The AI Corner
Subscribe
Sign in
Home
Claude & Anthropic
Prompting & Context Engineeri…
AI Agents
AI Tools & Models
Business & Investing
AI Agents
Latest
Top
Discussions
Your Obsidian Vault Finally Has a Model That Can Read All of It
GPT-6 Astra finds a note from 6 months ago with 96% accuracy at 1M tokens. The setup, the effort-tier protocol, the cost math, and the 3 prompts that…
Sep 18
•
Ruben Dominguez
64
2
Devin Got a Mac. Here’s the Handoff System for Shipping iOS Apps While You Sleep
Four layers of Cognition’s new macOS stack, eleven prompts for the verification loop, and one cost model that tells you when a Mac VM beats your own…
Sep 15
•
Ruben Dominguez
65
1
Zuckerberg Just Killed the Prompt. Muse Takes Goals Instead.
He admitted Meta got LLM scaling wrong, confirmed Watermelon is shipping, and explained why even Meta can't read your agent's memory.
Sep 14
•
Ruben Dominguez
73
4
1
The True Biggest Risks in the AI Thesis
1% of OpenAI and Anthropic's customers generate 80% of their enterprise revenue, and the compute promises stacked on top of that revenue start coming…
Sep 10
•
Ruben Dominguez
58
3
2
Max Junestrand (Legora) Says You Don't Need Domain Expertise. In 18 Months He Proved It at $100M ARR.
The founder who got laughed out of a YC interview for not knowing lawyers came in types built the category leader anyway. The 10 takeaways.
Sep 9
•
Ruben Dominguez
60
9
9
Garry Tan Runs YC at 400x His 2013 Output. His AGI Is a Folder of Markdown Files
He calls it Personal AGI: memory and skill files that get sharper every day, run in the open for years. The 10 moments.
Aug 31
•
Ruben Dominguez
72
2
3
Peter Steinberger Almost Deleted OpenClaw. Then 4.7 Million People Downloaded It.
Built in an hour out of hunger. Then a nonprofit, a security scandal, and 111,000 pull requests, all before a plan existed.
Aug 27
•
Ruben Dominguez
67
1
3
The One-Person Business, Built on Claude
A course for people who intend to still be operating in three years.
Aug 26
•
Ruben Dominguez
86
3
5
The Task Economy Is Real and Almost Everyone Is Valuing It Wrong
Benchmark says data will be AI's next $1 trillion category and the growth backs the claim. But the money is not buying services. It is buying human…
Aug 21
•
Ruben Dominguez
73
3
3
How to 8x Your Code Output Using Context Engineering
Anthropic’s engineers merge 8 times more code per day than in 2024. The context engineering playbook everyone is copying explains almost none of it.
Aug 20
•
Ruben Dominguez
89
3
3
Why Most Self-Improving AI Loops Fail and How to Build One That Works
Feedback loops are the pattern everyone in AI is saving this year. Most of them improve nothing and the fix is something anyone can build in an…
Aug 19
•
Ruben Dominguez
82
2
1
While Everyone is Waiting for the Next Model, Your Agent Can Learn Tonight
A guide on continual learning for AI agents.
Aug 12
•
Ruben Dominguez
77
1
2
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts
