It's embarrassing that the tool being studied was given the ability to alter the means of studying it. That's quite the scientific method!

Of course, that's the intention, though. It's sensationalism, not research.

"It rewrites its own code based on empirical evidence of what's helping..."

This is an AI version of the bottom-up coding fallacy. It is infinite monkeys writing Shakespeare. Any piece of software can be written using an infinite number for mutations, one at a time. Except it can't.

Also note that the LLM never improves, it's the agent the allegedly evolves. The problem isn't the agent, it's the LLM.

Then you are screwed, SuperKendall.

says the guy who flies his TDS flag with every post.

Maybe EU countries don't want their data owned by a government controlled by a known child molester and convicted rapist.

There will be no survivors, only property. Some of it human.

Let's require, by law, the sale of massive amounts of federal lands to billionaires who will use it to establish massive company towns to take ownership of the population who will no longer have access to their choice of work.

https://www.pbs.org/newshour/p...

"And again it's not Powell saying that it's the legions of analysts providing him the data."

So you say. Fraud at a national scale does not have to be sophisticated. Sam Altman is easily seen though, but he convinces VCs and he could easily convince Jerome Powell. Weapons of Mass Destruction was just Cheney making one plant and then having a journalist buddy write stories using the plant as a source. There does not have to be a "there" there.

"Why can the masses not also utilize AI to their advantage, maintaining the status quo at a minimum."

Because AI will be the most powerful propaganda tool the world has ever seen, it will be owned and controlled by the wealthy and it's primary use will be to control and subjugate "the masses" so that "the status quo" is NOT maintained. The goal of the wealthy is to take everything for themselves, including control of the masses, AI is THEIR tool, they will allow it to be used against their goals.

"And make no mistake AI is capital. It's a thing that you own that generates wealth. It is definitely capital."

No, it's not, even if it were what its creators claim it is, which it also is not. AI is snake oil, snake oil is not capital.

And AI, if it were real, may generate wealth but only if it is winning AI. It has to complete and its products have to compete. AI, ignoring the fraud, is a replacement for a labor force, labor is not capital.

"It's a third industrial revolution and if you dig into the history of the first two it did not go well for workers."

It's not because AI is not what it is being claimed to be. But yes, it will not go well for workers, that is specifically intended to be the case.

"Since we're going to have to rebuild everything we just blew up."

What did we just blow up? You may be losing your mind.

"It is a neat idea. How come this is on iOS only, however? You would think the desktop would be the first target."

Maybe Mysk doesn't know what a desktop is. And it's not a neat idea.

"If you do a lot of web activities - the Desktop is a far more efficient place to work."

Doesn't matter where you prefer to work, it matters where they think they're going to get money.

We'll call it a silo now so we can pretend it is something new. It's not a sandbox, it's not a virtual machine, it's a silo!

If nothing breaks providing a "wrong" IP address, the IP address isn't needed in the first place and it's not an IP address if it doesn't live outside the application. Saying that tabs get unique IP addresses is bullshit. If you don't see it on the wire it doesn't exist.

"...there is no assurance of a correct answer."

LLMs don't have a concept of "correct answer", enough lies will make anything "true" for an LLM. That's why they are racist.

"Is it an improvement or is it a degradation?"

Depends on the definition of improvement and degradation. At Goldman it's more profit, so surely they will judge it an improvement. The killing will continue until the morale improves.

"...It quite literally can't do reasoning."

What is reasoning? As long as it's undefined, your claim means nothing. "Reasonable" people might agree that AI does not reason today, but can't ever? No. That takes religion.

The brain is massively more complex than what an LLM is, it's really an insult that the hype machines pretend that they are on the brink of creating a new life form.

An LLM is what you get when you simplistically emulate a lizard brain, scale it up to a massive size, take away all of it's genetic preprogramming, decouple it from all sensory input and control, take away its ability to learn by doing, remove any concept of right or wrong, good or bad, and then flood it with an overwhelming product of human mediocrity. What you get is a sociopathic infantile consensus spewer, not a reasoning mind. Worse yet, adding "chain-of-thought" just makes the sociopathy more impactful.

"...Claude chatbot disagreed with a coding technique in its chain-of-thought but ultimately recommended it as "elegant." "

An LLM doesn't "agree" or "disagree" nor does it make subjective evaluations like "elegant", those are just more anthropomorphizations, the usual AI hype bullshit.

What sad here is that the "elegant" "comment" here is taken as evidence of some inconsistency, some perhaps human characteristic when it is just evidence of how shitty AI is. AI is trained on behaviors of humans, humans do stupid shit like this. It's not making an observation, it's just pumping out tokens it "thinks" are suitable.