351–360 of 664 posts
I am really confused as to what happened here. The use of ‘disabled organization’ to refer to the author made it extra confusing. I think I kind of have an idea what the author was doing, but not really.
Yeah, referring to yourself once as a "disabled organisation" is a good bit, referencing anthropics silly terminology. Keeping it for the duration made this a very hard follow
Earlier quoted context omitted.
The Pro plan quota seems to be getting worse. I can get maybe 20-30 minutes work done before I hit my 4 hour quota. I found myself using it more just for the planning phase to get a little bit more time out of it, but yesterday I managed to ask it ONE question in plan mode (from a fresh quota window), and while it was thinking it ran out of quota. I'm assuming it probably pulled in a ton of references from my project…
Self-hosted might be the way to go soon. I'm getting 2x Olares One boxes, each with an RTX 5090 GPU (NVIDIA 24GB VRAM), and a built-in ecosystem of AI apps, many of which should be useful, and Kubernetes + Docker will let me deploy whatever else I want. Presumably I will manage to host a good coding model and use Claude Code as the framework (or some other). There will be many good options out there soon.
The best you can get today with consumer hardware is something like devstral2-small(24B) or qwen-coder30b(underwhelming) or glm-4.7-flash (promising but buggy atm). And you'll still need beefy workstations ~5-10k.
If you want open-SotA you have to get hardware worth 80-100k to run the big boys (dsv3.2, glm4.7, minimax2.1, devstral2-123b, etc). It's ok for small office setups, but out of range for most local deployments (esp considering that the workstations need lots of power if you go 8x GPUs, even with something like 8x 6000pro @ 300w).
I've been doing something a lot like this, using a claude-desktop instance attached to my personal mcp server to spawn claude-code worker nodes for things, and for a month or two now it's been working great using the main desktop chat as a project manager of sorts. I even started paying for MAX plan as I've been using it effectively to write software now (I am NOT a developer). Lately it's gotten entirely flaky, wher…
> Lately it's gotten entirely flaky, where chat's will just stop working This happens to me more often than not both in the Claude Desktop and in web. It seems that longer the conversation goes the more likely it is to happen. Frustrating.
Earlier quoted context omitted.
Anecdotally but it definitely feels like in the last couple weeks CC tends to be more aggressive at pulling in significantly larger chunks of an existing code base - even for some simple queries I'll see it easily ramp up to 50-60k token usage.
This really speaks to the need to separate the LLM you use and the coding tool that uses it. LLM makers utilizing the SaaS model make money on the tokens you spend whether or not they need them. Tools like aider and opencode (each in their own way) use separate tools build a map of the codebase that they can use to work with code using fewer tokens. When I see posts like this I start to understand why Anthropic now b…
OpenCode is incentivized to make a good product that uses your token budget efficiently since it allows you to seamlessly switch between different models.
Anthropic as a model provider on the other hand, is incentivized to exhaust your token budget to keep you hooked. You'll be forced to wait when your usage limits are reached, or pay up for a higher plan if you can't wait to get your fix.
CC, specifically Opus 4.5, is an incredible tool, but Anthropic is handling its distribution the way a drug dealer would.
Earlier quoted context omitted.
Anecdotally but it definitely feels like in the last couple weeks CC tends to be more aggressive at pulling in significantly larger chunks of an existing code base - even for some simple queries I'll see it easily ramp up to 50-60k token usage.
I'm curious if anyone has logged the number of thinking tokens over time. My implication was the "thinking/reasoning" modes are a way for LLM providers to put their thumb on the scale for how much the service costs. they get to see (if not opted-out) your context, idea, source code , etc. and in return you give them $220 and they give you back "out of tokens"
I've done RL training on small local models, and there's a strong correlation between length of response and accuracy. The more they churn tokens, the better the end result gets.
I actually think that the hyper-scalers would prefer to serve shorter answers. A token generated at 1k ctx length is cheaper to serve than one at 10k context, and way way cheaper than one at 100k context.
Earlier quoted context omitted.
This whole API vs plan looks weird to me. Why not force everyone to use API? You pay for what you use, it's very simple. API should be the most honest way to monetize, right? This fixed subscription plan with some hardly specified quotas looks like they want to extract extra money from these users who pay $200 and don't use that value, at the same time preventing other users from going over $200. Like I understand th…
Consumers like predictable billing more than they care about getting the most bang for their buck and beancounters like sticky recurring revenue streams more than they care about maximizing the profit margins for every user.
I am really confused as to what happened here. The use of ‘disabled organization’ to refer to the author made it extra confusing. I think I kind of have an idea what the author was doing, but not really.
Earlier quoted context omitted.
Years ago I was involved in a service where we some times had to disable accounts for abusive behavior. I'm talking about obvious abusive behavior, akin to griefing other users. Every once in while someone would take it personally and go on a social media rampage. The one thing I learned from being on the other side of this is that if someone seems like an unreliable narrator, they probably are. They know the company…
"I'm glad this happened with Anthropic instead of Google, which provides Gemini, email, etc. or I would have been locked out of the actually important non-AI services as well." Non-disabled organization = the first party provider Disabled organization = me I don't know why they're using these weird euphemisms or ironic monikers, but that's what they mean.
Earlier quoted context omitted.
No, "another non-disabled organization" sounds like they used the account of someone else, or sockpuppet to craft the response. He was using "organization" to refer to himself earlier in the post, so it doesn't make sense to use that to refer to another model provider.
No, I don't think so. I think my interpretation is correct. > a textbox where I tried to convince some Claude C in the multi-trillion-quadrillion dollar non-disabled organization > So I wrote to their support, this time I wrote the text with the help of an LLM from another non-disabled organization. > My guess is that this likely tripped the "Prompt Injection" heuristics that the non-disabled organization has. A "non…