Update 4

I wonder if this competition will just become a reminder to me and Patrick that we need to be more structured about our free time in order to finish side projects and still balance everything else we have going on in life.

To be fair, we both have full-time work that we need to maintain, and it’s the holidays so both of us have been gone at different points. Time restraints aside, I’m still curious what could have been done differently. Every time either of us did anything for the GPSO, it did seem largely to be either ideating or doing very small dev tests. Potentially it would have been helpful to have some “checkpoints” throughout the course of the competition, like:

1. End of week 2: Have a landing page
2. End of week 4: Have the ability to accept credit card payment
3. End of week 6: etc

It couldn’t be overly specific, but it would ensure that either progress is being made or you forfeit the race early.

As an aside, one positive outcome is that I’ve been coming up with a lot more project ideas that I believe could provide real value to other people. A new idea I thought of would be to automatically create change logs from git repos. We could ingest the diffs within a given time range, and create a summary based on just the changes. We wouldn’t need to pass the context of the entire codebase, but probably we’d still need to provide some base context about what the software is doing.

Update 3 (I guess)

I’m finally back in one place and have no plans on leaving for the rest of the year, so now we can actually make Ship-Off progress.

I’ve narrowed down to 2 solid ideas that I would pay for if somebody else made them:

• the digital shadow
  • I still believe there’s real value here that doesn’t exist in current time-tracking solutions, because an LLM can “reason” with you about whether what you’re currently doing is actually relevant to your life, and then it can “explore” what causes those sidetracks
  • there may be a good amount of psychology involved in the prompt engineering
  • I think this is a great product that I personally want to use, but I think it may be the harder of the 2 to get paying customers for early on
• the AI debugger
  • plugs in to a CRM, has context of the codebase, and creates PRs based on bugs that the customers find
  • this one is very clearly saving users time and in a case like mine where I have software products that I’m very hands-off with, it makes it that much easier to let the business run itself

By next weekend the plan is to have chosen an idea and have the basic codebase set up, and then there are 2 more weekends left in the month. Not a ton of time, but it’s enough time for me to build an MVP.

P.S. I saw a comment on Hacker News that was a nice reminder that the GPSO really is putting pressure on us to build a product in the most efficient way possible:

always build the minimal thing that can be “sold”. Use AI to build the dirtiest MVP as fast as possible. Even better if your “MVP” is a deck and you can get people to put money down to wait. Figure out your GTM and messaging with that deck. If you are an engineer, you must resist that urge to build until you’re sure you can find enough people that have this problem and want to pay you to solve it.

Update 2

I’m making an update now for last weekend and this weekend, and after that I’ll be gone for 2 weekends so the next update will be in December.

I’ve been thinking about software that I’ve already built and how I could adapt them to be viable products. One such app is an “intelligent” task manager. There’s already a boat load (on theme) of task management solutions, But the app I made is just one list, and each item in the list has a checkbox, the task name, and a field for time required.

What it already does

• Allows you to “start” a task and decreases the time required for that task
• Allows you to set an end time for your workday, and optimizes the task list to get the most high priority tasks done that will still fit in the remaining time

What it could do

• compare new tasks to existing tasks
  • This could be comparing to the exact same task that was completed before, or doing some sentiment analysis to compare similar tasks
  • How many days did it take before that task was completed?
  • What was the time required for that task? Did that time required get adjusted?
• over time, try to point out patterns in time estimation and task delaying
  • What if fitness related tasks often take 15 minutes more than what is usually estimated?
  • What if writing tasks often get pushed off forever, whereas music tasks get done immediately? Is there a way to work the former into the latter somehow?

There’s also the reality that there isn’t much time until Dec 31st, so even 2 products might be pushing it but that’s my goal. They are going to be MVPs, but they’ll work (hopefully). The immediate goal is picking one idea that is technically simple to implement and provides enough value that I would pay for it if someone else made it.

Update 1

Almost immediately after writing my last post, I found out that one of my ideas was just launched (the same week) by someone else.

So I could still create an audio tool that does something similar for removing elements from audio, which I think is still useful and may actually produce cleaner results (akin to subtractive synthesis). That being said, this one is a bigger undertaking, and I only have 2 months where most of this GPSO work will happen exclusively on the weekends, so this idea is going in the trash.

That being said, I also have a new idea which is to steal one of Pat’s ideas and make it better. He mentioned he was working on something that would take existing customer support conversations and generate help documentation from them based on the answers provided. This got me thinking about LLM-based customer support tools, and then I started thinking about what kind of tools would actually be useful for my software business. Which led me to:

• a tool for SaaS products that takes in customer support requests and additionally takes in the codebase as context to make changes in response to the customers
  • if the customer reports a bug, the code would be updated to fix the bug
  • if the customer has a question how something works, the code would be read to understand what the customer is talking about
  • if the customer’s message appears to be a new problem by an existing customer, we could look at recent commits to focus in on specific changes

Pat also mentioned he’s been using Cursor to build everything, which made me think a bit more about how to speed up my development for this. Like I said, I have a limited amount of hours to actually dedicate to these before the end of the year. Some of the logic I need can’t be done cleanly by LLMs, but it is worthwhile delegating more tasks to my coding LLM tool (Supermaven).

generated with Google Gemini

I’ve been wanting to ship more software projects, mainly to get more reps in to decrease the time between ideation and distribution. That being said, I need some kind of pressure to actually work on those projects or else I’ll never finish them.

From now until December 31st, 2024 ( days from now), I’m challenging myself to ship as many profitable projects as possible. There’s no guarantee that the projects will make money, but the idea is that each one should be able to make profit. This focuses the ideas a bit more and helps avoid working on projects indefinitely with no clear direction.

To help with accountability and add an element of competition, my friend Patty Cake (he asked us to call him this) will be joining me. He just moved to Vancouver and I’m in San Diego, hence “Pacific”.

Rules

1. Products can’t be free. There has to be a way for each project to accept payments.
2. Each week we need to post an update on our blogs, even if it’s “I didn’t do anything this week”.
3. The deadline for building is the end of the year. Then we let the products generate revenue for the month of January. The winner is whoever generates more revenue within that month.
4. The loser’s punishment (and the winner’s reward) is that the user has to pay for all of the winner’s products for a month.

Potential Ideas

I have a few rough ideas so far, which I need to think about more to figure out the ratios of how much someone would be willing to pay vs how much time it will take to build. Some things I’ve been thinking about:

• a digital “shadow” that keeps track of what you do on your laptop and then provides insights at the start/end of each day
  • I have a Firefox extension that’s a very rough MVP, and keeping it as a browser extension might be the easiest way to get to V1. That being said, I don’t know if people actually pay for browser extensions.
• an audio tool to add or remove sounds to audio based on text input
  • It would probably be best for V1 to just pick a direction, either adding sounds (so being able to type out “snare drum with snappy transient”) or remove (“remove the snare drum”) sounds from audio samples
  • This idea seems very complex, but I know that there are people who would pay a good amount for this
• a Google Sheets app that enables listeners in spreadsheets
  • The idea is that you could supercharge your spreadsheets by adding a “listener” to a range of cells. If those cells change, or have a certain value, then some other action takes place. Basically like IFTTT for Google Sheets.
  • This is by far the easiest to implement because I already do this (but would still need to build a UI and more features)