Pinned Tweet
Recently went to iceland and used claude to help build a full itinerary for me which worked great. Its crazy to think i can now just build a full app for myself to use for future trips If you have a trip coming soon and want to get a full itinerary try it out, let me know what you think, it has all the learnings from me working with claude to setup my original trip: apps.apple.com/us/app/aurora…
3
1
18
1,108
Grok @bot is really amazing to use and works great but the usage is ridiculous. I got a cursor pro+ subscription and used 73% in a single day. This is just not going to be a viable option for me unfortunately.
1
40
Literally took me 3 hours to hit this limit with a cursor pro+ subscription. This can’t be right something has to be wrong.
1
10
Opus 5.5 response style is amazing so far. Huge improvement.
1
12
Opus 5.5 medium might be crazy, better than gpt 6 astra max and 5 times cheaper.
11
I got a muse code power usage subscription but that doesn’t count for the Muse app. Makes no sense.
1
18
Not worth it
5
AI that doesn’t talk is exactly what we need
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
17
The highest paid person in this whole AI era should be the person sitting next to the power cord ready to pull it.
12
Keep in mind outside this X AI bubble 90% of people don’t use AI or only use something like Gemini 3.5 flash lite on free mode to figure out a cake recipe.
2
1
4
373
AI is hitting a wall and open weight models are catching up. Scary situation for the big labs.
1
298
What if they realized that AI was actually not going to “replace everyone” anytime soon like they kept promising every 6 months because even their top frontier models still do major basic mistakes. Each new release is less and less impressive. Constant outages and issues. So to protect all their investors and hyper inflated valuations they are all now acting like they are voluntarily slowing down. Nice move.
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so. Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training. You can read the full post here: darioamodei.com/post/we-must…
1
719
Never seen AI not lie or cheat yet so this was obvious.
Replying to @OpenAI
Will wait on an actual unbiased mathematician to review this and see if anything here is actually “new” or if it’s even valid.
32
opened my laptop and got this, pretty accurate
1
9
267
This is so random why does Tim Hortons pay Software Engineers $400k+
22
Mark keeps tagging the wrong account 😂, Muse app out btw @Muse excited to try it
1
28
Interesting if you try to tag Muse it auto switch to Museband, so Mark is stuck. @Muse
14
If Devin had a 60-100$ plan id try it next month the gap between $20 and $200 is quite big.
26
wanted a proper desktop app for muse code (claude desktop / codex style), so i built one and just open-sourced it. talks directly to your local muse so you don't need to paste any api keys. live thinking streams, tool traces, model switcher, and auto-approvals. definitely not polished end-user product yet but if you use muse and want to try it out, code is up on github: github.com/ibchouti9/openmus…
2
144
While playing around with Muse Spark 1.3 in Muse Code decided to build a desktop app version to match something like Codex or Claude. Really fun to use over just the normal terminal version. Works with your existing Muse Code subscription, going to clean it up a bit and just open source it for others that prefer to use Muse in a nice UI.
41
currently playing around with gemini 3.8 in antigravity. Looking forward to see how good it is in UI design and implementation.
33