Just asked Fable (low) in three different chats for three different websites to see how it would improve the design, UI, UX - not one of the chats actually took a screenshot, even after the second reply from me said to do just that. I need Sol unlimited tokens please.
Fable (right) v Sol (left). Need Orbit feature debugged. The super app era is upon us.
Had a discussion with a founder last night about behind the scene war he experienced at the frontier labs (he & his brother are at different ones). Also mentioned extreme money flowing thru.
Where GPT is going (allegedly) connecting all apps to their marketplace, then replace them one by one. It will be the one stop shop - health, wealth, etc. They will crush enterprises faster than the little fellas.
Vibe Coding 10AM EST daily. Got to be the first East Coast Viber!
I’ve only used Claude for months, most of the time with two accounts - both on 20X max. After complete breakdown in quality I switch reluctantly to OpenAI.
I cannot switch back any time soon. Sol’s planning, output, design, detail of execution is far beyond Fable.
We're making Claude Sonnet 5's introductory pricing permanent.
We launched Sonnet 5 in June at $2 per million input tokens and $10 per million output tokens through August 31, and that price will remain unchanged.
Literally cancelled both Claude Max plans and upgrade OpenAI to $200 Max. Thought I’d never see the day, but here we are. I still believe in Dario and Boris but the product speaks for itself.
I'm going to cancel Claude. It's just so bad, I can't believe it.
It's just lazy. The most recent example: I have Claude check my inbox for important emails, summarize them, work with them, and send out replies if necessary. I caught Claude again simply not reading the email thread to the end and just ignoring the latest emails.
When I asked him about it, Opus 5 just said: "Valid point. I didn't read it."
I mean, seriously. What the heck? You have to babysit it every time.
Opus 5 on left, Fable on right. Fable succeeds where Opus failed for an hour. Opus has to be heavily throttled for a new model release from Anthropic this week.
Kimi not allowing sign-ups. Fable now part of the weekly usage. It's compute, will always be compute. Software will never be the issue. Kind of like cameras are not the issue, the internet delivering the resolution of the camera is.
Four American labs. Model releases will turn into software updates where - we’ll get to a point most models will roughly be good at everything and releases will be patches, incremental improvements and security guardrails. Then “big” releases twice a year. I suspect this in 2027.
Opinion on AI: Private capital is in endless meetings, scrambling and will be left behind because they cannot agree with what to sell and where to deploy capital. Whoever Pivots faster wins.