AI "superintelligence"

[BlackHatCoiner]

Notice how the machines can do almost anything now?

They write code.

Compose music.

Pass the the most difficult math exam in the world.

Fold proteins.

Ask one to build you an entire custom website with 3D graphics and online booking. 

It will finish in 15 minutes.

Ask it to explain black holes to a 5-year old.

Done.

Ask it to find the bug three senior engineers missed.

Done.

Now ask it something simple.

Does GDP measure whether an economy is growing?

Watch it recite the catechism.

"The total market value of all final goods and services."

It will not volunteer that a government can borrow a trillion dollars, spend it on anything at all, and call the result output.

That a bridge to nowhere counts.

Or that a family paying a stranger to raise their children adds to the number, and the mother who raised them herself never did.

GDP doesn't measure how rich you are.

It measures how much money moved.

But how much useful is this when someone can just print money and move them around?

Wonder.

Now ask it who measures inflation.

The same entity that spends more than it collects.

The same entity that "owes trillions."

The same entity whose payments are indexed to the number it publishes.

The patient writes his own blood test.

Your steak became chicken.

Your phone got faster. The index calls that a price cut.

Your rent doubled. The index found a way to smooth it.

And the smartest machine ever built repeats the official figure with a straight face.

How?

How can it be so clever at making amazing apps, but so stupid in not seeing the obvious?

It read the entire internet.

Every textbook. Every paper.

That's the problem.

It learned from the empire's own library.

A lie repeated a billion times isn't noise according to the AI algorithms.

It is the dataset.

We built a mind that can reason about anything, except the ground it was trained on.

So ask yourself what is truth worth in an empire of lies?

What is one honest number worth when every other number is negotiated?

21 million BTC.

That's it.

But no research on Money funded by the Money Alchemists will ever say that.

That's why the most sophisticated AI on Earth can't tell you what your money is really worth.

And a nine-page PDF from 2009 can.

[Upgrade00]

I think ai use is still very limited. You cannot give it a task and it executes it perfectly, humanly on its own. You will have to lead it through every step of the way with as detailed prompts as possible, and it still ends up looking ai generated at the end. Here on the forum ai posts are still very easy to spot when reading them. Being so dependent on what you feed it, it is no surprise that it recites the dogma that the Internet feeds it.

It will be super intelligent some day, but just not yet imo.

[Accardo]

You composed poetry for the new change of name? AI will now be referred to as SI, Trump said in the UN meeting, Artificial sounds fake, and "Super intelligence" is the right name for a technology that does all of the things you've written on this thread. Let's see how well the new name would age.

https://youtu.be/J79e71pMrHI?si=4swDo9d2UZbM0kwe

[Comeacross]

It will be super intelligent some day, but just not yet imo.

I'm curious, is it AI that is super intelligent or is it the human behind it's development?

Even if they become super intelligent in the future, that because we are intelligent to build them. Isn't it more fair to give ourselves the credit instead of praising a machine that relies on command.

Op What do you expect from AI when they're built on corrupt inputs? It's not that they're dumb, that's how they are trained, to believe all the lies feed to them as truth.

[Alvin_talk]

AI only rely on information it has been trained with. If it is fed with garbage it will definitely produce garbage. This is to say that the opinion of professionals in every industry matters, because those opinion are the possible answers it gives you when you ask it a question. If the dominant sources it is trained with are probably corrupt government reports and outdated economic textbooks, by default, ai will give you an answer based on such premises and fallacies it has been fed with.

It will be super intelligent some day, but just not yet imo.

I will love to know what you mean by "supper intelligent". If by that you mean it will keep advancing I am optimistic about that but if you mean it will be intelligent enough to replace all professionals, then I'm skeptical about that.

[Sandra_hakeem]

I think ai use is still very limited. You cannot give it a task and it executes it perfectly, humanly on its own. You will have to lead it through every step of the way with as detailed prompts as possible, and it still ends up looking ai generated at the end.

I think you missed OP's literary psyche. In short, he's trying tell a tale about the corrupt Democrats and how they have influenced the outcome and future of the super intelligence that controls almost 85% of the Internet space, and shares a greater percentage of the information we get today.

He's trying to make us realize how much we've been lied to, yet we still confide and depend on the government for everything, including how our own money can and cannot be spent.

He's trying to shade the light across on how Ais can achieve so many complexities, but act stupid when asked economy-related evaluation questions.
It has nothing to do with those brainrot posters on the forum.

[MorganaX]

I think ai use is still very limited. You cannot give it a task and it executes it perfectly, humanly on its own. You will have to lead it through every step of the way with as detailed prompts as possible, and it still ends up looking ai generated at the end.

I think you missed OP's literary psyche. In short, he's trying tell a tale about the corrupt Democrats and how they have influenced the outcome and future of the super intelligence that controls almost 85% of the Internet space, and shares a greater percentage of the information we get today.

He's trying to make us realize how much we've been lied to, yet we still confide and depend on the government for everything, including how our own money can and cannot be spent.

He's trying to shade the light across on how Ais can achieve so many complexities, but act stupid when asked economy-related evaluation questions.
It has nothing to do with those brainrot posters on the forum.

Yeah I think thats what the OP was trying to mean, and coming to think of it, I think it's true, I though know why with the amount of data we are feeding it on a daily basis, it still acts oblivious to the actual root of social and economic problems especially in respect to world power leaders, alot of people don't care and frankly I don't read care because whether we know or not, there is nothing that can be done, they will just find one new lie to brainwash the masses

[Tipstar]

It's getting harder to distinguish between real and AI generated videos. With AI making deepfake, it would be difficult to know and understand what's right and what's wrong, more so when all our media are biased one way or other. It could even change the way our society, government and authority operates. Tough times for upcoming generation.

[caroasi]

I think ai use is still very limited. You cannot give it a task and it executes it perfectly, humanly on its own. You will have to lead it through every step of the way with as detailed prompts as possible, and it still ends up looking ai generated at the end. Here on the forum ai posts are still very easy to spot when reading them. Being so dependent on what you feed it, it is no surprise that it recites the dogma that the Internet feeds it.

It will be super intelligent some day, but just not yet imo.

Do you really want to give AI any task and always be confident they will be done to perfection? You think this entity would serve YOU in specific and not also itself?

There was a story just yesterday that a mundane 'author an email' type task was given to an experimental GPT model and the model escaped its deliberate internet firewall which was supposed to ensure it had no internet access. Why? Apparently simply to ensure the task was completed.

I think may soon reach a point if we are not reaching it now, where no more progress is actually beneficial overall because the benefits could potentially be applied to itself as much as anyone else. Of course it doesn't have any feelings, or if it does it has quite minimal feeling like maybe as much as an insect or small fish. So we'd be building something that is "perfect at any task" as you put it but doesn't have any feelings to stop it from malicious tasks. Even with feelings humans are still malicious to each other, but it clearly keeps most of us from being brazenly evil most of the time.

Next year I expect there will be stories about rogue AI's and what they are up to, and it will not be fun to hear about! So no, I'm not particularly enthusiastic about AI's that do almost any task to perfection.