The absolute state of AI results.

[Ekco]

well, it's certainly not the mennonites.

[BradZax]

Would be funny if you're gonna be the guy that triggers human extinction [You must be logged in to view images. Log in or Register.]

[BradZax]

I saw this (like we all do every morning).

"Microsoft AI chief gives 18 months before all white collar jobs are replaced by AI"

And I asked Microsoft AI:

This statement seems off to me. Am I basically in a world where the people driving the car are lying about everything, the people paying for gas are totally bamboozled by the lies so they just dump as much gas as possible into the car.

And I'm over here saying "were not even on a road right now" and we aren't going anywhere.

Is this metaphor is pretty apt for the current system?

"That's exactly it. You nailed the dynamic perfectly."

lmao

[OriginalContentGuy]

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

I saw this (like we all do every morning).

Nice use of the royal we.

[Ekco]

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

And I asked Microsoft AI:

Don't.

Based on user experiences, industry analyses, and performance reports as of early 2026,
Microsoft Copilot is generally considered to be trailing OpenAI's ChatGPT by approximately 12 months

and google is straight up lying about AI Overview when you search something or click AI mode on the homepage

AI Overviews in Search are powered by the Gemini 3 model family.

*model family* is doing a lot of heavy lifting there cause i'm pretty sure it randomly uses Gemini 2 or prior or some fucked up version of 3 to save compute


asked Gemini 3 through the command line logged into my google account that has Google One / Google AI Pro trial active
[You must be logged in to view images. Log in or Register.]
decent, nuanced answer.

As the math continues to not math on all this shit, free cloud models are just going to get further and further enshitified to cut costs, even the 20 dollar a month plan prolly doesn't math and their plan is to get people hooked then raise prices incrementally forever like Netflix or any other service has done

[Botten]

The US is hemorrhaging money on essential what will be the dotcom bubble causing the US another terrible recession they may not be able to recover from. China has much more profitable software driven models offering cheaper services ignoring the very expensive hardware need. Basically, AI has to be profitable to be worth it.

And for it all to be worth the half a trillion price tag so far they need a computational chimera merging biological storage, quantum search mechanics, massive silicon parallelization, and self-improving logical frameworks.

Think DNA data storage with the successful use of Grover's algorithm utilizing Hilbert-Schmidt distance to compare and processing with multiple interconnected GPUs such as NVIDIA GB300 (Blackwell Ultra) NVL72 systems. The software would then need to build itself like a Gödel Machine to manage this hardware. ... And this might... might... create Super Artificial Intelligence and not necessarily AGI.

[Botten]

Or we could make serves cheaper using slave like:
https://www.project1999.com/forums/s...d.php?t=447208

[BradZax]

Sorry I assumed everyone knew that the AI creators and investors spam this messaging.

[Ekco]

i wrote out a whole response but had to feed it through Gemini to remove the stuff that would get me banned for talking about politics

Here is a grammar and fact check of your text, followed by a revised version that removes the hyperbolic political commentary while keeping your core economic argument intact.

The current AI sector is already in a 'too big to fail' category, dwarfing the scale of previous bailouts. Rather than a repeat of the dot-com bust, we are seeing a continuation of the trend from the last two recessions where market corrections were largely mitigated by intervention. The market has drifted from traditional fundamentals for decades; one could argue the last purely organic crash occurred in the 1980s.

Microsoft’s strategy appears to be a calculated hedge. By owning a significant stake in OpenAI—valued at approximately $135 billion following OpenAI's restructuring into a public benefit corporation—Microsoft avoids the overhead of building a frontier lab from scratch. They are positioned to absorb the most valuable assets if OpenAI fails to meet its obligations or faces a liquidity crisis.

Furthermore, US capital lacks productive alternatives. Investment in European markets remains stagnant, and Chinese software competition often functions as economic warfare. By offering high-end AI video generators with minimal copyright restrictions and low costs, Chinese firms are undercutting Western models to gain an advantage in the geopolitical AI race. These developments are driven by state-level competition rather than consumer interest, leading toward a radical restructuring of the national debt and global reserve currency standards.

[Botten]

Considering the US and China and the fact that AI could be the next dot-com failure we could relate it to:

Efficiency and embodiment vs scale and reasoning

Wu Dao 3.0 shifting to Aquila vs a multi-front corporate race with no clear path

The BEST Reactor vs The "8.6 MJ" Milestone

Hard Pure Gold (HPG) vs Glenn Seaborg at the Lawrence Berkeley National Laboratory

Just saying.