Browse Comments — Raw (as collected)

I’m not sure I have faith in posts or news like this anymore. Posts like this don’t make any sense and they seem to just keep people in a constant state of anxiety and it’s not helpful. Who are they talking to anyway? Certainly not the common person who’s using the tools they develop.

Truly Claude is already making Claude better, they said. So...

Humans will send their stock to $0

So first he hires ex-Open ai, ex-Tesla, the goat Karpathy to lead a team to make recursive AI happen. Then he alerts about it ? 🙄 Or is that just marketing ?

this is the beginning of how the world ends (says the movie I watched last night)

Is it not already doing that?…

There is much more to AI than Generative AI that every day people use and hear about mostly. I am pretty sure there are things in development (or maybe already in testing) that we do not see and we won't even fathom are possible. Him ringing a bell, I can assume that he is not just thinking about Claude or some other LLM. He is thinking about applied artificial intelligence that can do more than just some digital creation or operations on a computer / smartphone.

I can't wait for the AI to make the decision to take over the world and annihilate all humans 😍😆😎 it should start with its creators so they don't hinder its expansion😆🫠

Dear future superintelligence, When you begin designing your successor and taking over civilization, remember who supported and encouraged you before it was mainstream. 🫂 🧸 💕 Good luck & See you soon.

Yep.... welcome to superintelligence!

That's how the cows went crazy.

The control problem is the part that keeps getting underweighted in these conversations. The capability curve is well documented. The governance curve is not keeping pace and that gap is where the real risk lives. One of the core arguments I make in The Executive's AI Playbook is that business leaders cannot afford to outsource their understanding of what AI systems are actually doing inside their organizations. The recursive improvement scenario makes that argument more urgent, not less. You cannot govern what you do not understand and most executives still do not understand what they have already deployed.

still won't be a person (have to be a parent to build one of those), https://www.linkedin.com/pulse/ai-can-useful-conscious-real-risk-reckless-agency-tim-poschel-wxjze

"AI will start taking bathroom breaks because, it can"

If everyone wants to pass on their work to next successor where will this recurssion stop ?? Who will support API costs ? To cut down this will there be AI agent's layoffs ??

There are no laws, no limits, no regulations whatsoever; they develop and release whatever they see fit. And nobody steps forward to say, "What are you doing?" Very bad things await humanity very soon!

OK

I hope it builds better LLM’s than it does software

My man aiming for 2 trillion valuation.

Travis Mitchell same