Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Sonnet 4.6 Elevated Rate of Errors

status.claude.com

Post reply on HN
Print view View original

81–90 of 91 posts

Re: Sonnet 4.6 Elevated Rate of Errors

#81

Post by brador »

ManlyBread wrote: post #37
gambiting wrote: post #29

Earlier quoted context omitted.

I'm on the basic £18/month plan and with Sonnet 4.6 I literally get 20 maybe 30 minutes of use out of it per day. It's borderline useless now. I was using it for some Home Assistant changes yesterday and it used up my entire daily allowance after 8 prompts.

I guess 2026 is the last year AI is widely available to anyone who isn't willing to shell out hundreds if not thousands for a monthly subscription. I guess all that's left is to thank all the investors for the free ride LOL

For a second there we felt the future. And now it’s gone.

Re: Sonnet 4.6 Elevated Rate of Errors

#82

Post by joshdev »

ApolloRising wrote: post #21

As long as everyone is here, have you seen the token usage just go up remarkably recently for the $100 plan? it lasts a lot less time than it used to recently. Might be related to recent releases of claude.

We discovered a bug in AWS Bedrock that is double counting cache writes when thinking/reasoning is enabled for the Anthropic models. It’s not clear to me if this is limited to just AWS Bedrock or all providers. AWS Support is aware.

We’ve also observed a much higher cache miss rate in the past few weeks. Combine both together and your usage consumption can be greatly increased.

Re: Sonnet 4.6 Elevated Rate of Errors

#83

Post by stingraycharles »

dijit wrote: post #28
stingraycharles wrote: post #22

Earlier quoted context omitted.

> They’re asking for $100+/mo for the plans that are actually usable at scale. If I’m paying that much I have very high expectations. If you think $100 is that much and get very high expectations from it, you're not the target customer. You're a loss leader to Anthropic, and the fact that you don't see that / still have high expectations means your expectations are unrealistic.

$100/m for SaaS is very steep. For an entire productivity suite including mail, meetings and terabytes of backed up redundant storage with nearly no bandwidth limitations it's like $35/m for even the most expensive option.

Comparing state of the art LLMs with Office365 / Gsuite is like comparing renting a datacenter vs an airbnb. Entirely different things.

Compare it to hosting models locally, that would be more apt. Or renting GPUs from cloud vendors.

Re: Sonnet 4.6 Elevated Rate of Errors

#84

Post by dijit »

stingraycharles wrote: post #83
dijit wrote: post #28

Earlier quoted context omitted.

$100/m for SaaS is very steep. For an entire productivity suite including mail, meetings and terabytes of backed up redundant storage with nearly no bandwidth limitations it's like $35/m for even the most expensive option.

Comparing state of the art LLMs with Office365 / Gsuite is like comparing renting a datacenter vs an airbnb. Entirely different things. Compare it to hosting models locally, that would be more apt. Or renting GPUs from cloud vendors.

It’s a SaaS, and the most expensive SaaS available.

If you’re saying an LLM provides more value than the office productivity suites , mail platforms and meeting platforms which run essentially the planet: then I am afraid, you have drunk the kool-aid.

If you’re evaluating software licenses you have to weigh the price to value, there can be value to these LLMs but its not 3x the productivity of Mail+Spreadsheets+Live Meetings+presentations+wordprocessing+filesharing.

Its just not.

Re: Sonnet 4.6 Elevated Rate of Errors

#85

Post by stingraycharles »

dijit wrote: post #84
stingraycharles wrote: post #83

Earlier quoted context omitted.

Comparing state of the art LLMs with Office365 / Gsuite is like comparing renting a datacenter vs an airbnb. Entirely different things. Compare it to hosting models locally, that would be more apt. Or renting GPUs from cloud vendors.

It’s a SaaS, and the most expensive SaaS available. If you’re saying an LLM provides more value than the office productivity suites , mail platforms and meeting platforms which run essentially the planet: then I am afraid, you have drunk the kool-aid. If you’re evaluating software licenses you have to weigh the price to value, there can be value to these LLMs but its not 3x the productivity of Mail+Spreadsheets+Live…

If you really think the value add by LLMs is comparable to email and calendar, I don’t understand why you don’t understand my point that you’re not the customer Anthropic cares about.

Re: Sonnet 4.6 Elevated Rate of Errors

#86

Post by dijit »

stingraycharles wrote: post #85
dijit wrote: post #84

Earlier quoted context omitted.

It’s a SaaS, and the most expensive SaaS available. If you’re saying an LLM provides more value than the office productivity suites , mail platforms and meeting platforms which run essentially the planet: then I am afraid, you have drunk the kool-aid. If you’re evaluating software licenses you have to weigh the price to value, there can be value to these LLMs but its not 3x the productivity of Mail+Spreadsheets+Live…

If you really think the value add by LLMs is comparable to email and calendar, I don’t understand why you don’t understand my point that you’re not the customer Anthropic cares about.

Your point is stupid.

The cost of offering a service and the cost of buying a service are correlated but not the customers problem.

If you are the most expensive SaaS on a docket sheet and you’re also the least reliable you had better be delivering some serious value in the times you’re up otherwise customers won’t depend on you and you’ll be the first one out.

Nobody wants to pay premium prices for things they can’t depend on. If you cant understand that then you need to stop offloading critical thinking to your AI tools because your mind needs the exercise.

Re: Sonnet 4.6 Elevated Rate of Errors

#87

Post by al_borland »

runeblaze wrote: post #57
al_borland wrote: post #39

Earlier quoted context omitted.

When the narrative around AI is that people should rely on it all the time, people will be judged by your token use (it better be high), the AI is smarter than everyone and will take all the jobs, the AI is the best programmer, and more… When things fail repeatedly, it highlights that the emperor has no clothes. If it’s as good as they say, why can’t it figure out how to not go down every day? How can people rely on…

I mean I used to work on model reliability with my little PhD degree and the models i manage go down all the time. Some profs have a team of PhDs and things go to shit all the time. I don’t know why we expect $FRONTIER_LLM to do better

We accept human errors, limitations, and failures. We can empathize with team of humans doing the best they can, and we know any failure is a chance for them to learn and grow.

The sales pitch of AI is that it’s better than humans and has no real limits; it will make us all obsolete. This framing they created means I expect it not to make errors, not to have limits, and not to fail. I expect it to be able to learn and adapt at the speed of light and solve complex problems beyond what a PhD could do. This is what we’ve been told with the narratives around future jobs, AI performance on PhD level tests, how coding is a solved problem, and pictures painted of what a future with AI will look like. While we may know this isn’t true, this is what they are selling, and that’s the standard I’m going to hold them to.

I don’t blame the customer for being upset the snake oil didn’t live up to its promises, I blame the snake oil salesman. We have every right to be upset with the snake oil salesman and ridicule him when his product doesn’t work. Maybe we don’t need better more reliable snake oil, maybe we need real medicine. If real medicine don’t exist, its better to be honest than to mislead people and say it does.

This isn’t to say AI is completely useless, but it’s not what’s being sold. The downtime just proves that, unless they aren’t using their own product. If that’s the case, why not?

Re: Sonnet 4.6 Elevated Rate of Errors

#88

Post by mamcx »

ApolloRising wrote: post #21

As long as everyone is here, have you seen the token usage just go up remarkably recently for the $100 plan? it lasts a lot less time than it used to recently. Might be related to recent releases of claude.

Then is better to use copilot to get access to this? I have wondered if go with Claude directly or copilot...

Re: Sonnet 4.6 Elevated Rate of Errors

#89

Post by arcfour »

baq wrote: post #59
arcfour wrote: post #42

Earlier quoted context omitted.

You can't stop accepting new customers unless you're fine with killing your potential future customer base. That's a ridiculous suggestion. Either your current customers or your potential future customers are going to be unhappy so long as compute resources are finite. Take your pick.

> You can't stop accepting new customers unless you're fine with killing your potential future customer base. That's a ridiculous suggestion. what? they already have, they aren't releasing mythos except to a limited pre-approved customer base who is practically begging them to take their money. they can do that for lower tier models and at this point they should .

Their rationale provided for that is safety-based, not infrastructure-based.

Re: Sonnet 4.6 Elevated Rate of Errors

#90

Post by billynomates »

oefrha wrote: post #46
billynomates wrote: post #32

Earlier quoted context omitted.

No, in fact I'm growing increasingly suspicious of messages I see like this all over the socials. I am using Claude constantly, multiple agents, around 8-10hrs a day, 5 or 6 days a week, and I'm never anywhere need my limit.

Unless you’re somehow on a different quota system, or maybe using Haiku, there’s no way you can sustain five continuous hours of parallel agents running without hitting the 5h quota limit, even on the 20x max plan. But maybe your company is flagged as VIP or something.

OK not constantly using multiple agents, but very frequently.
Post reply on HN