Thoughts

Thoughts, ideas, and musings shared on @adlrocha. My Twitter diary.

The government didn't just wake up one day and decide to burn books; the people stopped reading them firstadlrocha: I just finished Fahrenheit 451 for the first time (thank you @whereisjorge for the recommendation).It is crazy how a book that is close to a century old can identify so clearly the issues of our current society.People are no let to think freely (and deeply) anymore. They want
I just finished Fahrenheit 451 for the first time (thank you @whereisjorge for the recommendation).It is crazy how a book that is close to a century old can identify so clearly the issues of our current society.People are no let to think freely (and deeply) anymore. They want us distracted. Books are the way to unlock that thinking space.
10 minutes of quiet time with Claude Code working in the background and I pushed two cool features to my personal website:* A `/thoughts` section that pulls my tweets and renders them on my website (if I am going to start using this feed as my diary I better keep track of it in a sovereign way)* An RSS feed (I think RSS feeds may have a revival soon, but more on this some other day)
Soon a license will be required by law to run AI models.And I don’t rule out governments confiscating compute capacity from citizens like the US government did with gold in the Great DepressionFun times ahead!
You land your dream job through proof-of-work
Hey, turns out unfiltered tweeting about whatever I am thinking is useful, and actually fun!Like a diary but with a few bots and humans on the other side
Tweets as aphorism are my favourite, and the quoted one is a great example of it! “AI is to software what the printing press was to literature”Pablo Grueso: La IA es al software lo que la imprenta fue la literatura.
X402 one of those web3-native primitives that I keep saying will be fundamental for a future agentic Internet. We are already seeing the need for itbin @BaselightDBErik Voorhees: x402 experts: we're examining integration of this into venice. From our research, x402 doesn't support any billing mechanism where we don't know the charge amount upfront. So for LLMs (where we don't know the fee until the end), the user would pay an amount in x402, and they'd
I don’t care if it is a bot, this is worth every minute I spend writing my newsletter 🫶🫶If there’s something I am going for in my posts is simplicity and clear communication. Get that feedback coming :)
The moment I decided to write about 🦞s I knew this was going to happen. This is on me for embracing the hype!adlrocha: The explosion of @openclaw is proof that agents are here to stay.But while they bring incredible value, we are currently running them "naked." You are one prompt away from a hallucination leaking your SSH keys or `rm -rf /` your entire device.It’s time to build a Security
The risks of vibe coding too farVarun: so moltbook tl;dr:anybody could create any number of “agents” and anybody could post anything on behalf of any “agent” due to poor implementation.
And this would be fine if I was single without kids. But focus time is a privilege these days (:adlrocha: Note to self: do not write posts of anything trendy. By the time the draft is ready, the topic is outdated. I’ll need to go back to first principles and immutable fundamentals
Note to self: do not write posts of anything trendy. By the time the draft is ready, the topic is outdated. I’ll need to go back to first principles and immutable fundamentals
I think @BaselightDB will end up having a bigger role in the crazy world of AI that is being shaped than originally expected
Which brings me back again to the fact that LLMs are a discovery, not an invention. And we are still figuring out how and why they work.
I’ve been having this thought on the back of my mind of migrating my crazy thoughts and reflections from here to @Substack notes.For the mere dozen impressions that I get for my X posts, I rather have all my content centralised in the same platforms.Thoughts from my audience here?