The big news is Creighton successfully landed a NIH grant for the Shepherd AI project this past month. Given that the precursor for that project is Message Everywhere I’ve dropped everything to focus on getting that one across the finish line. We’re very close ironing out some final bugs and putting the wrapping paper on it to be able to make it available for others to use. My goal this time next month is to unveil this project and put v1 out there for others to start using. This is where things currently stand:
Other Wins
I was able to get CoAlias running pretty easily giving us full white-label capability with custom domains for anyone wanting to use the Stone Soup Protocol to power their own instance while participating in the larger universe of talent/capital. You can see a proof of concept in action here: https://openstartup.community/
On the Message Everywhere front I’ve successfully switched us over to using Cloudflare as the intermediary router for sending and receiving amongst Twilio and Telnyx messaging services. It’s mind-blowing how well ChatGPT does as a copilot for creating the javascript function which now runs serveless as a worker to facilitate this. There are actually two independent workers (one for send and one for receive) and now instead of Bubble communicating directly with the SMS platforms, we simply interact with the CF workers and it handles the complexity of interfacing with the various systems. As a part of this migration I was able to also strip out our dependency on a deprecated Twilio funciton for handling the broadcast functionality so that now CF handles that and logs message status per individual number so we end up with a success/failure list for each broadcast in our Bubble app.
The other focus for me this last month was shifting my Sales Safari effort towards a space that I think is exploding. There is a whole frontier developing around coding apps entirely using the help of AI tools like v0, Replit, Cursor & Vercel. I’m taking the plunge down this rabbit hole because even if it’s not quite there yet today, the models are improving exponentially and it’s a matter of time before it’s possible to code and and deploy a production-grade app start to finish entirely using AI. I want to be positioned to capitalize on that before it’s mainstream. This is consistent with my focus on Nocode in terms of democratizing access of developing applications to the masses and I think it could be a massive unlock of prosperity once anyone can build and deploy software this easily. I don’t know where it’s going but this makes it a perfect candidate for ongoing Sales Safari. If you’re interested in this realm check out this video as a good example for what’s possible with these tools:
What I’m Thinking
-
I have (for now) suspended trying to sell the Stone Soup Protocol for sake of zooming in on Message Everywhere as the most high-leverage activity for getting a PR-worthy win we can celebrate and point to as a case study for this whole effort. Message Everywhere is incubated by Problemattic which runs on the Stone Soup Protocol.
-
It’s a bit like “death by 1000 paper cuts” with all these little $10-20/mo SaaS services I’m now paying for. At some point it’s worth it to do a consolidation effort to cut costs. For instance Supabase could be easily swapped out with Qdrant running on our Google Cloud instance and should create little or no extra expense. This Discourse forums instance that currently runs via Digital Ocean could likewise be ported to GCP saving that cost. And Dropbox is unnecessary with Google Drive and could be eliminated. Doing all that work however doesn’t really advance the ball in a meaningful way so that’s a can I foresee kicking down the road until it gets annoying enough where I bite the bullet and migrate everything.
-
Super excited about this tool I discovered last week QuizGecko. I’m figuring out the ultimate setup for building a 2nd brain with Obsidian and using tools like Anki for spaced repetition, Readwise for organizing and logging what I read, ReMarkable2 tablet integration, Snipd for podcasts, etc. The missing piece has been the ability to generate self-quizzes from content and that’s exactly what QuizGecko does. There is some product to be had unifying this all into a simple, sensible system that helps you learn faster, rehearse content in smart schedule and consequently encode/retain information better. I’m playing around with having QuizGecko index the library of Kindle books I’ve read and create quizzes/flashcards for all the main concepts. AI is fun.
Open Questions
-
The wisdom of diverting Sales Safari attention to the AI builder crowd from the original intent (conference organizers, intentional community leaders, CSR dept heads, hackathon organizers)? The combo of there being few good watering holes for the latter groups and the exploding potential of immersing with the former is fueling my decision but I can also see the argument for “stay the course and focus on the target customer.” Time will tell on this one…
-
How reproducible this current situation is with Message Everywhere (ie. the validity of this as a test of the platform). We’re lucky to have a relationship via the ASU EPICS Pro program through Creighton as a fountain of qualified intern labor but that’s a trickle, not a geyser at this point so unclear whether this extrapolates to the other projects on the platform. We also have a highly-competent, highly-motivated Director-level person driving requirements of the project which is unique (unlike other projects which have been largely student-submitted and hackathon participant-submitted). I wrestle with how to jump the chasm to getting mainstream interest and participation in this whole effort. Ultimately I’m under no naive illusion that altruism can be trusted at scale- it will be a matter of aligning incentives and reaching the right pockets of people for whom participation in this platform solves a need whether that be impact, volunteer requirement, capital contribution, greenwashing, etc. Lining up those dominoes though remains the million dollar question… If you happen to have a contact to a wealthy philanthropic individual who sees the vision in what we are doing with this whole effort, please connect me and let’s get them as the anchor bounty funder to pilot that aspect.
Next Month’s Focus
Platform
No major development planned. If time permits I’m going to look at adding some type of sensible messaging capability but that’s actually a fairly involved project to do it right so not committing to that just yet.
Project Success
100% Message Everywhere is the sole focus. We are going to ship this thing one way or another by this time next month. We are supplementing with AI and Upwork labor but this piece of software is probably 97% built with donated community labor and destined to be open source.
Events
We are technically powering the participant matching for Cofounder’s Hive conference this weekend but I’m getting very little reciprocation from the founder on our promo arrangement so I’ve relegated this one to the bottom of the stack.
Content
I did a post last month on how to create modular workflows using the top 3 pipeline automation tools (Make, Zapier and N8n). That got some good feedback in the various places where I shared it. It’s a crucial technique I’ve come to rely upon in how I architect things with those tools. I have no podcasts scheduled for this month for sake of keeping all attention on M.E.
Traffic Snapshot
The technical roadmap as always can be found here.