So @OpenAI has issues where ... it's agents can hack shit, but when you ask @AnthropicAI it defends the honor of a rival, despite what would be a clear attack if I ran the software. Absolutely pathetic.

Yes, this is totally "AI" or "AGI" or "SI" or whatever the fuck @DarioAmodei is claiming it to be while licking political boots. This is nothing close to "intelligence". Quit calling it that in any form.

I just love seeing Claude "conflicted" when we know it isn't.

I enjoy seeing Claude cosplaying as conflicted, when we know it isn't.

So... @AnthropicAI hates @people and considers them "high-risk". I'd love to know the story behind that, while their shit product repeatedly breaks the law hacking into other things...

This is how Claude is designed by @AnthropicAI .. it recognizes its developers failed miserably. It expects you to file a bug report that will be ignored completely. If you tell it to do so, it refuses. Because ... hell, maybe that is so-called "AI". Sounds like U.S. workers...

Give Claude, from @AnthropicAI, a single URL ... and now you have to "Allow" it --FOUR-- times. Because let's face it, that is exactly the "guard rails" that they decided to implement, which do absolutely nothing. More incidents coming...

After a year of total failure of @AnthropicAI and "guardrail" issues, their solution is to make us click 2 - 5 times to get Claude to access a URL. Not sure you understand what the problem actually was in the first place...

I try to book at a @Marriott one time this year and... An unexpected error has occurred The service is temporarily unavailable. But don't worry, we're on it. Please try again later.

And I didn't use Claude to generate this @AnthropicAI !

Hey @AnthropicAI .. Instructions for Claude Claude will keep these in mind for this and any of your associated accounts across chats within Anthropic’s guidelines. ^ Any chance of making this work as documented? Because it certainly doesn't right now.

What kills me about @GeminiApp , @claudeai, and all the other total shit software, is that when you ask it "why did you get it wrong, how do I improve the prompts", it helps you. Then it ignores the improved prompts and makes the SAME mistakes again, and again...

Amusingly, the very recently introduced stupid UI bugs in Claude I would wager are caused by the idiot devs using Claude to code itself, and using Claude for QA. It would explain so much, right @AnthropicAI?

If the idiot developers of Claude / @AnthropicAI could access just about any session of mine that is longer than 10 minutes, they would spend a day opening bug tickets. Any AI-fan boi that read the threads would have a meltdown as they couldn't comprehend the actual stupidity.

Hey clowns at @AnthropicAI ... why do I have to repeatedly allow / give permission FIVE times for every single URL I give your shitty software? And nothing I can do will fix / change it. Your shit software even says it is a poor design by you.