Hacker News Reader

Comment by SubuSS | Hacker News Reader

Would you believe someone who reports back?

FWIW I have a friend who is deep into spirituality, still works at a FAANG but definitely feels like he can attain higher states of awareness in his meditations. I don’t see him active on social media, he leads a reasonably ascetic life and so on.

This person doesn’t have much success in making others believe. At best, he gets ‘hey, I am happy that you’re happy’ from folks like me, most feedback is around ‘mystic woo’. I have another friend deep in this world, but she balances it with real world and doesn’t really proselytize.

I think it is very much a catch 22. You won’t believe it unless you’re doing it. So if belief is required for you to give it a shot, that’s a non-starter.

I would say try it out and make up your own mind! What do you got to lose anyway - stay away from holy men who make you uncomfortable, start with books or YouTube!

View on HN

Just a comment regarding belief in the context of learning: a lot (in fact almost all) of learning is through our innately held beliefs. Without having beliefs we would be no different than an ML model, which just keeps adjusting weights according to the data it receives. Each of us has some beliefs, which "seems right" to us and this is what shapes our understanding of what we see around us, no matter what the "data" says.

If the belief is weak (or false), it might lead to confusion and distrust, which will make it impossible to learn anything.

Also, experiencing heightened states of awareness is not necessarily beneficial, people on drugs experience it routinely.

Are you sure that our beliefs aren't "just" (the appropriate equivalent) of a stimulus or data point that has really high weight, e.g. because it was created during childhood and constantly reinforced?

Take the biases that we've heard about in image recognition where these ML models misclassify black people. If a person had done that you would accuse them of racist beliefs.

Why is the computer model so different? Because you don't want to believe that computers can reach consciousness? (I don't either at this point but time will tell. We don't really know enough - or I don't - about what consciousness is and how it works.)

Well, what about Gebru's findings? Data in and of itself is not beneficial, we do have certain underlying beliefs about appropriateness and goodness. There is an element of innate belief at play.

Otherwise, chatbots are very good at learning abusive and racist language. It's due to our belief that it's not right and decent that we train them using a bias. And it's only due to beliefs that we rein in the rogue AI which misclassifies black people.

Edit: If you still want to think of belief as an input node whose link has weight, then set the weight to infinite

Aka every human has the innate ability to distinguish good from evil. This idea is older than dinosaurs.

Reply on news.ycombinator.com