A short thread on Machine Phenomenology

I've been busy on Threads. Still figuring out how to cross-post ideally.

https://www.threads.net/@_jerng/post/DGwyeIaT0YZ?

A short thread on AI.

I'm not certified to have an opinion. You don't have to take this seriously. But maybe it can add colour to your view.

What do we mean by "artificial" "intelligence"?

Artifice : made by skill

Intelligence : to be like us

( because we cannot possibly have other definitions of intelligence )

1. What is the goal? To be like us, can we say, that we know what we are?

In philosophy/ metaphysics / ontology : we begin by proposing a manifest. A list of things that make up the human experience.

What is conscious experience?

The WhatList

How does it work?

The HowList

What a minute ... none of the business people are talking about the WhatList. They are talking about,

What can we do with AI?

The UsefulList

Wow. That rips things up, doesn't it?

2. The study of theWhatList is called phenomenology.

This is to talk about experience, and how it is experienced from the naive 1st person point of view.

The word phenomenon ( what is consciously perceived ) is different from noumenon ( what is not perceived ). Some people say, qualia. You can Google it.

And the richness of our lives comes from this, first person experience. Yes? Feel free to disagree.

But this is not commonly discussed today in AI. Why?

3. Solipsism means you know about you, but you don't really know about anyone else.

You talk to your friends, but do they really exist? How do you know that what they say, and what you understand, are the same? 

We do not. We ASSUME that we understand each other. This is a big part of why communication fails - sometimes we assume wrongly.

Some of us don't care about other people's experience, as long as we can get something useful from using them. ( The UsefulList )

Now we treat AI like this

4. This is sad. Not just because, we don't want to treat AI like proper people.

It is more poignant that we cannot achieve a proper implementation of AI unless we think about AI as proper people.

Why?

Simply because, self-awareness is what characterises us. So if we do not properly write the WhatList, we cannot properly build the HowList.

And we will always build only some half UsefulList, just not as well as we could if we properly built the WhatList first.

And thus we poorly craft AI, until such time as we begin to take the study of phenomenology seriously.

That will be seen as front and centre once we hear researchers and business people talking about what it feels like to be human, and how they are sure that AI feels exactly the same way.

As for me, I studied this a long time ago. I am just sharing my experience with you.

Thank you for reading!