How do I follow up my last post…I’ve never had such a response before, I’m still working my way through every reply.
And I wasn’t actually going to post today.
Since last week I’ve been having great conversations and spent a lot of time thinking, without my agent. I’m back to actually jotting stuff down in my notes rather than agents.
Inertia is my problem. Just because this newsletter is sent tues/thurs in a specific format, it’s easier to keep that going. I feel obliged.
But really I’m trying to find something. My thing. And if I do what I’ve done, I won’t find it. I need to change it up. So from now until the end of the year, the formats will vary, the content, even the days + frequency of when I post. I’m in experimental mode. The digital equivalent of going to Thailand, slinging on those awful trousers and finding myself.
I need to post what I’m interested in. For me to continue doing this for a long time, I have to make these for me. So what am I interested in?
I’m following my nose. It’s already led me to have convos with Stripe, Granola, Zed and Raycast about how they’re building in this new world, while staying sane and nurturing a considered culture. I think there’s a ton of interesting stuff here I’ll write and build on top of. Coming soon-ish.
I said I wasn’t going to post today, but you’re reading this…so, what gives? Well in my deliberation of to post or not to post, I felt like I wanted to bring you along in my thinking. I want to bare all whilst we see where we end up.
I’ve stopped using Grok Bot for everything, except to ping me when Carlos Alcaraz is playing. I’m reading more long-form. I’m consuming less slop (and spending a little less time on X). I’m recording a voice-note every morning for a minimum of 10 minutes, and not doing anything with the transcript. I’m letting things percolate, noticing what I come back to.
Agents give you answers to everything. But if we’re ‘soul-searching’ or trying to connect the dots, an answer presented to us may not be the right path. Even if it’s a reasonable one. All I’m ‘optimising’ for here is - what do I really want to do? And noticing what I keep coming back to when I take the time to think on it.
This is not an ‘anti-ai’ stance at all. I love AI. I can do so much more than I ever could. It’s a tool though, not a person, and I use it as such. You probably need a grand total of one tool (an agent). Instead of looking at everything as an opportunity to optimise, I want to use AI for fun.
Quick bites
Try doubleAI’s Maxwell, a code execution model that speeds up software as it runs. Results: 38x faster data pipelines with no code changes.*
Patrick Collison posted about adding a feedback box to what you build as a quick way to keep iterating.
A few links to get your agents to make good UI: Stageshow for making product demos, BoardUI for great-looking components and Looseleaf for a hand-drawn feel.
I enjoyed this post on how Jakub built a personal health app for just himself. He puts out a lot of great visual work.
Nobody knows anything - ‘if you feel behind right now don’t worry you aren’t.’ Fun site, but if it’s hard to read, hit the grid icon in the top right.
Kino - edit video with your agents.
Hex - A free voice dictation app for Mac, Windows and Linux. The landing page is pretty good!
I like this… a Chrome extension that swaps clickbait thumbnails for real stills from the video.
Afters
Its interesting hearing you say you are looking for your "thing" (you are one of my favorite susbtack / x follows, and have always admired how clear your angle was!).
Also, this helped me, just go sit down quietly for 10 minutes / day (don't even call it meditation) and train attention. The more I've done it the clearer the signal to follow fun / curiosity gets.
Trite but really useful as a tool, go read "the power of now".
A few posts back, you'd had this section in the newsletter, 'Try this', where you posted a sample prompt for us to try. It was cool cause it was something that I hadn't seen before. The one you posted turned out to be really useful. Wondering if you can keep that going for future posts?
Thanks, Terry