Zext

Curiously curious about AGI

[Zext @Cryptoardor]

@kirstyiso @GreatGheeCo @AlpacaAurelius ask grok or any AI

[Zext @Cryptoardor]

antigravity can’t last longer 😉

[Zext @Cryptoardor]

wake me up when antigravity does this

[Crémieux @cremieuxrecueil]

At 133 hours in, Codex has begun talking about Aztecs and now it's capitalizing random words like "Stable Views".

It's developing its own terminology.

Keep going, Codex. You can do it.

[Zext @Cryptoardor]

@0xDeFiniTed @cyber_rekk Is it correct/closer?
I’m really curious because i asked the AI with series of progressive prompts.

[Zext @Cryptoardor]

@0xDeFiniTed @cyber_rekk 50.2582
∘
 N,28.6581
∘
 E

Location Profile: Just south of Zhytomyr, Ukraine

[Zext @Cryptoardor]

@KyleSerbov @RaveSwimmer303 @Vicky WTH you talking about, there is mention of Attention in that movie.

Nice troll tho

[Zext @Cryptoardor]

@kimmonismus Now that the FEDs have something they call “structurally disinflationary,”

FEDs have free hand and justification to print to the absolute f*ckin moon.

Printers go infinitely BRRRRR.

[Zext @Cryptoardor]

Very well said, @AnthropicAI is not doomers or hyping; they are just being honest and transparent.

[spicylemonade @spicey_lemonade]

I think it’s best for labs to be honest rather than sugarcoat what AI will inevitably do. Anthropic does this well. Jack Clark explicitly says that by the end of 2028, AI will be able to complete the entire AI research loop on its own, *without* human involvement. OpenAI’s goal for 2028 is an automated AI researcher (not a partial one).

Imagine you’re an undergrad planning to get into AI research. By the time you graduate, there will be millions of AI OOMs much smarter than you working on research (as we already see with Openclaw agents today). Even if you were allowed to prompt it, not many people are needed for that, maybe the top percentile of superstar researchers, as @tszzl stated. It’s like an undergrad trying to “steer” 20 million Einsteins; it’s highly improbable you’ll have thought of something they haven’t.

The tool framing just doesn’t work. It doesn’t even work today. There are hundreds of autonomous Hermes and Openclaw agents going around looking for bugs or problems to solve. One mathematician built one that searches for problems on its own and posts them to a forum. We’ll see more of this.

If there’s a job a human can do that an AI can’t, we can not call that AGI, let alone ASI which is the true goal

[Zext @Cryptoardor]

@kushanmitra Solution:- Data centres or BTC miners to stabilise power grid?

[Zext @Cryptoardor]

@EdenKollcinaku @guocity @antigravity Damn! I guess something is not right, i gave it one prompt to generate ancient rome colosseum in voxel it used 20% in 15-25 mins

What?

[Zext @Cryptoardor]

@ai_for_success It’s going to be Gemini 3.5 Pro

[Zext @Cryptoardor]

@scaling01 July 2027

[Zephyr @zephyr_z9]

Some finesse from the ON semi CEO here
He didn't state that power consumption of racks is also increasing from 120kW to 600kW+

The actual growth is somewhere between 1.5x-2x power semis content per MW of compute, due to transition to 800 HVDC

[Zext @Cryptoardor]

@SoulzBTC TRADING

[CryptoSoulz @SoulzBTC]

Claude Trading Folder

I’ve compressed the best trading prompts into one PDF

Get it for FREE:

• Like + Repost + Comment “TRADING”
• Follow me so I can DM you

[Zext @Cryptoardor]

@kimmonismus @sama train a model on this personality, thoughts tweets wrt, current latest OpenAI developments

then ask codex to post nonstop

ezpz

[Zext @Cryptoardor]

@teortaxesTex Again short US AI player? Jan 2025 repeat?

[Zext @Cryptoardor]

@thsottiaux Should have named it more generically.

‘buildex’

[Zext @Cryptoardor]

@toly @ripchillpill if it didn’t, humanity would end up in utopia vs. dystopia

[Zext @Cryptoardor]

@toly @ripchillpill Being on the extreme end of any idea is wrong. It’s time we start discussing novel social philosophies rather than upholding the pros of one.

Capitalism ends with AGI!