HN user
Anthropic literally don't allow you to use the 1M context anymore on Sonnet and Opus 4.6 without it being billed as extra usage immediately.
I had 4.5 1M before that so they definitely made it worse.
OpenAI at least gives you the option of using your plan for it. Even if it uses it up more quickly.
update: he's in shambles upon hearing the news
or OpenSpec https://github.com/Fission-AI/OpenSpec/
I think it's much better
After 2.20.0 of Claude code where they started not showing what files are read / searches are made by default .. I fucking love how easy it was to ditch Claude code for pi.
They're too small to screw you over. But you've got more time until they do at least.
I had a Claude code instance using 55 GB of RAM yesterday.
I capped my extra usage to that free 50$ and hit 108% usage. Nice.
There's no way you're still talking about verbose mode.. this is insane.
Codex feels much faster. For a while after the rewrite (to rust also I think?) it was bad because you couldn't copy anything from the terminal but since then it's gotten much much better.
Saw this the other day and loved it. Especially seeing Opus 4.5 degrading prior to the 4.6 release (IIRC) and Codex staying very stable and even improving over time.
But FYI the blog post is not about the actual model being dumbed down, but the command line interface.
"It appears minor on the surface but their response to all the comments tells you everything you need to know."
I mean I hope it's just a single developer being stubborn rather than guidance from management asking everyone to simplify Claude Code for maximum mass appeal. But I agree otherwise, it's telling.
Exactly how I feel. I'm happy that more people are using these tools and learning (hopefully) about engineering but it shouldn't degrade the core experience for let's say "more advanced" users who don't see themselves as Vibe coders and want precise control over what's happening.
Yeah just that it's not real time and you have to toggle to see it. It lags a bunch also in longer threads. Definitely a downgrade.
Sorry I'm dumber than the average Anthropic employee, might just take me a few more days for it to "click" that I'm no longer seeing useful information and that this is good.
Jokes about vibe-coded CLI aside, I think that's the issue for me, the defaults are being tailored to vibe coders. (and the general weirdness of trying to fix it with verbose mode)
I like that people who were afraid of CLIs perhaps are now warming up to them through tools like Claude Code but I don't think it means the interfaces should be simplified and dumbed down for them as the primary audience.
Sure you can press CTRL+O, but that's not realtime and you have to toggle between that and your current real time activity. Plus it's often laggy as hell.
I run a loop where I have 4 agents review in parallel after each implementation phase. It just increases the odds of finding issues.
I've switched this over to a team of 4 now that talk to each other to discuss issues they find and it's amazing. They confirm between themselves and if they wrongly identified something the others correct them.
Oh god am I glad to read this. Thought it was my microphone or something.
AI slop commenter account above
A few more years and some more ram on these earbuds and we'll be able to run some nice local earbud kubernetes clusters
Is there any way to report AI slop accounts on HN? This entire account is generic AI slop comments.
I've always dual booted windows with some Linux and used it like 90/10.
I haven't even tried windows 11 even though my PC is compatible.
Went full Linux and I'm not sure what I was missing at this point that I needed from Windows.
Ran Pop OS (cosmic) which is the new Wayland based one but unfortunately it's still buggy and then I switched to a gaming focused Linux called Bazzite which has been perfect.
Tiny learning curve because it's an "immutable" OS but have everything I need running on it plus everything gaming related works out of the box.
I mean marriage is a global concept but it feels like the US makes a huge deal about it.
Like two people can't be together without being married.
But mostly it's a low effort low with quality comment that adds zero value and implicitly passes judgment on those who are not married and don't have kids.
As if married people with kids are the happiest people in the world lol.
Unfair or not, same thing for me.
Then I'm not even focused on the content more than I'm scanning through it for signs of AI slop writing so I don't have to waste brainpower consuming that which took no brainpower to produce.
Also unfair perhaps but I think writers in particular, like the author of this post, should be aware enough of the patterns of AI written slop to consciously avoid them nowadays.
It doesn't matter if you used to write like this, the reality is people will question you now if you do.
Never heard of graphite before today. Were they built specifically for AI code reviews or it's a pivot / new feature from a company that started with something else?
Unfortunate that hacker news doesn't have reply notifications but I'm curious what you did when retiring it.
Just recycle the parts? Was it your main and only server?
I have that server running Truenas, I have another PC I had built for friends and family for Plex only, and I have a third one running an ethereum validator which is the most powerful but only does that.
It's not stuff that would sell for any price i'd care to get and just throwing it away / recyling it feels bad since it still works.
Agreed on Bryan Johnson. Before I actually watched a bit of his content, I just thought he had nutjob vibes and looked weird. No offense intended, if possible.
But honestly he just seems like a guy enjoying a fun project. He seems calm and happy in his videos.
Barring any hidden issues with chronic depression, it would be unlikely that he's unhappy. He's very well off financially, has a nice beautiful girlfriend who's with him in his journey, he sleeps a ton, works out, eats well and in general experiments with life.
I'm running Truenas Scale on my old i7 3770 with 16GB DDR3.
Obviously got a bunch of datasets just for storage, one for time machine backups over the network and then dedicated ones for apps.
I'm using for almost all my self hosted apps.
Home Assistant, Plex, Calibre, Immich, Paperless NGX, Code Server, Pi-Hole, Syncthing and a few others.
I've got Tailscale on it and I'm using a convenience package called caddy-reverse-proxy-cloudflare to make my apps available on subdomains of my personal domain (which is on CloudFlare ) by just adding labels to the docker containers.
And since I'm putting the Tailscale address as the DNS entry on CloudFlare, they can only be accessed by my devices when they're connected to Tailscale.
I think at this point what's amazing is the ease with which I can deploy new apps if I need something or want to try something.
I can have Claude whip up a docker compose and deploy it with Dockge.
You didn't read the article. It talks about the tumbling of share prices of digital asset treasury companies.
Is this sarcasm
Valve could learn from Google here