As I sit here staring out my window contemplating the world’s tiniest typewriter, I’m remembering this founder interview with Zeb Evans of Clickup in which he shared his daily journaling and annual review process. He writes daily and at the end of the year reads his last 5 years-worth of journal entries, reflects and makes a plan incorporating the learnings. He cited this practice as a key driver of Clickup’s ability to iterate which has led to its massive success.
I’m not intending to do anything nearly that extensive but I’ve been writing monthly #buildinpublic updates the past few years and taking inspiration from Zeb I’m doing a nano version of his process having just read through all of my 2024 updates. I wanted to take a few moments to reflect, recognize the accomplishments, distill the learnings that might be valuable to others and then give some insights into what changes I’m making for us in 2025.
First here’s the outline of what I’m going to cover here:
Learnings
Let’s start with the key lessons and ideas that could map to others’ circumstances and therefore have some transferrable value.
The primary challenge I face with this whole project is one that is common amongst first-time technical founders and they call it the “Field of Dreams Fallacy.” It’s the same mistake we made with JumpBox and ScratchAudio (building a product first then retrofitting how to sell it afterwards) and yet… here I am in the same position nearly two decades later…
Image credit: Leading in Limbo
The problem is I have an unshakeable vision of a future in which anyone can easily plug-in and contribute his/her unique skills towards building new impact initiatives, for-profit ventures, open source hobby projects or bolstering existing charities and businesses. I see a pervasive hunger in corporate employees for more meaning in their jobs and uncertainty on how to create it. I also see where the puck is headed with public and private companies getting pressured towards increasing efficiency via staff cuts and AI. IMO we’re moving towards a gig economy where eventually the norm will be to build up your personal brand, carve out a hyper-niche specific skill and sell your products and services to multiple companies any of which would have otherwise employed you full-time.
Now in that gig economy-based future world, some portion of clients would ideally be enlightened enough to see the value of getting skin in the game from their contractors by cutting them in on some form of ownership (options or equity). Maybe they do hybrid deals where contractors take some % of options in lieu of cash or perhaps they offer a performance-based stretch goal bonus in the form of equity grants but the smart companies will seek better alignment from gig workers by giving them a path to be on the cap table and share the upside. My aim is to accelerate getting to this reality and systematize how it can work at scale across organizations. The sooner we can live in a world that has this landscape, the better. It gets people to financial and temporal freedom so they can be more sovereign and selective in how & where they invest their time and therefore increases the percentage of people who wind up doing what they were born to do in the world. And that can only be a good thing.
I also see a day in the near future where the low-level “brick-laying” type drudgery of one’s daily work are delegated to agentic AI frameworks freeing us up to be architects and operate more in the playmaker role vs the quarterback role. At a personal level the wealth unlock here will be nothing short of transformational. David Friedberg gave an excellent breakdown of how to think about the evolving role of AI with respect to one’s job in this interview. I was an early Photoshop user back in the day and remember well the advent of Kai’s power tools – his analogy is spot on in terms of how to think about AI. At a societal level this AI-driven wealth unlock represents unfathomable prosperity assuming we navigate the transition right. There is endless FUD around “AI will take our jobs” – some of it justified, some of it not. I am no more worried about AI replacing good engineers then I am worried about calculators replacing good mathematicians. The types of jobs that can be easily obsolesced by AI Agents are the types of jobs people should not want anyways. This reset is coming and it will wash away the drudgery and usher in new, more-fulfilling roles.
My thesis here is “Let’s help people navigate this sea change of transitioning to the gig economy and coexisting productively with AI Agents effectively. Let’s pave that road and make it a clear path and if we do it right, some percentage of those travelers you help will pay it forward and participate on the impact side with Problemattic.” Once people have complete freedom and domain over their time, they should in theory gravitate to the impact projects for even more purpose and meaning.
This vision is 100% clear and inevitable to me in the way I imagine it must have been to Travis Kalanick and Brian Chesky once they saw clearly the latent value hiding in plain sight for Uber and AirBnb respectively. There is latent, untapped genius hiding amongst knowledge workers in this world that will emerge once we break the 4min mile here and demonstrate this vision of co-op-built startups is possible. It all hinges upon having the right mix of people in the first cohort and then supporting them steadily with the right environment consisting of inspiration, capital, tooling and mentoring resources to fast-track the path to a working product. Sprinkle in some Agentic AI accelerant sauce as we continue to refine our process and tooling and we should slowly-then-all-at-once arrive at a reliable engine for creating cashflowing ventures via this trustless, distributed approach. We will have at that point created an unprecedented, citizen-powered problem solving machine that can be harnessed in various ways and in so doing play a key role in facilitating this societal wealth unlock.
That is admittedly an ambitious vision. I see the destination. I see the path from reverse goal planning to sequence the dominoes in the order we need them to fall. Actually tipping them over one at a time is always the rub and that’s what I’m slowly chipping away it via running our events and building out our community and software.
Offloading the promotional burden
I realized somewhere late Q1 of 2024 that building Problemattic and running our events was a sisyphean task that was going to break my back if I tried to do it alone. I could easily schedule the next event series but the movement at that point wasn’t taking off under its own propulsion – it required an untenable level of grunt work promotional practices from me to make it go. I tried out a “street team” concept, various guerrilla promo tactics, ads, partnerships… it was an uphill slog. More importantly, we had maxed out the learnings from the events themselves so even if I executed the next one, there were only marginal gains to be had in terms of refinements to the event formula. I knew something needed to change at this point but wasn’t sure what the right move was.
About this time a chamipon of Problemattic within the HATCH community had made a suggestion to their board that they consider using our system to power their next Arcosanti conference. This was a very interesting idea that made me realize Problemattic the movement was separate from Problemattic the tech platform.
I went heads-down for the entire month of March refactoring all the Problemattic-specific aspects of the app that were hardcoded into the platform expressing to make stuff more generalizable and white-labelable. I also applied what I was learning with LangChain and n8n to build an AI Agent powered system for suggesting mini dream teams within the community with the goal of “not leaving serendipity to chance” by compiling detailed dossiers on everyone and using LLM’s to ask the right questions and connect the people who needed to meet each other. This capability (which was really just a bolt-on afterthought) became the killer feature that users actually wanted. I finished everything just in time to accommodate HATCH with a custom impact portal exclusively for their Arcosanti conference and unveiled it onstage.
With some more modifications after the conference I was able to implement custom domains and spun up sites for Polylogue, the Nocode Lisboa Meetup group I run and a new initiative I established called “Open Startup Community.” I rebranded the underlying tech system as Stone Soup and made the Problemattic community the first customer of that platform. I figured out how to implement this concept of Tribe-specific filter profiles allowing each community to project a specific view of the underlying data while retaining everything as multi-tenant and rolling up under the Stone Soup universe. People still found the process of creating custom dossiers for their members too confusing at this point so I implemented a chatbot-based UX that made it more intuitive which led to an uptick in signups and cracked that equation.
AI Builder tools will leapfrog and obsolesce visual Nocode platforms
This is maybe a bold prediction but I’ve played around enough with tools like Bolt.new, Replit, Cursor and V0 at this point to understand that these are not just force multipliers for amplifying professional developers, they are a democratizing force for lowering the barriers of who can write software.
I wrote our entire platform using Bubble.io and n8n. It’s an integrated system that delivers the best of Meetup, Kickstarter, Coinbase, Match.com and Upwork so I know intimately the capabilities of nocode tooling. I’m predicting this year that the AI builder tools referenced above surpass the visual drag-n-drop programming environments like Bubble/Webflow/Glide/Adalo in terms of speed of development and we see an exodus from those nocode platforms to traditional apps built with AI. I have nearly 3yrs invested in our Bubble-based tech stack but I’m contemplating rewriting the entire thing using Cursor/Claude. It’s not yet the most high-value move at this point but at some point the tradeoff will be worth it for the month it would take to recreate the entire platform as a progressive web application using these tools. For some short term dev pain and halt of forward feature progress, you get the extensibility, scalability, freedom of vendor lock-in, access to wider dev pool, etc all while retaining the agility that a nocode platform gives you for iterating (even improving speed of iteration frankly). And you overcome all the limitations that a nocode platform like Bubble has in terms of accepting constraints for the sake of convenience.
AI Builder tools are not just force multipliers for amplifying professional developers, they are a democratizing force for lowering the barriers of who can write software.
I highly encourage you to read this essay which is a hot take on Addy Osmani’s piece. It’s probably the best summary of all the dynamics in play here when it comes to AI replacing software engineering jobs and while they got most of it right, I disagree with their conclusions about the democratizing nature of this stuff. I’m using it now to build NextJS and Node.JS applications and I have no idea what I’m doing but if I can get to revenue and working software and then use that revenue to pay more qualified devs to refactor my initial version, then that’s pudding proof that I don’t need to be a professional software dev to use these tools to build product. It’s no doubt advantageous to have some familiarity with software design concepts and programming principles but at this point you can acquire enough understanding to operate as a product manager and have Claude play the role of CTO, Cursor be your lead dev and v0 be your designer. If this is doable now realize that today is the worst AI will ever be– it will only improve from here as the models become more capable and chain of thought reasoning improves.
This talk below I did at the last Nocode Lisboa meeting was the most-attended talk by a factor of 3 (we had 80 attendees who stayed the entire time). It’s a good overview of these tools where I do a “live cooking show” making an app from scratch in front of a room full of people:
Key Wins
I’ll break these out by realm here: platform, project success and business development:
Platform
- The biggest win was decoupling the community from the tech so that now any community can use this platform to align its members around outcomes.
- Configuring TypeBot as a chatbot-based onboarding flow was key for simplifying the process of completing one’s dossier that is used to make the smart matches with your dream team.
- In March I spent an entire week creating a pair of animated explainer videos to convey the grand vision for Problemattic succinctly and clearly to all the different stakeholder types. This was pre-whitelabel so I hadn’t yet conceived of the ideal of decoupling the movement from the underlying tech at this point but it explains the game theory dynamics well with using bounties and tokenized projects to attack the problems.
- Employing Agentic AI: the smart intros feature I built for the HATCH conference in March has been popular with people. You can see this talk I did for Nocode Lisboa that opens the kimono and shows this stuff works.
- November was the first month I was able to finally extricate myself from the development role and recruit external technical help to offload some of the dev burden – a huge productivity and psychological win. Bringing in outside help on the technical stuff frees me up to try and land some deals that should generate revenue, rinse repeat. Getting extricated to bring in the first person is always the hardest. We have all the systems in place at this point to do multi-person dev which is great.
- Other highlights include: coming up with an elegant “self-healing infrastructure hack” in Bubble and creating a way for members to directly chat with the vector store database to ask questions of the network.
Project Success
- Message Everywhere is scheduled to go live tomorrow. This is the culmination of two years of work and collaboration amongst 20+ ASU computer science interns and the ProNocoders team. It will be first deployed in Phoenix, Arizona for Creighton Community Foundation and their community of 15k constituents but soon after will be packaged up for distribution via the Bubble marketplace and made available to other non-profits and NGO’s around the world to use as a SMS-based communication platform for managing their communities.
- We have successfully landed a NIH grant for the Shepherd.AI project which promises to amplify the utility of the Message Everywhere platform by adding diagnostic and interventional AI-powered agents for getting better outcomes for the clientele of M.E.
Business Development
- We closed our first fully-funded bounty and are scrambling for a Feb 28th delivery date to provide the first end-to-end pilot of the bounty mechanics in action for aligning teams around a specific outcome. There’s still a lot to be done to figure out how to make this scaleable and repeatable but getting through the first cycle is the first step.
- I made a strategic investment in a company called Tribe who we’ve partnered with and will be running our events out of their spaces going forward.
- I spun up a new brand called “Open Startup Community” to address the demand I saw at our events for ability to build for-profit ventures via this same model.
- Polylogue let me spin up a member hub for that community in Lisbon giving another feeder into this network. The more feeders like these we have the more diverse a community we build bringing wide-ranging backgrounds and perspectives to bear on these issues.
- Lisbon Digital Nomads has agreed to be the petri dish for Lunch Mixer in testing this with their community of 2k+ nomads in their Slack. I’m working now to get that product across the finish line and implemented so we get yet another interesting feeder of talent into the ecosystem.
2025 Roadmap
- By now it should be obvious I’m insanely bullish on the prospect of Agentic AI this year, both from a software creation perspective as well as an invokable tool within SaaS applications to perform complex tasks. The way I architected things with projects and tasks lends itself really well to the possibility of using tools like OpenDevin to have AI agents knocking out well-defined dev tasks. We built Behalf.bot in 24hrs at Startup Weekend as a proof of concept of what’s possible with Agentic AI in short order. Expect us to leverage agents wherever possible to streamline things and free people up to do their best work.
- I’m intending to run a pair of monthly events (one for Nocode Lisboa and one for Open Startup) each month staggered by 2 weeks. NCL events will be these “live cooking sessions” like the one I did back in November and should be a fountain of app ideas that then get gestated via the OpenStartup incubator. I’m intending to record and potentially live stream the events as a way to proliferate these outside of the local Lisbon scene and be a global feeder to this talent pool. I’ve looked at Twitch, Twitter Spaces and YouTube Live streaming as options here.
- You can always find our public roadmap in Trello that shows the current focus. We have some cleanup work to do and then pruning back some of the various SaaS we’re paying for will be a priority (death by 1000 paper cuts).
- The strongest thing we can do at this point is to make a few of these OpenStartup initiatives work and make it possible for a few people to quit their jobs and live on sustainable revenue from their portfolio. This is the A#1 north star this year: making some of these OpenStartup ventures work. The intent is to get a nice cadence of monthly Nocode Lisboa meetings in which we demo the AI Builder tools live while building apps attendees want built. Then we throw those up on OpenStartup and use that for continuity of ongoing dev to get it deployed and working. Some will work, some will fail and all of it is an ever-evolving experiment and of the ones that work, hopefully some set of those folks will become active participants on the impact side with Problemattic. I’m hoping to partner up with a promising venture studio here in Lisbon this year to feed them qualified working MVP’s that they can then take and grow via their methods.
- Recruiting talent & capital to the platform will be #2 priority and becomes significantly easier once we generate some successes.
A shameless ask
If you have connections with folks at any of these entities below and are willing to make an intro, please reach out or leave a comment. I’ve identified these as incredibly promising partnership candidates where we’re well-aligned and able to help advance their cause by getting them into the Stone Soup universe. I’ve made it easy to quick check if you have contacts at each – just open each link and see if you have any contacts or extended connections there in LinkedIn:
- Rotary International
- Lions Clubs International
- Hult Prize Foundation
- Kauffman Foundation
- Techstars Startup Weekend
- Long Now Foundation
- Kevin Kelly
- Tony Robbins Foundation
- Peter Diamandis
- Xprize
- Junior Chamber International
- Kiwanis International
If you’ve read this far thanks for your interest in this journey. I wish you a productive 2025 and invite you to come into our world and get involved in whatever capacity you find most interesting. We have lots of ways to engage and it starts by telling us about yourself so do that here now.
Thanks for reading. Could you answer a few questions below? This is incredibly helpful for understanding my readers and helps me create better content. thanks so much. -Sean
