I might not be using the correct terms. But, here is what I was seeing:
- in a running session text display would slow down intermittently. Like 2-3 words a second
- occasionally terminal would stop, some times in the middle of showing a response, and would need to restart the terminal to get it back. But I tailed the session logs and Claude code didn’t think anything was amiss. It was proceeding like normal
After switching to Ghostty these issues went away and has been rock solid since then.
The issue the OP mentions was likely a known memory-leak in Ghostty that has since been fixed. Claude was triggering the bug in Ghostty but the bug was in Ghostty. Ghostty's author blogged about it
Not necessarily. I'm thinking of TV shows now (yes I need to go out more) where people lose their inhibitions (House did it iirc), and they "became" assholes - or they were assholes all along.
But if you're an asshole on the inside but it never comes out, are you really one? This is a pretty dangerous trap. Compare abusive situations where someone needles someone else for ages until they've had enough and the target gets angry. What would be a good response? "I'm sorry I made you angry" or "I knew you were really like this all along!"
(there is no correct answer by the way, it depends on many factors and relationships can't be abstracted away that easily, let alone by an external person on the internet who hears only one side of the story)
Anyway, if Hitler was Hitler but had filters so he never actually said what was really on his mind, we wouldn't have had WW2 and co. In theory.
I expect it will be in ui and integration as the platform builds out further. Then you'll get people who don't want to switch because they don't know how to use other things, like with OSes, phones (android vs iphone, heck even Samsung vs other androids for some folks), office products (why is no one using free Google docs over paid m365), etc.
Opus 5.5 is the frontier today. But there are more ways for Anthropic to lose than to remain in that position. I'd take Anthropic stock if you gave it to me but I sure as fuck wouldn't buy into it.
The tool is not a moat when you can tell Opus to build you a better tool.
based on how bad these models are at one-shotting complex real-world software engineering problems -- it always has to be a conversation -- there is absolutely no way they're one-shotting complex math proofs.
Also, goes without saying, but put Sam Altman in jail, obviously.
Rooftop solar (ideally with battery backup) allows you to be independent of the grid (yes, even if you're interconnected, if you do it properly), which means PG&E's garbage infrastructure can largely become irrelevant if your system is properly sized.
If you're gonna rank them: Anthropic > OpenAI > Google > Meta > (everyone else).
Gotta think Grok's cognitive dissonance due to the need to rewrite reality for Musk's political and moral biases is always going to keep them behind the frontier, but this is just speculation. The amount of money and power they have should put them at the top and I can't otherwise explain why they're consistently behind. Maybe Musk is simply too cheap/control freaky to throw enough money/benefits at good employees.
reply