Nicholas Christensen

Your notes, projects, and history — always there, always yours. Project is called Metacortex. github.com/chrishonson/Me…

Built it with MCP + Google Cloud (Firebase, Firestore, Vertex AI). Being AI-native means jumping between 5+ models and burning out on context switching. The fix: a single persistent memory layer that ANY AI can call via MCP.

It worked! I added MCP tools to ChatGPT so it can save + retrieve memories straight to my personal cloud brain. No more copy-pasting context between ChatGPT, Claude, Gemini, etc. One central hub = everything stays in sync. Demo attached 👇 #Metacortex #MCP #AI

2.9M views. 6k followers. all in one day. wtf. Karpathy just called my agent swarm setup "brilliant or severe AI psychosis." my wife said it's both. a lot of you just got here. so this is my full story: I have a 4-year-old and 6-month-old. I'm a solo founder building a SaaS that competes with companies with hundreds of employees. but this isn't a hustle story. I watched too many people sacrifice their lives for someone else's exit. miss their kids growing up to hit arbitrary revenue milestones. I was that kid. I'm not doing that to mine. my rule is simple: business supports life. not the other way around. everything I build has to pass one test: does this give me more time with my family, or less? AI give me more. I can take a walk after a customer call. come back to PRs ready for review. ship features same-day. be present for bath time and the small moments that disappear if you're not paying attention. > that's the how. and here's the why: good ideas die in silence every day because the people who have them can't afford to be heard. the PR infrastructure is broken. big incumbents like Cision and Muck Rack charge $10k/year just to access their journalist database. these are table-stakes tools - you need them just to know who to email. so what happens? funded startups get coverage. bootstrapped founders stay invisible. corporations control the stories that get told. rich people pay their way out of the ones that shouldn't. freedom of speech is a nice concept. but freedom of reach? that costs money that most people don't have. I've watched founders with incredible products struggle to get a single piece of coverage while well-funded competitors with mediocre ideas land in techcrunch. this whole system is a paywall built to extract rent at every step. I'm tearing down that wall. > so here's what we're building the whole industry is broken. not just the gatekeepers - the entire system. journalists get 300+ pitches a day. most of it is spam. they're drowning, so they ignore everything. PR people respond by blasting 1,000+ journalists and praying for one reply. which creates more spam. which makes journalists ignore even more. it's an arms race to the bottom. everyone loses. the AI tools that exist? built to spam harder. "relentless follow-ups." "scale your outreach 10x." nobody stopped to ask if that was the problem in the first place. what if we point AI in the opposite direction. not "send more pitches faster" - but "find the 10 journalists who'd actually care." - for founders: get the same quality of outreach that funded companies buy, without the $15k/month agency retainer. - for agencies: stop burning hours on research. handle more clients without burning out your team. turn expertise into leverage with AI. - for journalists: fewer pitches. better ones. from people with stories actually worth covering. the big spam factories aren't going to like this. everyone else will. > what I believe one person with AI can now compete with teams of 100. not a prediction. I'm doing it. the world is splitting in two. on one side: specialists getting more specialized, protecting smaller slices of expertise, competing for jobs that are slowly disappearing. on the other side: generalists getting more powerful, building bigger visions with smaller teams, creating the jobs. I'm a generalist. I can code and market and sell and iterate. most people see that as lack of focus. I see it as my unfair advantage. the specialist uses AI to go deeper into their specialty. the generalist uses AI to become unstoppable. > not a unicorn I'm not trying to raise a series a. I'm not optimizing for an exit. I want three things: - location freedom: work from anywhere with a laptop and wifi - time freedom: control my schedule. no one tells me when to show up - financial freedom: never take a job I don't want. never compromise my values for a paycheck the path to all three: bootstrapped, profitable, sustainable. no VC telling me to 10x in 18 months. no board diluting my vision. no boss questioning my decisions. just me and zoe. my skills. my choices. most people see that as limiting. I see it as the whole point. > follow along I'm documenting everything. the wins, the failures, building in public while the world figures out what AI actually means. no course to sell. no community to upsell. just a guy with two kids building something real. if you're someone with a story worth telling - you're exactly who I'm building for. the playbook is being rewritten in real time. might as well write your own chapter. - Elvis

@gregisenberg This was extremely interesting because graphic design is such a huge blind spot for me and I'm trying to branch out of "just engineering"

how to make your ai app not look like it was made with ai (weavy ai, claude, figma etc) use weavy ai to generate custom assets, color systems, and logos across models like flux + ideogram, then compose it properly in figma then paste into claude code/google ai studio etc watch

Here's a free Valentine's Day card maker. Just answer a few questions about your relationship — how you met, what you love about them, a favorite memory — and AI designs a completely unique card around your story. card-maker.space

@unclebobmartin Not sure how to feel about that because at this new level of ambiguity, problems are less about the machine and more about the human factors

Static or dynamically typed? Who cares? OO, Structured, or Functional? Who cares? High Level, low level, declarative, or procedural? Who gives? What matters are the scenarios and the testing discipline, the coverage, and a bit of refactoring to keep the token count and context compressions at bay.

@anitakirkovska @openclaw I think it's increasingly important to protect your reputation and having a bot use your accounts blurs the line. I have also created a separate identity for my autonomous coding agent

I think I’m using @openclaw a little differently than most people. I'm not sharing any of my credentials or giving access to my personal local files. Instead, I set Ava to be a totally separate “person” with her own Apple ID, phone number, Google account, and even its own dedicated Mac Mini. With this, she can create her own accounts with apps like Vercel or Amazon or setup an X account (which she did: hi @avareed_1994) to help me with my tasks. I only share things like reminders, our messaging chat, and calendar access, so it can text me, order groceries, and keep me accountable to go to the gym.. and anything else we figure out in the meantime I know that means I’m missing some powerful persona/work automations for now, but I’d rather move slower than worry about security. and also, this is still so powerful and fun

@openclaw It seems like this will work WhatsApp message app idea ↓ Clawd bot clarifies requirements (conversational) ↓ Break down into discrete tasks ↓ Write tasks.txt to target project ↓ Launch night-shift-agent ↓ Monitor progress / report back

Just to see what would happen I texted Henry my Clawdbot to make a reservation for me next Saturday at a restaurant When the OpenTable res didn't work, it used it's ElevenLabs skill to call the restaurant and complete the reservation AGI is here and 99% of people have no clue

Remember how slow the internet was and when tabs came to browsers we could click links in a new tab and then let them load in the background and we could have multiple tabs loading at the same time. This is how it feels to work with multiple agents. It's hard to sustain for long

Time for a Saturday night AI coding sesh 😁

I had my night shift agent build for me for 3 hours and it exhausted my Gemini API limit so I switched it to Claude code. But really I want it to automatically exhaust my subscription limits first before then switching to open router

@GrousARK @CathieDWood The cable companies weren't ready for it either. As developers were used to change. This time it's a big change! Own your agent framework, leverage compute, and ship!

The software industry is going through its printing-press moment. Historically, the cost of writing code has been high, not because developers were paying to type lines of code, but because of the educational and time investment required to become proficient. A computer science degree alone can cost tens of thousands of dollars, and even for self-taught engineers, the opportunity cost of the time spent learning is substantial. That constraint is now breaking. With Claude and other coding agents, everyday people can increasingly create software for themselves, and more importantly, for others. As a consumer analyst, the closest analogy I can draw is content creation. When the cost of producing content fell toward zero, we saw an explosion in the creator economy. Platforms built on user-generated content, like YouTube, Instagram, X, TikTok, and Roblox, reached massive scale almost overnight. This is what happens when something that was once expensive and difficult becomes cheap and easy. The biggest difference this time is that software isn’t just another creative medium (or maybe you could argue it's the most important one?) It’s the foundation of nearly everything we do across every industry; there is almost nothing software doesn’t touch. So when anyone can create and manipulate software, you’re not just unlocking a new market, you’re reshaping how the modern economy operates. User-generated content creation reshaped culture; user-generated software will reshape the economy I don’t think we’re ready for it.

Maximize Developer Efficiency: Leverage compute of your Gemini Pro and Claude Pro plans; wake to vetted PRs with self-generated reviews, robust against failures through retries, checkpoints, and auto-reverts. Open-sourced for adaptation.

Treat AI as a Secure Collaborator: Uses a dedicated bot identity with limited GitHub PAT scopes (push/PR only, no admin), enabling real remote CI/UI tests while preventing unauthorized merges.

Own your coding agent. I’m sharing my autonomous “Night Shift” coding agent for Kotlin Multiplatform mobile dev. It’s powered by Gemini and Claude Code CLI wrappers. It takes a task list you define and executes while you sleep. Then it creates a PR with passing quality checks

@joecarlsonshow First! I wish I bought more under $300

Hey it's me. I'm from a few years in the future. People are now talking about how they wish they had purchased more Google below $300. I came back in time to tell you.