171–180 of 222 posts
Earlier quoted context omitted.
> I doubt many support agents have access to editing user records Why do you think that's the case?
Just because agents aren't immune to prompt injection doesn't make it so that they aren't fantastically capable
It only has a ~1 in 20,000 chance of working but at scale it'll go through!
In all seriousness, shouldn't Anthropic be heavily dogfooding this sort of use case? I'm also not a huge fan of Amazon's support system, but they at least seem to be using their AI tools a lot for support responses (which has its own issues, but credit where its due). Every conference talk on this stuff seems to suggest that we're all way behind the curve on AI implementation, but I suspect its mostly smoke and mirro…
Companies that operate this way figure their customers are either so entrenched they’ll never leave or that it’s cheaper to get a new customer than expend a human’s time fixing a customer’s issue.
I hope OP either files a chargeback with their card or files a small claims court suit. Either way, they should take their money elsewhere.
But they probably won’t, because Claude is the best coding agent. Just like Google is the best search engine/free email/etc.
I had a similar experience. Pretty ironic that we can't reach a human given that "Anthropic" literally means "involving or concerning the existence of human life".
There’s nothing Meta about Meta.
There’s nothing soft about Microsoft.
Apple is a fruit, not a phone.
I'm on $20 Pro plan, I only use Claude through the web chat interface at claude.ai. I do not use Claude Code, the API, or any third-party integrations. so far for this month "$81.07 spent (Resets May 1)" just 8 days. For basic web-based conversations, accumulating $81 in overage charges within three days(April 5, April 6, April 7) is unreasonable.
Earlier quoted context omitted.
What if they built their company with poor support, so they don’t have to hold up to any standard ? But others companies have historically good reputation for good customer support, and maybe AI can help them automate easily 80% of easiest requests
Clearly they have sales and other teams as the important people within the company, with customer services being down the pecking order. They don't need AI to automate their customer service requests, they just need decent forms with a standard issue helpdesk system. It takes some work to get right, but anyone with experience of building customer support services will be able to do that, to put most of the customer s…
I'm on $20 Pro plan, I only use Claude through the web chat interface at claude.ai. I do not use Claude Code, the API, or any third-party integrations. so far for this month "$81.07 spent (Resets May 1)" just 8 days. For basic web-based conversations, accumulating $81 in overage charges within three days(April 5, April 6, April 7) is unreasonable.
Do you have extra usage enabled? Where are you finding this info?
So they're clearly playing some tricks here when they give you rebates - it turns on the overusage again.
Same, I tested Claude Code CLI and it crapped out and took my money, and they haven't responded to my billing dispute yet. Meanwhile JetBrains replies within hours and Junie is LLM agnostic. I'm a huge fan of JetBrains for AI coding.
Earlier quoted context omitted.
I don't think it's helpful to think about this as the company "trying to steal from you". There is no intention here. It's just something that got lost in a bad IT system. You gain nothing from issuing a chargeback. You imperceptibly nudge some statistic and a "banned for life" flag might automatically get flipped in a database. There's no righteous comeuppance here. You try to contact support, pester them a bit, cal…
> There is no intention here. You don’t think it’s funny how the mechanism for taking the money is never broken? Work with a large company who won’t pay your 30 or 45 day invoice for 90 days before you broadly decide this.
I dunno, sometimes it is.
The most broken I've seen in my favour was a ~$600 purchase where the order flow broke partway through. Customer support was a major pain to get in contact with in order to figure out how to give them my money. When I eventually managed to talk to someone, they advised that maybe their third-party fraud algorithms didn't like my email. I changed my email, the order worked when I placed it again, and I received my product a week later.
Several months later, without any communication from the company, I received a second product in the mail, presumably from the first order that I didn't pay for. Based on how much of a pain it was to contact support the first time, I wasn't about to do so again based on their mistake. To be charitable, I kept the package in my garage for a couple months in case the company contacted me to arrange return shipping. Not hearing from them, I just sold it off.
It took like 40 minutes and worked like this:
1. Sign in to the site, get the onboarding screen, download clients.
2. Run a client, it triggers an email with a link to open on the site so i can authenticate it.
3. Instead of authenticating the client, the site sends me to the onboarding screen again.
4-20: repeat the above loop
21. finally get a code I can paste into the client.
I'm sure someone at Anthropic posted a blog entry with how fast they vibe coded the authentication code. Back before they claimed to have vibe coded a compiler that can do a linux kernel.