After 2 years of its booming.. my thoughts on A.I

 

• Generative A.I is best in 2 things:

> NLP: Good at communicating acoustically and scripturally:

○ Acoustically:

Vocally: has a good NLP, thus will be good for Communication with humans; e.g.: Chatbots, Interface for Vocally impaired, translation, and Singing

Aurally: Simulating Speech, Music, Tunes, Sound, Noise, etc

○ Scriptually: Writing Texts of prose & powems

> Visual-esque: Simulating psychedelic, hallucinogen, delirium, and psychovisual experiences & illnesses for medical and recreational reasons.

POV: Visuals comes off a little bit as the byproduct of hallucinating when the A.I can’t reproduce visuals with the real world’s sense of depth, intuition, and naturalness (You can always spawn & market that as an intentional art form or such)

 

 

• Generative A.I is just a piece of the jigsaw puzzle to allow us to create the 1^st artificial conscious creature

 

• Generative A.I seems to not have conscious, sapience, or sentience:

 

• Given the Generative A.I’s advanced NLP, a human could easily be deceived of what it’s capable of; this is because historically, a sophisticated horizontal (vocabulary) or vertical (speech) manipulation of words a language by a human, correlates positively with the developed cognitive capabilities of the said human.

Given this, when you’re dealing with an agent whose sophistication come from the objective experience of being built for such purpose VS that of a human whose sophistication comes from his subjective experience, it becomes hard to decide on whether current A.I is sentient, sapient, or even conscious.

This is enforced by the realization that some of the biggest A.Is aren’t open-sourced, which means the marketed A.I, may not be the same as the A.I behind these corporations: whether they overpromise on the A.I, or they restraint it.

Realistically though, and regardless of the political aspect of things, it wouldn’t be in the interest of the corporations to restraint their A.I, especially that the competition right now is getting tighter.

So my guess would be that the only depth similar to that of reality  that current A.I has, is that of Language.

Thus I will have at the moment to say that current A.I is neither conscious nor sapient..

Concerning Sentiency, I’ll have to see it interact with humans randomly, i.e. without requiring to be enable with a scriptural or acoustic prompt

Unless this happen, I can’t say it’s a real agent either; it may have a level of agency of its information crawling, but that’s equivalent to every Narrow A.I, and every computer program humans have ever made.

 

• I can’t say that Adversarial A.I is proof of the A.I not being sentient, sapient, conscious because even living beings have certain subconscious -& physical- switches that can corrupt their data i/o

 

• I also can’t say the A.I not sentient, sapient, conscious of it didn’t defend itself: many creatures who lived in isolated islands without any predators went extinct when a predator got into the island just because such creatures didn’t had a psychological fight or flight mechanism.

Moreover, even if an A.I developed defensive mechanisms, they may not be efficient at the moment (say for example an ant wanting to defend itself when you lift your foot to stomp it).