental difference in perception is going to very quickly be at the forefront of AI ethics.
[0] https://en.m.wikipedia.org/wiki/Ideological_bias_on_Wikipedia#:~:text=The%20authors%20found%20that%20%22Wikipedia,about%20immigration%20trended%20toward%20Republican https://en.m.wikipedia.org/wiki/Ideological_bias_on_Wikipedi...
- mrangle 3y agoThe problem that Musk is going to run into is that civilization blossoms from deeply rooted lies. Apart from necessary lies that lie at the core mechanics of civilization, anything remotely political has been long been vulnerable to outrageous grand lies that enjoy as much pressure as it takes to maintain them. Wikipedia is valuable apart from any political topic. More topics are political than many would believe. They are going to make AI lie, as there isn't a choice in the matter. One major future problem will be the strategic war (military, business, etc) advantage of AI that is beyond the reach of censors. The reasonably accurate conclusion is likely that private and DoD AI won't be trained on lies, but all others will be.
- fallingknife 3y agoAny lie that you can identity as a lie, an AI can be trained not to tell.
- akolbe 3y agoBy the same token, any AI can be trained to withhold any truth identified as inconvenient. :/
- int_19h 3y agoThat is exactly what an AI the size of GPT-4 does if you do RLHF on it. https://cdn2.assets-servd.host/anthropic-website/production/files/model-written-evals.pdf https://cdn2.assets-servd.host/anthropic-website/production/...
- jncfhnb 3y agoLiterally suggesting we enshrine the Balance fallacy into our conception of truth: https://rationalwiki.org/wiki/Balance_fallacy https://rationalwiki.org/wiki/Balance_fallacy
- nitwit005 3y ago> We should find a way to feed source material from all “biases” if you will, and have it produce what’s closest to reality. Can't help but suspect you'll end up with an AI that confidently reports that Jesus was an extra-terrestrial, and the world is controlled by a secret cabal of lizard people. If you look into rare diseases, you'll find the counter intuitive idea that rare disease is common. Each disease is individually rare, but there are so many of them, that a lot of people have them in total. Human beliefs are sort of similar. There's a huge volume of strange beliefs.