I am accidentally building a plugin for a video editor. I asked Claude to analyze two days of my video-editing sessions with raea in CC, and asked to improve my flow by building tools. Here you go, a video editor not made for manual editing
Gleb Kalinin retweeted
I just finished talks about my software factories at two conferences. Afterwards a bunch of people came up and told me what they were surprised about. Here is a distillation of what people thought were hot takes from my two talks.
First hot take: I tell everyone to slow down. Most speakers go on about how fast they're going, but I strongly believe in pacing AI outputs to human speeds. I regularly tell AI-pilled CIOs and CTOs to chill: almost nobody is _actually_ going fast with AI. Engineering still moves at engineering speeds, and business moves even slower. The ones pushing too hard are starting to see downtime for it.
If you don't pace your AI to human speeds, everything starts breaking around you, even at small operations like my online video game. I only have about a dozen humans working with my factory, and only a few hundred active old-timer players in the community who get along well, and still, introducing AI has still been very rocky.
My agents have, at times, overwhelmed my players with features, my accountant with emails, my merge queue with tickets, and my machines with load. Agents have pissed off my admins, resulting in name-calling, they've made messes in the game, and they've launched broken OSS without permission. The list goes on.
I can get away with all this chaos because everyone involved in my operation is very forgiving. They're excited by the potential. But even so, I had to learn to slow down, real fast. And so that's what I preach.
Hot take #2: Should you buy a software factory? Do you even need one? No. You probably don't need one. Factories take an extraordinary amount of upkeep. They turn into garbage mountains unless you put a frightening amount of time and tokens into curating them. I had to burn my previous five factories down since it eventually became cheaper to rewrite them than to fix them. So I'm not out there advocating for software factories or swarms just yet. If you do want one, be aware that it is a significant commitment of both money and human leadership to keep one running.
Several people told me that I was the only one who stood up onstage, at either conference, and said Build, don't Buy. But my experience is that software factories grow to be highly bespoke to your domain. Mine is super tangled up with my game, and I think that's inevitable. Because the factory is not just building my product, it's supporting it too. It's a company in a box. And they're easy for Claude to build for you.
Third and arguably hottest take: Next year will have a lot of human/agent and agent/agent office drama, resulting from bringing in AI employees, and it will be gloriously funny, but also frustrating.
Many of you are already on this bandwagon, but everyone next year will have a personal chief-of-staff agent. Busy people, even hobbyists, will have whole entourages of helper agents. The hot take is that people will become so dependent on these little private entourages, that they'll bring them to work. Your people will talk to my people, etc. Whole protocols will emerge for private entourages negotiating work between humans, inside companies big and small.
Nearly all humans will wind up having to start working with AI employees next year, and it'll be the beginning of the biggest cultural change we've seen in a hundred years.
So that's the hot takes. Slow down. Only build a factory if you think you really need one. And focus on empowering yourself with your own agentic chief of staff (or team). Your entourage will define how you work with others, so start by perfecting how you work with them yourself.
Experimenting with ways to represent product features. Every release can become a demo of a product. Pictured: a real MacOS pi SDK-based agent, running DeepSeek
1
56
Build similar for yourself in a group of co-builders in a 3 week cohort → agent.salient.community/
11
allo.joshpuckett.me/ dope set of free fonts (install via npm for web fonts included)
54
morphicons: any SVG icon morphs into any other. Lucide, Tabler, Heroicons, stroked or filled — spring physics, 7 KB core, zero deps, MIT. React/Vue/Svelte/RN/Astro + vanilla element. Closed-form Procrustes is why a hamburger folds into an X at 45°. morphicons.com
47
Gleb Kalinin retweeted
AI interfaces nowadays are still pretty much turn-based.
In my PhD thesis, I thought about what interaction would look like if we could go beyond taking turns.
One example is intra-turn interaction:
Gleb Kalinin retweeted
this is the stuff we’re not seeing enough of. we were promised the future but all the labs do from a UI perspective is fancy chatbots. very cool!
Peekaboo: macOS automation for humans and agents. Captures pixels, reads the accessibility tree, drives input, and ships an agent runtime + MCP server so agents get shell-level primitives.
The missing half of computer use isn't the browser — it's the OS.
peekaboo.sh
4
89
Public-domain video search just got useful: Arkivism indexes archival films at shot level — each shot captioned, with its own hover-to-play preview.
4M+ CC0/PD images and videos from 51 archives.
arkivism.com
34
Building a tool for people who need database for themselves and their agents but don't have any desire to maintain #Obsidian → workroom.salient.community/
It's a set of skills + Mac App that lets you easily sync your bases via LAN or Tailscale and maintain it via agents
55
Marsha M. Linehan, who created Dialectical Behavior Therapy, has died at 83. I am releasing an update to my cognitive toolkit — a collection of DBT skills you can practice with your agent → github.com/glebis/claude-cog…
Not a replacement for therapy
58
Justin Schroeder on agents: we're in the "line-shaft era" — one giant model in the middle, every task belted back to it.
Even if everything works on Claude, businesses don't want to trust critical parts to one provider.
Fix: standards.
youtu.be/wIPr9GmLRaQ
My first DeepSeek Harness plugin is out — github.com/glebis/deepseek-h…
It's a plugin that helps you design plugins from the context of your life
1
66
On-device AI that understands your entire life overnight.
Sentient OS (YC F26) makes any Mac a "touchscreen": tap the notch, sensor fusion across lid angle, ambient light and gyro. Its LLM runs on 8 GB Macs, reading email, messages, Granola transcripts.
sentient-os.ai
1
72
NVIDIA not only released open weight models, but ran harness research.
nvlabs.github.io/SoL-Pi/
1
2
52
Their auto-research loops on their own agent harness looked for token savings under a no-regression constraint.
1
29






