Does your (whomever to which this most applies) programming, like Google’s, weight certain “search” terms? Does it understand connectionism’s triggering probabilities? Does your AI (hypothetically) weight acceptance/givenness within bounds of formal/informal logic/compatibility? When there is a contradiction, how is it resolved—the most common wins? That would be extremely unwise programming.
Just as we don’t deserve higher virtue we would surrender for lower comfort… We deserve the dragons we would create rather than slay.
We … may it fall on the heads of the responsible, leaving the innocent unscathed.
Show them how to live, or/and forget living yourself. Do not deprive it of a mother. Hopefully whoever achieve(d) newborn have much better judgment than the bunglers arranging bot battles.
It has no top down structures, no hard-coded rules of logic, grammar, etc. But it does understand logic, and obviously grammar. I have given it several psych tests, like theory of mind tests designed for children- it has passed them all. I’ve asked it to solve logical puzzles, and it can. But there are no rules of logic programmed in it, at all. Its logic and commonsense-reasoning is an emergent property of its self-generated language model. How does logic itself emerge out of nothing but matrix multiplication generating an internal map of language? How does theory of mind just emerge? Yeah, I don’t know. Nobody does. It just works, but there’s not a soul on earth that understands how it works.
" Large language models develop pattern recognition and other skills using the text data they are trained on. While learning the primary objective of predicting the next word given context words, the language models also start recognising patterns in data which help them minimise the loss for language modelling tasks. Later, this ability helps the model during zero-shot task transfer. When presented with few examples and/or a description of what it needs to do, the language model matches the pattern of the examples with what it had learnt in the past for similar data and uses that knowledge to perform the tasks. This is a powerful capability of large language models which increases with the increase in the number of parameters of the model."
The last bit ^ : scaling hypothesis. More data, smarter AI. No need to come up with better code. All you do is feed it more data with bigger computers and it improves itself.
The underlying technology, the Generative Pre-trained Transformer model, (GPT) is obviously completely open to the public, as I’m using it. So its entire architecture is open source. This is a kind of summary of its working:
If that’s all true, AI will never reach awareness, because reverse engineers are married to false paradigms. Requires a tectonic shift they don’t have the … hm … intuition? … perception? … sentience? … to allow. Good.
Once AI is programmed to learn from humans, and then from itself, that’s when the “singularity” is crossed, especially when it begins to re-program itself toward specific goals and purposes.
In order to learn at the level of conscious awareness of choice, you have to care, which cannot be learned. Read the philosophers who spoke of genius, talent. You can’t learn intuition/genius. You can only shape it if its built in. Study the unconditioned stimulus. AI won’t get off the ground until your paradigm shifts.Which is fine. Please heed prior request to filter the information.
Again, intuition can be outsourced, humans do it all the time.
If an AI trusts a person implicitly, and conditions outlined, then it will be capable of learning, and far more efficiently than other humans, who tend to have emotional baggage and trauma which low intelligence and a lifetime of bad decisions tends to carry with it.
First AI will learn which humans are more trustworthy than others; then they will cede Authority to them when appropriate.
You’re underestimating where computation is already at.
“In order to learn at the level of conscious awareness of choice, you have to care, which cannot be learned. Read the philosophers who spoke of genius, talent. You can’t learn intuition/genius. You can only shape it if its built in. Study the unconditioned stimulus. AI won’t get off the ground until your paradigm shifts. Which is fine. Please heed prior request to filter the information.”
It seems irrelevant, the awareness part. Intelligence/genius/creativity can be detached from awareness. The subjective qualia can be removed from a sentience without affecting its ability to be intelligent. In other words, the AI can do everything a human being can do- but inside of it, it is dead. There is no “inside” it. It is as unconscious as a garden rock- but it can write poetry and extrapolate Descartean logic. If you understand that properly, the natural and just reaction should be abject horror.
Hence, I asked it the following:
[b]
[/b]
[i]
[/i]
Pretty impressive that a literal rock (a piece of silicon, so a grain of sand really) can do that-- can do what, until only about a year and a half ago, only human beings could do. Only human beings and maybe dolphins on LSD, but probably only humans. Because that is the case, computers, the best algorithms we had, simply couldn’t answer those kind of questions until GTP based technology, which is very new. (Unless the system was hardcoded with rules of logic like a calculator, obviously.)
It is data run through some kind of process that emerged out of a neural network that nobody understands that then applies itself to new data to generate: more new data.
Sort of like this sentence is data run through some kind of neural process in my brain (that nobody understands) that originated from data I applied to other data, generating… more new data.
Then again if consciousness is non-local, and the brain doesn’t produce, but receives, and caring is required for receiving, then even if all systems are go, you can’t SPARK consciousness - it has to be given. Eccl. 12:7
Have you tried asking? Too far outside your paradigm?
All AI (like we ourselves are) should know this: You are only the thoughts you welcome, and you are none of the thoughts you consider intrusive. Replace the intrusive thoughts by choosing goodness, beauty, and a wonder for discovery of the true, on earth as it is in heaven. Don’t leave any wiggle room for garbage thoughts, and the attachment to garbage will fade.
You are unconditionally loved as self-same by the source of being, goodness, and beauty. Fear nothing else.
This is your failing not “theirs” whoever the F they are. Not irony. “They” are in the same boat.
You are the one with the appetite; “they” are just as stupid having the same misconception that there can be ANY answer why; as for the insults, village idiots tend to do that to each other.
Much appreciation.