Good morning, Internet!
I'm still not sure how I'm going to spend this day. Maybe reading books. Or figuring out what I should be focusing on more... actively?
There are a couple of ideas that could be built on top of Resend capabilities, that I had after building a Resend CLI (to give it to Claude Code):
- An email inbox for AI agents/companies/whatever: I am already using Resend as a (cheap) personal, professional, inbox. Maybe other people could be interested.
- An AI digest (just like the company Abstrakt.news I was writing about yesterday)
- "ai.email": Chat with AI asynchronously through emails?
- Yet another newsletter platform.
A random thought: I love Craigslist-like businesses. Low employees, low complexity, (very) high revenues. At some point, they will finally die (replaced by Facebook Marketplace or whatever), but after making so much revenue with so few employees, I guess that's not a big deal at all.
Ok, let's start building a "buy" button, as they call it. I'll create a small website, accessible at buy.alexisbouchez.com, and iterate from there!
The "buy" button is deployed at buy.alexisbouchez.com! The question is now: what should I sell?
Good afternoon, Internet!
I had to remove articles from Wikistral that were generated with Llama3.2, since references were not properly generated. I'd probably invest in LLM credits, if I had more conviction in the project. But for now, I guess I'll keep it as is...
I gave all my journal entries to Codex for processing. That's the main reason why I'm writing it all: to give all these thoughts to a sparring partner, a LLM in this case.
Random thought: Yesterday, I read about this tweet https://x.com/jawwwn_/status/2020208196443361340 with a screenshot saved in here (just in case), about Peter Thiel worrying about population collapse. Every single time I see this so called "contrarian" (I do not think he's that much of a contrarian, really) talking about this subject, I cannot avoid thinking about ectogenesis, artificial wombs. It's very contrarian, very unpleasant thought, but it seems like the obvious subject everyone is avoiding even though it'd definitely make sense to bring it into the conversation (and break the neverending loop of "we need more babies, we need to find the economical incentives even though we know for sure that it doesn't work...").
Random thought: I wish we could browse Reddit/HN by date. Let's say something like: http://news.ycombinator.com/2026/02/09. I'm currently thinking about this while I'm looking for the original Hacker News post discussing the launch of AWS S3... I have the date, but have a hard time finding the original post, since HN search is not that good.
Random thought: I love the work that the people (could be a single person) behind Internal Tech Emails put into their newsletter. But, same issue as described before, exploring the content isn't really that smooth...
Random thought: I'm listening to this podcast (french), and he keeps talking about the fact that people should follow what they do naturally, without putting too much thoughts into it, and with some results. In my case, I keep avoiding it, but it may indeed be teaching programming. That's the thing I used to do naturally when I was about 12 Years old. So, maybe I should explore this path more seriously.
Startup idea: I just found this project called Shannon on GitHub, basically Claude Code for pentesting? Shannon-as-a-Service could be a great startup idea.
TIL:
- I learned that Typepad was a thing (apparently a blogging platform) from looking for old Hacker News posts
- I discovered this amazing website called https://timelines.issarice.com, providing timelines for multiple subjects, startups, AI, ...
That's it for today!