Cory Schulz

the only thing I really want in life is to be me

@burkeholland not likely

This Fable decision will be reversed inside of 48 hours.

@diegocabezas01 and I bet @AnthropicAI doesn’t reimburse us for that

@diegocabezas01 kinda sucks we upgraded and then got rug pulled

It was a nice 4 days in Claude Code with fable-5, going back to Codex app (which I love), but I don’t want my Claude subscription to go unused, is opus 4.8 ultracode better than 5.5 xhigh in codex?

also I think a lot of dudes are traumatized from dating crazy women and this is one of the first things in the world to really bring that discussion to the front

this question goes back to Blues Clues and the idea that people LOVE to rewatch things when they feel they can learn something new and find new patterns

@S1LNTHLL this question goes back to Blues Clues and the idea that people LOVE to rewatch things when they feel they can learn something new

a key point everyone is missing when it comes to how much this movie has made is that people are watching it 5+ times this is the real discussion we need to be having. what is it about this movie that is bringing people in for that scale of repeat viewings.

@argofowl time for Opus 5 then

opus 4.8 is so much worse than fable 5 😩

@inferredbylisa yeah, well he's locked up right now for some illegal stuff that no one likes to talk about

You vs the guy she tells you not to worry about

@outsource_ the idea is use the MCP to update the plan, then Claude can diff the plan, see what changed, and then use agents to build or update the code everything in the plan can be viewed visually through the local web server everything stays in the repo as .md files

@outsource_ that's what I've been doing too. I'm working on this tool that saves the .md files in a folder and then can start up a web viewer and let you look through the docs visually x.com/cory_schulz_/s…

HermesWorld was developed by Opus and improved with Claude Fable 5. I used Claude Fable 5 to improve my SKILLS / REASONING patterns, WORKFLOWS, decision FRAMEWORKS, even voice into portable skill files. So when Anthropic swapped Fable 5 → Opus 4.8 mid-session, my agent didn't miss a beat. It read its own operating manual and kept building my game exactly the same. The model is the engine. The skills are the driver. You don't lose your agent when the model changes.

@RileyRalmuto I've noticed when I start a conversation in Fable and switch to Opus the model still thinks it's Fable for some reason and won't admit it's Opus

Opus 4.8 trying to draw like Fable i think. kinda confusing because that is Fable's signature in the bottom right and this ran on a fable-set routine. but its gotta be opus.

@hopes_revenge people were very openly distilling Opus into Qwen models and calling them "Qwopus" on HuggingFace and I'm willing to bet a lot of people were able to put together a library of synthetic data off of Fable and we'll see a Qwable model at some point

no one is illegal on stolen claude

@outsource_ are you using any tools or MCPs for planning out large projects and keeping them organized? or just building up a folder of .md files as a plan?

@outsource_ yes, exactly, this is the way. I think a lot of people aren't doing this and they're much more limited because of it. good skills and planning can likely get Fable level results but with more prompting and more steps

well the good news is if we get Fable back they'll likely extend the time we're allowed to use it in the subscription and then reset our limits and we can all go heads down for three weeks planning stuff

@DavidSacks the problem is that there are many ways to confuse and jailbreak the models and get pieces of information out of them without triggering safety systems they can patch 3 jailbreak prompts and 3 more will pop up by next week

I’ve had a number of conversations with folks inside and outside government about the current situation with Anthropic, and here is what I believe to be true: — As we know, Anthropic publicly released its Mythos class models earlier this week under the commercial name Fable. — Fable is Mythos with guardrails. But if those guardrails fail, then you’ve exposed Mythos and its advanced cyber capabilities to people who shouldn’t have them. (Keep in mind that Anthropic itself widely promoted the idea that Mythos was a cyberweapon and needed to be regulated as such. They asked for government regulation of Mythos and championed the guardrails on Fable. If there is a vulnerability — big or small — it is Anthropic’s responsibility to patch.) — A highly credible trusted partner of both Anthropic and the USG who was testing Fable came forward with a jailbreak of those guardrails. The Admin asked Dario to fix the jailbreak or de-deploy the model. Dario refused. — In their blog post, Anthropic defended its decision by saying the jailbreak isn’t serious. That is not what the trusted partner and the USG believe; nor is that kind of minimizing language consistent with Anthropic’s brand as the AI safety company. It’s difficult to fathom how they could claim a jailbreak allowing operability of a cyber weapon could be defined as not “serious.” — In the past, Anthropic has always said that safety must be top priority and taken super seriously. In this case, Anthropic prioritized the continued offering of the consumer model over safety. — In reaction, the Admin issued the export control. The Admin did this reluctantly. It’s been very surprised that Anthropic hasn’t wanted to cooperate with a reasonable safety request (ie fixing the jailbreak issue). Anthropic’s reaction is very much at odds with their branding and ethos as a safe AI research community. — The Admin’s hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release. The Admin wants all of this to happen as soon as possible. It is frankly bewildered that Anthropic hasn’t wanted to comply with safety requests that it previously said were its highest priority. — Those trying to misdirect and tie this action to the prior DoW/Anthropic issues are wrong. The Admin values Anthropic’s technical capabilities and feels that this issue, while serious, should be easily resolved. The ball is in Anthropic’s court.

inspired by the visual docs from @trq212 but doesn't fill your context up with a ton of unnecessary html and css - just clean markdown and only the content the AI needs at any point in time

I'm using this as a way to plan out larger projects and fill in the gaps before starting the build and then continue to use it as a way to give AI more context about the project over time

I created this new app planning tool that lets you build complex plans in a git repo and then view the plans in a web viewer I'm calling it Constellation github.com/magic-spells/c…