Max (@talkaboutdesign) header
Max (@talkaboutdesign) profile picture

I have a day job. I still ship products. Apple featured my last one in 20+ countries. Design + speed, in public.

New York
Joined September 2008
Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)Photo by Max (@talkaboutdesign)
Delicious! 🤤
Introducing Geist Serif.
57
No way it gets a PSA 10. Nope. What do ya'll think? xitter.gq/yoxics/status/21160208…
1
105
You can clearly see some damage here. 👇
18
🤯 Cursor found a modified, non-Anthropic version of the Claude Code VS Code extension in its extension marketplace. Cursor says it affected at most 5,000 users. It took the marketplace down, worked with providers to neutralize the malware, and then brought it back. If you use Claude Code for VS Code, update it to 2.1.314. It's another reminder that editor extensions are now part of your attack surface, especially the ones that run agents with access to your machine.
We identified a security incident affecting a modified Claude Code VS Code extension accessed through our Cursor IDE extension marketplace, impacting at most 5,000 users. This malicious version of the extension was not developed by Anthropic. It does not impact other extensions downloaded from the cursor.com/marketplace, Grok Bot, Grok or our other SpaceXSI products. In response, we took down our extension marketplace this afternoon, worked with relevant providers to neutralize the malware, and have reinstated the marketplace after confirming there is no further risk. If affected, please review instructions in this blog post to understand how to find the malicious package and establish a path to mitigate any potential impact. We are asking all customers update to Claude Code for VS Code version 2.1.314: cursor.com/blog/claude-code-…
4
119
Max retweeted
It’s almost incomprehensible how wide the distance is between someone who’s working at the frontier with AI and basically everyone else right now. The crazy thing is these lines are not converging, but only growing farther apart. If the curve of AI progress had leveled off, it would certainly have given most people and companies time to catch up or implement their workflows in a standardized way. The technology would have neutralized competitive advantage because everyone would be using it the same way. But what’s happened is the opposite. The curve of AI progress is basically going vertical right now, which means if you had codified your processes 3-6 months ago you’re likely already behind again. Now imagine most companies that kicked into gear a year ago and thought that they were done. Even further behind now. I’ve never seen a technology in my life and career that improves at this rate, which consequently means you need to upgrade your expectations at a similar rate. Our brains weren’t wired for this, and our organizational change management processes certainly weren’t either. With any form of recursive self improvement, things just get wilder. It’s a time when you will need to be constantly reevaluating the things that didn’t work just a couple months ago to see if they now can be solved. AI native companies certainly have a strong advantage due to this right now, so now it comes down to can the bigger companies keep up. Wild times.
675
856
346
6,922
2,274,480
Key here is "layers." Plenty of AI design tools look great until you open the file and it's 400 flattened groups named Frame 1287. If Brush actually edits the real layers, that's the whole game. Super curios about this.
Introducing Brush: Our first vision-language model for design. 🖌️ Brush sees your design’s layout, layers and vibe the way you do, and edits alongside you. Fix one line of text, restyle a whole deck or build from a blank page. See what Brush can do 👇
3
2
231
Hot take: the Figma file is becoming the spec doc nobody reads. The real design now happens in the browser with an agent open next to it. Figma is where you explore. Code is where you decide.
5
1
243
Quick I have roughly 45% of usage left on my Grok Bot which resets tomorrow. I need some ideas on how to burn it. Give me your best use cases in the comments.
3
157
Great seeing OpenAI doing this for their pro subs. Following in the footsteps of Anthropic. This is going to enable someone like me to create more tools for personal use powered by these powerful models without spending additional money. Bravo!
Starting today, eligible ChatGPT Pro and Business subscribers can claim up to $500 per month in API credits to build apps and agents.
1
1
221
Holy shit. This new Shopify Editions site is insane. It's using the new html in canvas that Google Chrome now supports. If your browsing the site in Chrome and hit the ~ key the whole website transforms. What is this sorcery! shopify.com/editions/fall202…
4
5
517
Love everything about this but touch. - Glowing Apple logo is back! Big win. - OLED Screen with brighter display! Yes please. - Smaller notch. Win win. - Lighter design. Give it it me now. - Touch screen? Meh. Can't wait for them to update the full lineup next year. Right now no reason for me to switch from my late 2025 M5 Pro, 64GB monster.
14-inch is $2,899, 16-inch is $3,399.
11
1
30
10,635
Or just give us macOS on 12.9" iPad Pro.
518
OpenAI says its models produced three research results in math, CS, and physics.They include a possible refutation of the Exponential Time Hypothesis via a new k-SAT algorithm, a proof of the local smoothing conjecture, and an existence proof for quantum field theories like QCD. Math and CS details are on GitHub with Lean proofs, and outside experts are verifying them. Meanwhile, more than 1,000 mathematicians have joined a new Association for Human Mathematics that commits to human-only work.
172
My Grok Bot can now work on my coding projects in T3 Code, with any model I've set up there. The setup is T3 Connect on my Mac mini at home, so the T3 app doesn't even need to be open. Grok Bot sends the work to the Mac, it runs there on my own hardware, and I get a PR back to review. I can be anywhere. I can follow along from my laptop or the T3 Code iOS app, or just let Grok Bot handle it in the background. Today I asked it to fix light/dark mode on my site. A few minutes later the PR was open, the preview passed, and it was merged. Wouldn't be possible without @Tailscale and the T3 Code MCP 🙌 @theo
4
5
370
Doesn't seem like this will hold up. They updated bots vm and the settings were wiped. Hey @poteto is there any way to make this work reliably? Or have some official way to connect to t3code or even claude code cli?
1
55
Excited to finally be able to give this a go. If you used this already how does it compare to Grok Bot, Dots, Instinct?
143
Best software for Mac to record videos looking at camera while reading a script?
2
203
X is amazing. One day everyone is an immunology expert , next day everyone is an expert in foldable phone hinge design. How did we get so lucky?
2
220
Work hard in silence. Let success be your noise. 🏆
18,946
212,174
18,814
1,346,039
27,665,852
I got my Grok Bot coding on my own Mac mini today. No cloud coding agents, just my own machine over a private network. Here's how 🧵 1/ The setup. Grok Bot has its own Linux computer in the cloud. T3 Code runs on my Mac mini and exposes an MCP server. The problem is that my Mac isn't on the public internet, and I don't want it to be. 2/ Tailscale fixes that. Install it on the Mac, then ask Grok Bot to install Tailscale on its own computer. It gives you a login link, you approve it, and the bot's machine joins your tailnet like any other device. It's private and encrypted, and there are no open ports. 3/ Start T3 Code on the Mac. Its MCP endpoint lives at your Mac's Tailscale address on T3's port, at /mcp. Give that URL to Grok Bot. 4/ Sign-in is OAuth. Grok Bot registers itself as a client and sends you a T3 approval page. You enter the pairing code T3 shows you, choose the access level, and approve. The bot gets a token that lasts 30 days. 5/ Keep the access tight. I didn't give it full access, so it can launch and run coding threads but can't add or clone projects. I add repos in T3 myself. 6/ Test it. I had it start a read-only thread on my blog's repo. T3 started Claude in a separate worktree, mapped the codebase in about 15 seconds, and pointed out that I'm missing a sitemap, robots.txt and an RSS feed. Good catch. 7/ Make it the rule. I told Grok Bot that all coding goes through T3, never cloud agents. It passed that to my other bots, including the one that runs my blog. If the Mac is unreachable, they stop and report instead of falling back. 8/ The upshot: my bots code on my own hardware, with my own models (Claude and Codex in T3), in worktrees I can inspect. A reminder is already set for when the token needs renewing. Your AI agent doesn't need your machine on the public internet. It just needs to join your tailnet.
13
1
8
852
I am curious if there is an even better way to do this? Or is this optimal? @theo
84