Hearth & Code

I never got into using voice to dictate my prompts. I have a tendency to be verbose and detailing in my writing, very precise and detailed. I love writing, but suck as a writer. But for me, prompting an AI is less about trying to trick the AI but weaving a story that the AI can understand. So, I guess in a sense, my dream of one day being a writer came through, the only problem is my audience are AI agents, lol!

I actually stopped talking to AI and started typing my prompts again. I realized I think much more clearly when I type instead of speaking. Anyone else notice this?

I used to be a tutorial junkie, Youtube videos and Udemy courses were like a blood drip feeding into my veins. I spent more time looking at other people build things instead of building things for myself. Honestly, it was AI that changed that for me. I'd probably still be stuck in tutorial hell if it weren't for the tools and possibilities that AI tooling allows.

Some people spend 5 hours watching AI tutorials. Others spend 30 minutes building. Guess who improves faster.

I'm building out slowly intentionally. I'm going into a lot of this blindly, like navigating a labyrinth without a guide. I do worry about someone coming in and beating me to the punch, but I'd rather build intentionally and with conviction than bump around in the night and stub my toe. That really hurts!

Founders obsess too much over speed and not enough over decision quality. One thoughtful decision repeated every week beats dozens of rushed ones. Trade "how quickly can we move?" for "how can we make this the last time we solve this problem?"

@divaagurlxw You seem like my kind of people, I dig it. Totally vibed with that full list, been tackling transhumanism a lot in my mind lately. Sorry, but with a list as comprehensive and mind-blowingly awesome as that, I had no choice but follow. Hope to connect!

People obsessed with engineering, computer science, robotics, AI, AGI, machine learning, deep learning, neural networks, LLMs, SLMs, transformers, RAG systems, vector databases, agents, automation, DevOps, MLOps, distributed systems, backend architecture, scalable systems, Linux, Arch Linux, NixOS, Docker, Kubernetes, Kafka, cloud computing, cybersecurity, cryptography, reverse engineering, system programming, low-level coding, Rust, C++, Python, APIs, startups, indie hacking, SaaS, product building, hacking ideas together at 2AM, UI/UX, minimalism, branding, typography, industrial design, cinematography, philosophy, existentialism, absurdism, stoicism, psychology, neuroscience, cognitive science, mathematics, physics, astronomy, quantum theory, futurism, transhumanism, biohacking, economics, game theory, geopolitics, Dostoevsky, Kafka, Camus, Nietzsche, Orwell, literature, poetry, storytelling, dark academia vibes, deep conversations, cafés, notebooks full of ideas, overthinking reality, and changing the world with weird ideas automatically become my friends

@chaosengineerr More like conducting a symphony of agents, weaving a beautiful masterpiece into existence.

founders, vibe coding or AI coding?

I mean, come on, I'm making my own personal soundtrack? We should all have our own personal soundtracks! How cool is that?!?!

I make this not to just self-promo, but to demonstrate what is possible with AI now. Honestly, I only spent a couple of hours with Claude and Hermes to develop the style and the lyrics. Hermes asks me questions about my experiences, and uses my answers to shape the lyrics. I don't want people to be afraid of AI because they feel they will be replaced, I want people to embrace AI to allow their dreams and visions to become a living reality for them.

I really don't know where to begin with this tweet. It's a long one, but I would be honored if you stuck through. Anyway, do you remember the one Family Guy episode where Peter makes a wish or something to have someone around playing a theme song for him? I think the skit worked out this way, Peter goes around with a theme song playing behind him all the time until someone gets annoyed by the music and beats him up.

I just delegate all my work to my agents so I can spend more time posting on X.

I do. Not all the code, necessarily, but I try to grok the high level conceptual overview of it. I might poke in a file here and there and read what it generates. Honestly, my space has been more in workflows and markdown files, getting the agentic harness setup to a point where I feel confident to start coding. In that regard, I do try and review everything the AI generates for me. Outgoing posts gets a verification from me and public facing tweets I generate with AI get a "voice and brand approval" from me before I ship. Nothing goes out automated, I manually review and process each piece of information. My session logs for my self-directed PhD thesis is all managed by AI agents and has a locking mechanism in place in the workflow that once I sign of and verify a session log as being reviewed and approved by me, the AI agent isn't allowed to make any edits or addenums without a formal review and amendment process verified and approved by me.

Be honest. Are you actually learning from AI's code or just copying it

Seriously, I want to read more. I want to poke and prod at ideas, find inspiration, and find interesting content to ingest and grok. I get bored easily, my mind needs constant stimulation, and I love when a post makes me click on "expand for more" and leaves me with a wall of text to read.

You know what's sometimes frustrating, when other people confuse you for an LLM. It seems to have happened to me on a few occassions, and I feel sometimes I leave the wrong impression. So let me set the record straight. I can't just write a single sentence. I see all this single sentence tweets, and I'm like "how can I capture all my ideas and thoughts in a single sentence?" I have to be loud and proud when I share, so I type and I type and sometimes I don't know when to stop. But I do find that I have a particular style of writing and communicating. My neurodivergent brain loves structure and homogeneity, I'm not afraid to use an emdash here and there if it looks aesthetically pleasing to how I want to structure my thoughts. Yes, I do sometimes use AI to generate and shape my public facing tweets. Yes, I do use AI to generate posts for me and content for me. But at the end of the day, I'm still human, I still put significant effort in producing stuff that I can be proud of. I don't shy away from using AI, and I don't shy away from expressing myself fully. Perhaps X might not be the best platform for long-form content, but its how I express myself. Character limits is one of my mortal enemies, and does a disservice to a brain that operates like mine. So with that in mind, lets bring long-form content back. Lets bring discourse and discussion back in the mix. Lets move past one-sentence posts and engagement bait and dive at something deeper. Lets explore ideas together, share workflows, shoot the breeze, and gather around a digital campfire and share stories. Let us not lose out on human connection simply because a platform prefers short-form posts.

I really struggle with just starting somewhere small. I have a tendency to want to design the full system, every layer, every detail, every component. I tend to iterate over and over on the design and the specs and never actually follow through with implementing just an MVP. So for me, developing the discipline and expertise to know how to really start and execute a task through to finish and deliver something tangible and user as a baseline is where I need to improve.

What's the hardest part of building SaaS today? • Getting first users • Keeping users engaged • Standing out in the market • Pricing correctly • Building the right features • Anything else?