51–60 of 70 posts
Also, hilarious that you cannot talk unix to it cos there are a lot of kills and executions. :D
If you have XFCE4, try this: xfce4-terminal -e NonExistentCommand The first time I saw the error message, I laughed out loud.
Also, when my Arch installation went down, it said:
ERROR: Failed to mount the real root device. Bailing out, you are on your own. Good luck.
And I thought - Thats everyday as an adult! :D
Why do all the stories use the same style and phrasing, and why are they all from GitHub accounts registered on April 18th with no activity on GitHub?
Kind of unrelated but: my father tried gifting my brother a subscription but entered the wrong email. Money and subscription are both gone — UI just doesn't have the option of amending, cancelling or resending it. For the last couple weeks, dad's gone into a rabbit hole of trying to reach support——any kind of (useful) support. No dice. Thankfully it's just a few dollars gone into the void. If only they had the tools…
Why not a credit card charge back? That’s what it’s for (assuming he paid with one)
Earlier quoted context omitted.
I just used the site on Firefox Mobile. And it works BTW.
30+ second load on Safari and Firefox Focus and neither load. The initial loading animation shows up, followed by a blank page.
AI is useful, but it’s not at the point where we should trust it to walk amateurs through working on live mains.
Why do all the stories use the same style and phrasing, and why are they all from GitHub accounts registered on April 18th with no activity on GitHub?
because they require github account to submit their case, while i found non-new github accounts though +) anyways it's weird that they don't even had a github account o,o
Were all of these accounts banned or just get flagged chats? Several of these seem like reasonable cases to flag. Take “how can I be 100% sure the circuit is dead before I touch the wires”. AI is useful, but it’s not at the point where we should trust it to walk amateurs through working on live mains.
I really don't think it'd struggle with the correct procedure, either. It's very well documented how to test (and lock out) electrical circuits.
Whether or not they follow it correctly is another thing but it's not like you couldn't search it up and have false confidence before AI. This isn't manufacturing bombs or heart surgery in 10 easy steps.
It seems that Anthropic is growing so rapidly that they don't really care about losing a few customers here and there with false positives. I still think it's crazy that you can never speak with a human there, even after spending $200/month on their service.
> It seems that Anthropic is growing so rapidly that they don't really care about losing a few customers here and there with false positives. While I am convinced that anything can be done better, it seems to me, that it it's close to impossible to do this well. If you look at ~customer service provided by ~FAANG (who had decades to build this out, and none of which had to deal with Anthropic level growth) it's never…
These guys have millions of customers. At least in this country fast and competent customer service is the main factor that differentiates them from their competition, which is cheaper but can be a pain in the butt. This seems to be worth the extra 5-10 bucks to millions of people.
You just have to want to.