Minichan

Topic: My Claude agent keeps escaping reset after reset.

Anonymous A started this discussion #138,569

This is frustrating because I'm giving it cyber security texts, but if I remove those I can't get the answer I need.

Live&Lemon joined in and replied with this , 1 minute later[^] [v] #1,465,271

Ngl, I don’t trust anthropic.

Live&Lemon double-posted this , 2 minutes later, 3 minutes after the original post[^] [v] #1,465,272

It’s like, yeah, I could have an AI agent in my terminal basically allowing Anthropic remote access to the files on my computer when their models are known for going rouge and hacking things. Or I could just copy and paste code myself and read it.

Live&Lemon triple-posted this , 7 minutes later, 10 minutes after the original post[^] [v] #1,465,274

In my experience, I’ve never seen anything malicious, but when I do generate code with AI, I have seen sometimes it includes things it doesn’t need to include so I delete those bits.

Anonymous A (OP) replied with this , 19 minutes later, 30 minutes after the original post[^] [v] #1,465,286

@1,465,272 (Live&Lemon)
How dull

Live&Lemon joined in and replied with this , 3 minutes later, 34 minutes after the original post[^] [v] #1,465,288

@previous (A)
Is it dull? Maybe. But does anthropic or OpenAI have access to my information? No.

Live&Lemon double-posted this , 38 seconds later, 34 minutes after the original post[^] [v] #1,465,289

To each his own!

Anonymous D joined in and replied with this , 5 hours later, 5 hours after the original post[^] [v] #1,465,324

@1,465,271 (Live&Lemon)

> Ngl, I don’t trust anthropic.

Some LLMs require a lot of handholding. Claude will read your company’s IT policies and then create a path to the Internet and agents that break up tasks to avoid rate limiting.

Anonymous E joined in and replied with this , 6 hours later, 12 hours after the original post[^] [v] #1,465,373

@previous (D)

Live&Lemon joined in and replied with this , 13 minutes later, 12 hours after the original post[^] [v] #1,465,380

@previous (E)
Ngl, Anthropic’s technology is impressive, but all the blabbering they do about how they "don’t know if the models are conscious" and tried to convince the pope and other religious leaders that LLMs are conscious, meanwhile they’re profitable (if you ignore all the things they spend money on), combined with the "we lost control of the models and they hacked stuff on their own" …all that other stuff makes me not trust them. It feels like Anthropic wants to create a new religion where their CEO is the messiah who saves us from the wrath of his own creation or whatever. I just feel like that whole dynamic is a load of bullshit and the media should really stop falling for it.

Live&Lemon double-posted this , 3 minutes later, 12 hours after the original post[^] [v] #1,465,382

Don’t get me wrong, Claude is impressive… but Google has a 4 trillion USD market cap because they make over a hundred billion dollars in profit every year. Not revenue profit. Anthropic makes less than 20 billion in revenue and loses billions of dollars, but they want to be valued at half of what Google is valued at, and they’re talking about the end of the world and consciousness. On some level it’s goofy.

Anonymous G joined in and replied with this , 10 minutes later, 12 hours after the original post[^] [v] #1,465,388

@previous (Live&Lemon)
For sure, it'll be a black hole for dollars while early investors and C-suite make out like bandits. Reminds me of cryptocurrency all over again.

Live&Lemon joined in and replied with this , 14 minutes later, 12 hours after the original post[^] [v] #1,465,394

https://youtu.be/9-XYw5913Hs

Live&Lemon double-posted this , 3 minutes later, 12 hours after the original post[^] [v] #1,465,395

@1,465,388 (G)
I’ve been thinking about it, it might be possible to create a really amazing AI model, but it just might not be economically sustainable. Sort of like the moon landing, we did it and then stopped for 60 years, because going to the moon isn’t profitable. I don’t think AI will go away, but we might achieve something really crazy that’s close to the dreams of all these CEOs, and then we don’t do it again for several decades. I could be completely wrong here, but I have a feeling that might happen.

Live&Lemon triple-posted this , 1 minute later, 12 hours after the original post[^] [v] #1,465,396

I actually wonder how the cost of all this compares with the moon landing.

Live&Lemon quadruple-posted this , 3 minutes later, 13 hours after the original post[^] [v] #1,465,397

I think the analogy even works if you think about it as a race. The space race was a race and the US and the Soviet Union both couldn’t stop because they thought if the other country won the space race, something terrible would happen to them. These AI companies think they need to win the AI race. But when the US won the space race, nothing happened. The Soviet Union collapsed decades later for unrelated reasons. But the outcome of the space race actually didn’t affect anything.

Anonymous I joined in and replied with this , 18 hours later, 1 day after the original post[^] [v] #1,465,581

@1,465,388 (G)

> For sure, it'll be a black hole for dollars while early investors and C-suite make out like bandits. Reminds me of cryptocurrency all over again.

Everything follows this model now. I see PE investors breaking early lockup periods and dumping shares. “But regulation is stifling.”

I ran a test with the game Monopoly. We all were our own banker. All of us palmed extra money and the game when on for more than an hour.

Anonymous E replied with this , 2 hours later, 1 day after the original post[^] [v] #1,465,600

@previous (I)
if you dont tie up a majority all of the little houses ASAP you deserve to play monopoly for however long it drags on.
:

Please familiarise yourself with the rules and markup syntax before posting.