The absolute state of AI results.

[BradZax]

Originally Posted by OriginalContentGuy [You must be logged in to view images. Log in or Register.]

Came across this paper touching on what I think that we discussed fifteen pages ago.

The breakdown of why the top 10% of global consumers (which includes anyone making over roughly $65k/year globally) drive a vastly disproportionate share of global ecological damage compared to developing nations.

1. Outsourcing the "Dirty" Work (Embedded Footprints)

Bu-bu-but muh tariffs r bad!!!!

[OriginalContentGuy]

Don't make me go Daniel Plainview on you, Eli Brad.

[BradZax]

Originally Posted by OriginalContentGuy [You must be logged in to view images. Log in or Register.]

Don't make me go Daniel Plainview on you, Eli Brad.

Hey you and I agree, manufacturing in 3rd world countries to get around paying livable wages and offloading those costs into excess carbon emissions is bad.

I just don't refuse to do anything about it because I dont like my dad.

And that 65k earners (teachers) are the problem lol

The "Base" of the 10% (The 9%): This is made up of roughly 500 million people—teachers, nurses, software engineers, and managers living in the US, Europe, and developed parts of Asia. Their environmental footprint comes from standard modern lifestyles (driving a car, running air conditioning, eating a Western diet, and buying consumer goods).

LOL

[Reiwa]

I would also like to enslave a nation or two.

[BradZax]

We are not so divine that the laws of thermo dynamics wont affect us.

The more, comfortable people that there are.

The more fuel you need to feed them.

The more fuel you need to power them.

The less comfortable people there are.

[BradZax]

This is amazing. Check out this video. It's a great video alone, but at 17:31 you can see the best part.

This will validate all the AI haters so much, so enjoy.

[You must be logged in to view images. Log in or Register.]

This screenshot shows a snip from the part of this google document about the future progress of AI.

The section lists the top 5 major things that could slow AI development, the causes, and potential solutions that could stop those things from slowing AI development.

The guy making the video doesn't stop to notice this section, but I did lol

I love how it just straight up says solutions to society deliberately slowing down ai development are:

1 fuck up the economy
2 push racial division in politics
3 exploit ww3

lmao now lets all argue about which extreme we should support instead of coming to an actual solution.

So I guess this what taking "don't be evil" off the marque does [You must be logged in to view images. Log in or Register.]

[Ekco]

Only half the planned data centers have started this year and our energy infrastructure is dogshit, Elon has mobile gas turbines running in the parking lot of the colossus data center because the Memphis grid can't supply it, we're currently compute restrained with only 10 million ai power users in the country

When evaluating the market through this strict lens, the estimate of 10 to 12 million true, deeply integrated AI power users in the United States is incredibly accurate.

China has 60 nuclear reactors with 40 under construction with a population pool of non retards orders of magnitude larger to pull from, if AGI is brute forceable by just scaling they've already won

Lucky for us AGI and certainly ASI by any of the like 100+ definitions out there is some sci-fi bullshit that transformers doing linear algebra matrix multiplications isn't going to achieve

[BradZax]

Originally Posted by Ekco [You must be logged in to view images. Log in or Register.]

Lucky for us AGI and certainly ASI by any of the like 100+ definitions out there is some sci-fi bullshit that transformers doing linear algebra matrix multiplications isn't going to achieve

not according to that google doc.

I also like in that google doc that it debunks the myth that ai progress has flattened and it shows that actually, it has increased.

[You must be logged in to view images. Log in or Register.]

[Ekco]

It's already no longer exponential and plateauing like moore's law with CPUs and other hardware, that's literally what your graph shows of them all bunchin up together at the top of the curve there, 4.6, 4.7 fable/mythos chatgpt 5.5 the new China one are all still on the same tier the difference between their capabilities and the previous ones are a fraction compared to the previous ones and another tier down

They've already shifted strategy from one giant parameter model to a bunch of smaller MOE mixture of expert models in a trench coat because they literally did hit a ceiling with the previous model training method

[BradZax]

Originally Posted by Ekco [You must be logged in to view images. Log in or Register.]

It's already no longer exponential and plateauing like moore's law with CPUs and other hardware, that's literally what your graph shows of them all bunchin up together at the top of the curve there, 4.6, 4.7 fable/mythos chatgpt 5.5 the new China one are all still on the same tier the difference between their capabilities and the previous ones are a fraction compared to the previous ones and another tier down

They've already shifted strategy from one giant parameter model to a bunch of smaller MOE mixture of expert models in a trench coat because they literally did hit a ceiling with the previous model training method

they are not bunching up, what you're seeing is there are more companies developing more AI, the progress over time is doing exponential growth.

Power consumption aside, that curve will solve that problem on its own when it needs to.

[You must be logged in to view images. Log in or Register.]