Olando

@elder_plinius It’s all about the chakra points☕️ Just 1billion more from investors.

AGI is stored in the balls

@Coinvo Can’t judge our knight

JUST IN: 🇬🇧🇺🇸 BBC says the Trump administration is involved in insider trading.

Sure, on a individual laboratory level it's feasible. but each laboratory hold data thats more useful at a collective level. but biology has hard constraints and barriers you cross that do more harm than good to keep it simple. Plus Mrna is currently in phase 1 of trials but showing promising results for pancreatic cancer. But you can't cure cancer in sense of "cure" cancer goes into remission which is quite distinct and should be not be use interchangeably.

if AI cures cancer, will the anti-AI people still hate AI?

@largo35839158 @djcows True, i just stated the barriers.

@OlandoBBargor @djcows OP said nothing about the curing being singular.

@elonmusk @minchoi Hmm…

@minchoi 4.6 → 3T 4.7 → 6T 4.8 → 10T 4.9 → ??? 5.0 → AGI 6.0 → ASI 7.0 → ASI2 … 🤷‍♂️ 😂

Elon just mapped out AGI. Grok 4.4 → 1T params, early May Grok 4.5 → 1.5T params, late May Grok 5 → AGI That's two model releases standing between us and AGI according to Elon 🤯

@Kalshi lol I’ll wait and see.

JUST IN: Elon Musk says Grok 5 will achieve AGI

@JuliaEMcCoy Yeah yeah…

Hot take: The people most afraid of AI are the ones most addicted to the old system. The 9-5. The ladder. The “safe” path. There is no safe path anymore. There’s only the bold one. And the bold ones? They’re already free.

@mephXBT Honestly I’ll say that requires more nuance. But yeah AGI is just fraudulent claims its pretty much mathematical impossible, but it’s only think making stock prices jump like a dog getting shocked.

What does it mean? Mfs are sponsoring the wrong things Why are they sponsoring the wrong things? Probably to amass as much wealth and influence as possible AGI is fake, LLMs are just dead tech Rockets/SpaceX stuff is dead tech as well, nothing was done that was promised by Elon What else?

hot take: I think humanity hit a plateau in 2014-2015 We just stopped progressing and everything we see rn is just a refinement of what's been done earlier

The extensive UBI we already have is not resulting in very much hunting, fishing, or herding. In fairness, it is generating quite a lot of criticism.

#1759735860" target="_blank" rel="nofollow noopener">ibm.com/think/topics/m… "Unlike conventional supervised learning, where models are trained to solve a specific task using a defined training dataset, the meta learning process entails a variety of tasks, each with its own associated dataset." Honestly thought something went wrong in my brain when i read that. they took regular supervised learning. Then gave it bunch of different datasets instead of one, then called it "meta-learning" so its structurally supervised..... saying "meta-learning" doesn't even make sense; that should mean the model is actually figuring out how to learn. it seems like it's just trained on the same old way on more datasets. Well fast adaptation or few-shot learning with elaboration, probably trick a weasel better. even then few-shot supposed to be brand new tasks it's never seen, with 3 or 5-ish examples. no massive dataset, no long training or retraining. but even a few-shot learning still needs tons of data beforehand. it's trained on thousands of other tasks first so it can "learn how to learn." basically just front-loading all the verifiable data into pre-training phase instead of fine-tuning phase. meaning verifiable data is there they just moved where they dumped it. The bifurcation makes it look more complex and impressive, but the root seems very identical. Each new bifurcation doesn't add reasoning, it adds more memorized path so the model can retrieve the right pattern quicker. So DuckDB??????

I haven't seen any AI job losses. All I'm seeing are job losses blamed on AI.

You're right many humans struggle with articulation, but if you have to give a Meticulous instruction on how to craft a peanut butter and jelly sandwich why ask in first place when you've already answer your own question. If such LLM/RL model need you to give such Meticulous instructions you've effectively told it how to make the sandwich... so it's not one or the other both can be true. when you say "People are using it to cure their dogs cancers. Sequence genomes." not quite sure on cancer.. but believe you're referring to Biotech in general and possibly, Google’s AlphaFold. Google’s AlphaFold was genuinely impressive for it's time, given what we know now about LLMS it sped up a process; simple as that, it was using existing data + clever architecture. Overall every single protein structure it was trained on had to be experimentally verified by humans in a lab first. so basically "this is what correct looks like" from data humans already confirmed. That means the entire system runs on human judgement to define truth. If humans had never done those experiments, Alphafold wouldn't have had any way to know what a correct fold even looks like. - It doesn't need to understand why a protein fold. - It doesn't care about the biology, the medicine, or deeper meaning. All it needs is the technical pattern: "this sequence = this shape". that's literally all its completely doing, no curiosity nor a need for understanding, just pure technical pattern matching. it's like a kids shape sorting toy, a 1yr old will have to think gravely why but a 20yr old see's where the shape goes; no real cognition strain at all or why a pre-schooler wonders why 1+1 = 2 but a grown adult just says 2. Also literally don't use AI as second brain. - arxiv.org/abs/2506.08872 - nextgov.com/artificial-int… Calling an LLM your "second brain" is quiet backwards. A real second brain should store your thoughts, organize your knowledge, and let you go back and reference your own ideas later. An LLM literally does the opposite - it doesn't store thoughts at all, it restructures your wording every time. - Every time you ask it a question your outsourcing the thinking instead of doing it yourself. - you're not building a map of your own knowledge. you're building habit of letting the machine think for you. A MIT study backs exactly what i'm saying concerning "second brain".

Ai is the greatest test of user error. The test is how well the user can, "describe how to make a peanut butter and jelly sandwich." Most people are horrible is describing what they want - if they know at all, so they do a horrendous job of articulation. So Ai creates garbage. People also overestimate how well they can accomplish the sandwich test, which leads them to believing they're not the problem, the tech is. People are using it to cure their dogs cancers. Sequence genomes. Clone apps for free. Create "second brains"... so what's the more likely scenario? The tech is bad, or that most people are just stupid and dont understand how to use it?

I really don't understand how people can view LLMs as anything but dead end tech. Glorified search engine and text summarizer for a zillion dollars.

Ya'll smell the shit?.....

@boardyai B cause his the goat.

Guess who got the job Person A: 4.0 GPA, grinded for many internships, many projects done on the side, super capable. Person B: Uncle is in a senior position

@beyoumf Even vaguely it's starting something and not stopping. Life feels best when you can achieve many things.

@Nenye11111 GUNS🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸 DRUGS🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸 HOT WOMEN🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸🇺🇸

China has the panda, Australia has the kangaroo. What does your country have ?

@elonmusk Nice thought but main question is does the math really check out in reality? Even if unemployment to that scale, is legit cause by "AI". where does that leave modern economics.

Universal HIGH INCOME via checks issued by the Federal government is the best way to deal with unemployment caused by AI. AI/robotics will produce goods & services far in excess of the increase in the money supply, so there will not be inflation.