.@AnthropicAI , you were on such a good roll! A/B testing 20% less weekly usage.
SemiAnalysis says Anthropic is now A/B testing Claude Max limits: roughly 20% more usage per 5-hour window, but 20% less per week for some accounts. tl;dr Same subscription price, less total usage for heavy users.
Starting next week, sessions across Claude web, desktop, and mobile apps will run in auto mode by default on Pro, Max, and Team plans.
Rather than asking you to approve each step, an automated check blocks risky actions. Switch back to manual anytime.
Obsidian 1.15.0 (early access) is now available to Catalyst members!
- Mermaid diagrams use our new rendering engine, Pele, which renders up to 500 times faster and uses your theme's colors and fonts.
- Voice capture on iOS: Quick Capture can record and transcribe voice notes as you speak, even from the Lock Screen.
- PDF export on mobile.
- Smarter list editing: indenting or moving a list item now moves nested items too.
- Accent-insensitive search: cafe matches café in suggestions.
- New URI and CLI commands to append or prepend content under a heading.
- Filters in Bases can now be reordered.
- Community plugins load faster when Obsidian starts.
We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent.
It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.
Introducing Eleven Realtime, a new class of Speech to Speech models.
Voice agents usually make you choose between quick replies or precise answers.
Eleven Realtime gives you both. It starts talking in half a second and gets spoken reasoning questions right.
I built a Command & Conquer style RTS called Battle Drones, mainly using Claude Opus 5.5, which runs in the browser and includes device-to-device multiplayer over an audio link https://t.co/OUGfRDk7DU
Had fun writing this talk for our MPhil students today:
"How to avoid AI Brain Rot"
Discussing how to avoid intelligence atrophy from AI use in work and study. And how to stand out on an ever-narrowing bell curve
This is Pele, my new rendering library for diagrams and charts, compatible with the latest Mermaid syntax.
- 100-500× faster than mermaid.js
- 6× smaller bundle
- No dependencies
- Easy to theme
- Responsive to device width
In pursuit of making Obsidian insanely fast and cohesively designed, I always found the Mermaid library to be too big, too slow, and too hard to theme. Pele will be the new renderer in the next version of Obsidian, but you can start using it now.
Pele [ˈpɛlɛ] is named after the goddess of volcanoes and fire. It's open source under the MIT license. Try the playground!
Six projects from the last few weeks that caught our attention:
A wildfire analysis prototype that maps the fire’s spread over time and tracks hot spots, made with Opus 5.5.
It's almost impossible to keep up with the advances in AI for creative work (seems like there’s a new breakthrough product or model every day!!)
So I pulled a list of the top 40 accounts who are actually hands-on experimenting w models and video/image generation and talking about how AI is reshaping how creatives work.
Let me know who you think I should add!!
Hands-On Creators & Experimenters (Image, Video, Workflows) 👇👇👇
@DavidSHolz: Midjourney founder
@ShaneHegde: Founder/CEO of Air, AI native creative ops
@Gossip_Goblin: AI horror creator
@c_valenzuelab: Runway co-founder/CEO
@medeana: Co-founder Boys Club
@pridesai: ops at Tribute Labs
@bilawalsidhu: Creative technologist building & reviewing AI workflows
@TheoMediaAI: Filmmaker testing and reviewing the new creative-AI stack
@dillonrolnick: CEO, Nous Research
@techhalla: Prompts and practical generative techniques
@heavypulp: Cinematic AI visuals and VFX-style work
@egeberkina: High-quality daily AI image drops and prompts
@AmirMushich: Brand/ad design and usable generative workflows
@PJaccetturo: Viral AI trailers and commercial-style work
@henrydaubrez: Creative direction, festival shorts, Google Labs context
@charliebcurran: Viral AI filmmaking and distribution insights
@lexx_aura: Cinematic AI animation studies
@Salmaaboukarr: Higher-end AI content and brand work
@Mho_23: AI content systems and production roadmaps
@cfryant: Artist/filmmaker perspective on AI commercials and craft
@hollyherndon: Experimental music/voice AI and artist-side generative work
@simonw: Practical testing of new models and tools the day they ship
@emollick: Evidence-based takes on AI (including generative tools) in real work
@levelsio: Shipping with AI, practical generative experiments
@burkaygur: Founder/CEO, @fal@gorkem: Founder/CTO, @fal@robrombach: Founder/CEO, Black Forest Labs
@andi_blatt: Co-founder, Black Forest Labs
@weberwongwong: Founder/CEO of Flora, node-based AI media tool
@yoland_yan: CEO of Comfy, node-based AI media tool
@alexmashrabov: CEO of Higgsfield, leading genAI tool
@mjlbach: Founder/CEO of Hedra Labs
@gmharhar: Founder/CEO of Mirage (formerly Captions)
@nyn531: Founder/CEO of Creatify, AI Ad-maker
@sxmawl: Founder/CEO of cardboard, agentic video editor
@rezakhadjavi: CEO of Runneth, AI teammate for marketers
@viccpoes: CEO of Krea
@sashastiles: Experimental AI artist
@n0w00j: CEO of Melius
@fofrAI: Generative image/video experiments
@JohnBrueckner@Kingsley_codes Hi John, really interested in your fully GitHub-based second brain (without Obsidian as the main knowledge base). How do you structure it so Claude Code and Grok Bot can both access the same info? Any tips on the repo/setup would be appreciated!
@Kingsley_codes Sorry to hear about this! This is why I built a second brain on GitHub. Both my Claude Code and my Grok Bot accounts can access the same information. If I switch to Codex at some point, it should go smoothly since the imporant data is on GitHub. DM me if you want help with this.
I am curious what were the challenges that prevented Google Docs from shipping Markdown support 15 years ago and how they have finally overcome those hurdles
@weswinder You can setup the default reasoning effort.
I've been using Claude Projects for the past few weeks, and no, it's not burned the limit faster.
It is now getting very hard to track where your work and conversations are in Claude and ChatGPT: Local computer? Cloud? Which computer? Phone? The computer connected via Dispatch or Remote? The one you are talking to in Tag? Is that in a project? The current UX is little help.
We just let in every Pro and Max user from the Claude Code Projects waitlist!
If you're new to Claude Code Projects, here's a 4 minute walkthrough to get you started: