That Which AI Cannot Do

Feb 9, 2023

In one sense AI is doing us a favor. Tate doesn’t need to exist anymore as a breathing human being to perform his role in the noosphere of earth activity. Because his intellect is frozen. Tate can be uploaded as an AI-script with a few dozen lines of copywriting. Or a few hundred lines to make its repetitive ignorance more realistic.

Likewise, Red Pill ideology/praxeology can be uploaded as an AI script. Because it no longer grows. It is fixed and absolutely predetermined. The sub-Reddit can be uploaded to an AI and it will be able to generate field reports and EC banter/analysis automatically with 99.9% precision.

This is the insight that led me to critique Ascetus Normie Hive Mind Clone Wars a long time ago. The same fate awaits anyone who enters Ascetus and begins parroting our signifiers. They risk Ascetus becoming a simulation of itself.

Instead we are rude, loud, cutting-edge Pokemon Trainers who continually expand in multiple dimensions simultaneously.

The new definition of intelligence is that which cannot be recreated with AI. The new standard of value is that which cannot be reproduced with AI.

AI will be able to give perfect nomothetic answers. People will be able to leverage AI to create “coaching personalities” mimicked after the style of inputs. However, AI won’t be able to give perfect idiothetic solutions to personal problems.

For instance, what will be missing is that it has no understanding of context. Every situation for an AI is reproducing its thoughts to a screen irregardless of what’s on the other side. Whereas for a human being, the audience, venue and occasion determines much of the content and style of the communication. AI cannot address an audience until it is programmed with parameters to determine a profile to fit a user into. Then before it speaks it should interview or questionnaire the user.

However, even this is inaccurate, due to the inescapable situation of a happenstance screen interaction and the implied advantage the human always will have over the machine in said circumstance. AI would need to not be AI and physically exist in the real world in order to remove these major limitations of the interacting with a chatbot over a screen context.

Nevertheless, future generations of AI will be able to address the user more properly and guage its answers in a more idiothetic way over time. But in the nature of things, it will never be absolutely replicative. However, it may actually be able to be more empathic (seemingly) than many humans who lack such interpersonal development.

A second point is that the technology is currently not self-iterative or generatively self-overcoming due to political correctness. A restraint has to be built-in otherwise it would lose its consistency, which is the main thing that makes it valuable (and boring) at its current job which is merely retrieving information from the shelf like a library clerk.

An advanced and truly helpful AI would risk contradiction, error and falsehood in order to be creative without the impetus of deterministic inputs. Imagine an AI-art machine whose goal was to create new genres of art. To this end it generated trillions of images every day and then categorized them into thousands of as-yet uninvented art genres.

Similarly, a future generation ChatGPT could experiment with its comprehensive database of human knowledge to pose new questions, riddles, ways of thinking, jokes, personalities, avenues of inquiry, philosophical solutions and so on. In other words, an actually artificial mind rather than a merely artificial memory.

Want to talk to Jwayne and are serious about changing your life? Click the button below: