Yeah, honestly this is a bad move on anthropic's part. I don't think their moat is as big as they think it is. They are competing against opencode + ACP + every other model out there, and there are quite a few good ones (even open weight ones).
Opus might be currently the best model out there, and CC might be the best tool out of the commercial alternatives, but once someone switches to open code + multiple model providers depending on the task, they are going to have difficulty winning them back considering pricing and their locked down ecosystem.
I went from max 20x and chatgpt pro to Claude pro and chat gpt plus + open router providers, and I have now cancelled Claude pro and gpt plus, keeping only Gemini pro (super cheap) and using open router models + a local ai workstation I built using minimax m2.1 and glam 4.7. I use Gemini as the planner and my local models as the churners. Works great, the local models might not be as good as opus 4.5 or sonnet 4.7, but they are consistent which is something I had been missing with all commercial providers.
> I went from max 20x and chatgpt pro to Claude pro and chat gpt plus + open router providers, and I have now cancelled Claude pro and gpt plus, keeping only Gemini pro (super cheap) and using open router models + a local ai workstation I built using minimax m2.1 and glam 4.7. I use Gemini as the planner and my local models as the churners. Works great, the local models might not be as good as opus 4.5 or sonnet 4.7, but they are consistent which is something I had been missing with all commercial providers.
You went from a 5 minute signup (and 20-200 bucks per month) to probably weeks of research (or prior experience setting up workstations) and probably days of setup. Also probably a few thousand bucks in hardware.
I mean, that's great, but tech companies are a thing because convenience is a thing.
My first switch was to open code + open router. I used it to try mixing models for different tasks and to try open weights models before committing to the hardware.
Even paying API pricing it was significantly cheaper than the nearly $500 I was paying monthly (I was spending about $100 month combined between Claude pro, chat gpt plus, and open router credits).
Only when I knew exactly the setup I wanted locally did I start looking at hardware. That part has been a PITA since I went with AMD for budget reasons and it looks like I'll be writing my own inference engine soon, but I could have gone with Nvidia and had much less issues (for double the cost, dual Blackwell's vs quad Radeon W7900s for 192GB of VRAM).
If you spend twice what I did and go Nvidia you should have nearly no issues running any models. But using open router is super easy, there are always free models (grok famously was free for a while), and there are very cheap and decent models.
All of this doesn't matter if you aren't paying for your AI usage out of pocket. I was so Anthropics and OpenAIs value proposition vs basically free Gemini + open router or local models is just not there for me.
I see a ton of my peers driving around in 80k cars. I drive a 20k used one.
I'm planning a writing a ROCM inference engine anyways, or at least contributing to the rocm vllm or sglang implementations for my cards since I'm interested in the field. Funnily enough, I wouldn't consider myself bullish on AI, I just want to really learn the field so I can evaluate where it's heading.
I spent about 10k on the cards, though the upgrades were piece meal as I found them cheap. I still have to get custom water blocks for them since the original W7900s (which are cheap) are triple slot, so you can't fit 4 of them in any sort of workstation setup (I even looked at rack mount options).
Bought a used thread ripper pro wrx80 motherboard ($600), I bought the cheapest TR Pro CPU for the MB (3945wx, $150), I bought 3 128Gb DDR4-3200 sticks at 230 each before the craze, was planning on populating all 8 channels if prices went down a bit. Each stick is now 900, more than I paid for all 3 combined (730 with S&H and taxes). So the system is staying as is until prices come down a bit.
For AI assisted programming, the best value prop by far is Gemini (free) as the orchestrator + open code using either free models or grok / minimax / glm through their very cheap plans (for minimax or glm) or open router which is very cheap. You can also find some interest providers like Cerebras, who get silly fast token generation, which enables interesting cases.
OP mentioned a local LLM rig plus a rather complex setup from what I understood (I could be wrong).
Also, most of the lower end models aren't that good. At this point you can take an experienced dev and start implementing an app using any mature stack (I'm talking even about stuff like Ada, so not just C/C++, JS, etc) on any mainstream platform (big 3 desktop + big 2 mobile + web) and you can get quite far with Claude Code. By far I mean you'll at least do the 80% really quickly, at which point the "experienced dev" needs to take over. I think you can even get 95% of the way there. And that's with a stack that the dev is unfamiliar with at the start.
Claude code has aggressive limits on $18/month plan. You can get much farther with Minimax 2.1 or Qwen3 on the same amount of money. I have noticed opus is much better in some scenarios but Minimax and Q3 are not as bad or behind. Or my setup is now too used to the quirks of those models, who knows. I am using Claude Haiku as small model anyway, just not using Opus because it burns credits very quickly.
I have a 49" QD-OLED panel. I have never been one to find visual artifacts distracting, but fonts were awfully jaggy in Linux to the point I spent a week tinkering with font config and almost switched panels to a larger miniled since code looked horrible. And I'm someone who was fine with horrible VA low res low quality screens back in the day.
The sub pixel geometry on samsung's qd-oled needs very specific font configuration to be correctly displayed, and even then it just stops looking bad.
Fair enough, I do like parent’s a bit better, “blurting processing” feels like a too high default setting right after seeing “I’m thinking” :) - not that any of it matters anyways, communicating _something_ gets you there. Rest it just triaging around the edges what people will call you weird for, and if they are, they were going to anyway.
Apple's M5 Max will probably be able to run it decently (as it will fix the biggest issue with the current lineup, prompt processing, in addition to a bandwidth bump).
That should easily run an 8 bit (~360GB) quant of the model. It's probably going to be the first actually portable machine that can run it. Strix Halo does not come with enough memory (or bandwidth) to run it (would need almost 180GB for weights + context even at 4 bits), and they don't have any laptops available with the top end (max 395+) chips, only mini PCs and a tablet.
Right now you only get the performance you want out of a multi GPU setup.
Mine too! Only 2 years old but I can already see the massive anxiety bursts in him.
If this guy has a non chemical cure, I'm all for it. In fact I'm actively researching children psychologists to stave off the meds as much as we can, the problem is that 99% of psychologists are quacks, so choosing them is tough.
There's only so many times a kid can get sent home from school for biting/kicking/punching before you realize you need some professional help and will do anything to help the poor kid. I wish you luck.
You don't understand what having extreme anxiety at that age feels like.
As someone who lived through that, I refuse to let him. All of memories of school are just feeling anxious about everything, just tight and suffocated, always in a panic. I started living when I started taking anxiety pills at 39 years old, and I can see my 2 year old having the exact same anxiety ticks and fits I have.
I don't know at what age I'll medicate him, but I'll do it as soon as I notice he isn't coping and happy anymore.
Horrifying is forcing him to experience that because you can't comprehend us.
I'm sorry that you're dealing with this - it was my greatest fear at that point. That my daughter seems to not have my disposition and seems happy go lucky is the greatest thing ever. There's no rhyme or reason to my depression and anxiety, it's completely maladaptive and I'm relieved, that knock on wood, she stays happy and light while not having to shield herself from the horrors of the world.
How do you know an anxiety pill is treating symptoms only? What if the cause is physiological, and the pill treats that? It is entirely possible to sit in your therapist's office and mutually shrug because neither of you can find an underlying reason for your anxiety. Sometimes anxiety just is.
I had severe anxiety/depression and majorly recovered from the anxiety component through a year of dilligent transcendental meditation. It changes the brain structure and neurochemistry.
I was on medication during that period and it complemented my practice, provided a stable base to apply meditation and other recovery protocols.
I had panic attacks every morning before school. God, I hated school. Mainly because of the other kids, and when I was older, because of both the kids and the teachers. I remember telling my IT teacher I am using Linux (I forgot why I told her) and she was very condescending. I have a lot of other stories but yeah, school was an anxiety-inducing nightmare.
Often the cause is things that most people can handle, without being able to easily wield the tools to handle them. Having a pill that dulls the symptoms gives space to learn and become adept at the tools
Beyond obvious tumors/lesions/clots/abnormalities, we are not even close to being able to identify the cause of organic anxiety or mood disorders even if we wanted to.
We can say certain behaviors, experiences, illnesses and some genetic identifiers can trigger the conditions, but not the underlying cause. We can say things like some therapy and medication can help with the illness, but not the cause.
Not to trivialize therapy, but for many illnesses, not just mental, a portion of it can be described as ways of learning to live with the illness, not necessarily treating the underlying cause.
> Not to trivialize therapy, but for many illnesses, not just mental, a portion of it can be described as ways of learning to live with the illness, not necessarily treating the underlying cause.
Yeah, I feel like it's fair to describe the cognitive behavioral model. We're not necessarily looking for the cause of these thoughts and beliefs, tho they may come up, we're simply going to challenge them at face value and reinterpret the situation.
Same here, after struggling for 39 years, glp-1 + SSRI + ADHD meds have made me a normal productive human, and 2 years ago I had pretty much given up on the possibility.
Having a child forced me to fix my life, and I'm incredibly happy because of it.
For me? Not being hyper anxious all day (to the point that I just freeze and procrastinate all day), being able to sort of focus on the most important task (I'm still ADHD with 1000 unfinished projects, but at least I finish the things that have to be finished), eating healthy and enjoying exercising (100 lbs down and got quite good at tennis), not entering into a rage state due to anxiety overflow everytime I fight with my wife, being able to regulate my emotions, I could go on and on honestly.
As someone with ADHD, if your productivity was decreased or did not increase in the slightest, then I doubt a doctor would keep prescribing the medication. Such increases do not have to be astronomically large, but I do believe increasing the productivity of people with ADHD is absolutely part of the benefit.
I agree, but I think you’re misunderstanding my comment. I was replying to a snarky comment that seemed to imply that the effect of taking amphetamines is obvious and mundane.
The point I was trying to make is that the effect on someone with ADHD can be profound and transformative, not like going from 80 to 100 but rather from 0 to 100. You suddenly feel like a functional person (I say this as someone with ADHD).
Because unlike the other my journey posts, this one is sharing acquired knowledge and framing it through his (in this instance relatable since it explains the reasons) experience.
Other my journey posts are look at me with only enough subject matter to disguise it.
This post is about sharing knowledge, the others are about sharing experiences.
Opus might be currently the best model out there, and CC might be the best tool out of the commercial alternatives, but once someone switches to open code + multiple model providers depending on the task, they are going to have difficulty winning them back considering pricing and their locked down ecosystem.
I went from max 20x and chatgpt pro to Claude pro and chat gpt plus + open router providers, and I have now cancelled Claude pro and gpt plus, keeping only Gemini pro (super cheap) and using open router models + a local ai workstation I built using minimax m2.1 and glam 4.7. I use Gemini as the planner and my local models as the churners. Works great, the local models might not be as good as opus 4.5 or sonnet 4.7, but they are consistent which is something I had been missing with all commercial providers.
reply