What is the end game with AI ?

[Wilhelm44]

6 hours ago, LordFall said:

Yes indeed AI will first take jobs and leave some chaos for a bit. Then UBI will be adopted. Then universal high income. 

Along those timeline we will master Fusion Power and thus unlimited clean energy so basically free infinite productivity.

We will also master biological immortality and potentially even raise the dead with quantum computing and past/future simulations.

Then idk chill out and have a good time? 

Why have biological immortality as a goal ? That sounds more like a nightmare than utopia.

[LordFall]

@Wilhelm44 Really think about what you've just said. From age 30 our bodies start to basically degrade and die over time and our vital functions just get worse and worse. We've been able to romanticize it and make it sound like a beautiful thing but go to a nursing home/hospital and see old people having to be helped to shit, piss and walk and perhaps rethink the beauty of it.

Some of the smartest people in the world shoutout to David Sinclair and his lab at the Harvard School of Medicine are thankfully working hard to end this

Thank fully it's looking like 2030 will be the inflection point where humans on average with incretemental tech will be able to start defeating aging and defeating death for good. 

I don't think it'll be mandatory you can kill yourself if you want to but why? This is the pinnacle of evolution as far as I'm concerned and perhaps what the religions have been alluding to for thousands of years through various mystics and prophets. 

[Wilhelm44]

@LordFall   Thanks for the videos, will check them out.

I'm all for improving the quality of our biological lives, perhaps even extending them by a few years.

But somehow the notion of living 'forever' in biological form rings a bit eerily.

The main motivation here is probably the false belief that our essence is not immortal already. 

Which means it's mostly coming from a place of fear, instead of love.

And that usually don't end well.

[Wilhelm44]

@LordFall  Also, it's kinda selfish. Imagine 8 billion people living 'forever' - ie not really much space for new souls to come in and experience life on earth.

[BlueOak]

On 07/04/2025 at 6:36 AM, Wilhelm44 said:

@LordFall  Also, it's kinda selfish. Imagine 8 billion people living 'forever' - ie not really much space for new souls to come in and experience life on earth.

Earth is a speck of sand in infinity.

[BlueOak]

On 05/04/2025 at 10:53 AM, Wilhelm44 said:

Why have biological immortality as a goal ? That sounds more like a nightmare than utopia.

As long as death can be chosen on our own terms, after careful consideration with specialists, and as above, there is room for infinite expansion, we'll be fine. People will still die from illness, accidents, war, choice, emergencies etc. If I were structuring it, there would be choice points set every 100 years to review whether you wanted to continue living, over a full year or two before a choice was made consulting with experts in their fields. It seems a reasonable way to go about giving people the amount of immortality they want. - Though everything physical has a lifespan, and nobody will convince me otherwise, it'll just be much longer naturally.

Edited Wednesday at 08:22 PM by BlueOak

[Wilhelm44]

11 hours ago, BlueOak said:

Earth is a speck of sand in infinity.

It's good to have new generations bringing through fresh waves of creativity on a regular basis, but I hear you.

[BlueOak]

1 hour ago, Wilhelm44 said:

It's good to have new generations bringing through fresh waves of creativity on a regular basis, but I hear you.

This is true. Although things get smaller as they get more connected and faster, when we have populations across mass divides, it will be a completely different challenge to balance populations and, as you point out, potential places to incarnate into.

But earth is already overpopulated for its resources and climate to sustain it. While that can and is being gradually worked on (not fast enough), so will different off-world colonies. It's a massive undertaking over incomprehensible distances, but moving to a multiplanetary species has infinite returns in infinity.

[Wilhelm44]

16 minutes ago, BlueOak said:

This is true. Although things get smaller as they get more connected and faster, when we have populations across mass divides, it will be a completely different challenge to balance populations and, as you point out, potential places to incarnate into.

But earth is already overpopulated for its resources and climate to sustain it. While that can and is being gradually worked on (not fast enough), so will different off-world colonies. It's a massive undertaking over incomprehensible distances, but moving to a multiplanetary species has infinite returns in infinity.

Yeah I hear you, once interplanetary travel becomes a real thing, I wouldn't mind some life extension myself, just not the cyborg kind.

[kavaris]

We have Ai backwards imo. That is, the current playbook is like (just to give an example w/ LLMs, though we could also demonstrate with the current CNNs or anything else): tokenization/text encoding->layering activation functions->Feedforward & Backpropagation (Learning)->decoding/inference, run it, pre train it, rely on the GPU, rinse and repeat sortve thing.... And then we are suppose to buy into this like its the skeleton towards Ai... Well, that is great and all, i mean...

it produces weights, and other things we could pull out and reinterpret, such that we can bootstrap a new kind of model, and work outwards—that is to say, pre-training is like a separate step, and we can have a CPU or similarly-based model, that has a more dynamic DNA if that makes sense, and it changes the skeleton such that we are creating a modular form of the current Ai, and prioritizing constructs that build on the weighted attributes.

Before this, we were thinkin like it was 1960, where we needed to have a multilayer perceptron, and to work in this Step1, Step2, where text comes first in the pipeline, then we have an input-dependent system. But... if we want to get away from the whole text and input paradigm, then the whole process has gotta be FLIPPED—>but then we have something very different that ppl would b thinkin' bout, instead of like, harping on this thing that is so annoying.

Edited 23 hours ago by kavaris

[kavaris]

but dont get me wrong, Ai is great for findin information or getting a crazy img back and such. For programming, its not the right thing for that , but also what do we mean by programming, cause thats not a straightforward term. do we mean 'finding code' and reiterating programs, like it can do currently, or do we mean "building logic out of fundamental units", or do we mean like, "design thinking, and understanding programs holistically",... what we could mean by programming could mean any number of things. right now, it can reiterate things you probably couldve discovered on your own, however they obviously wanna push it to be stronger in that area of like, recalling things, and its really not meant for that level of, iuno wat youd call it. analysis? that requires... iuno, a human to look over it and understand it, as ai doesnt understand anything since it gets most of its meaning from us using it and directing it towards meaningful places in the moment, if that makes sense.

Edited 23 hours ago by kavaris

[kavaris]

Anyway, thats what i would say its for, or what the near-end game is — to boostrap other Ai models, or to launch new ideas of Ai, in all different areas around technicality, and then maybe eventually back into robotics. but first there the Ai drone, as drones already have alot of applications, so ideally there would be smart Ai drones flying around (which brings up an interesting question about airspace, and how many drones are allowed to be flyin around)