The idea
It's about finding optimal approaches to problems.
Someone expresses a prompt ("Improving in chess for one month one hour a day") and must also come up with a metric ("Rating on chess.com at the end").
Other people can then submit approaches, so what the person should do.
Every approach if followed will on average result in some rating.
We'll use prediction markets to determine this average.
The best approaches then get rewarded based on the problem demand which can be boosted by others.
Resolution
If I decide work on this for 3 months full-time it resolves YES iff it has >100 daily active users at the end.
If I don't decide to work on this resolves N/A.
About me
I have received a bronze medal at the IOI and have been programming all sorts of things since I was 13.
Besides that I've just finished high school and this is my first attempt at a startup.
Links
The current implementation of the platform is accessible here: https://approdict.com
You can read more about the idea and why I think it has potential here: https://medium.com/@patrik.cihal1234/approdict-finding-optimal-approaches-to-problems-e06491785134
Other examples
Prompt: Increase happiness levels in one week
Metric: Average of daily subjective evaluation from 1 to 10 for that one week
Prompt: Increase sleep quality in one month by allocating only 9 hours a day for sleep
Metric: Average sleep quality as measured by Whoop
Prompt: Working as a software engineer full-time for one year without college degree
Metric: Total net-worth from the job at the end
[redacted, I didn't understand the idea here correctly]
@YonatanCale Thanks! But no I did that manually this time... Also regarding the first message, since you redacted it, I'll at least respond to why I'm not selling the idea more: 1. I'm working on something else now, which, I think, has larger potential. 2. This idea seems very hard to attract people to, as you need an active community for it to be valuable, but active community needs it to be valuable, therefore I think this is something which requires significant capital because it won't grow by itself... Which is not something I have right now.
@YonatanCale I was thinking of reading Lean startup, and have even read few pages, but you've convinced me to read it more.
@patrik +1 to working on whatever you find fun!
Re getting a community: Yeah that's a big problem for many projects, but I'd also say:
Capital (marketing) isn't imo a good solution (to start with). Doing something-like-marketing to many users before you have a few excited users ( = product market fit, sort of) will probably lead to people trying the product and quitting quickly
You can skip the "get a community" by using manifold, as I see you started doing (very nice! an amazing solution to a very hard problem)
More generally, I think attracting people to your project at all will be hard if you don't talk to users a lot (with the goal of figuring out your blind spots around some stuff which I think is described pretty well in "the lean startup")
I'm taking your positive comment as an indication that such opinionated comments of mine are welcome so I'll make a bit more of them
@traders I created few experimental questions here on Manifold that attempt to replicate the functionality of my platform. You can check them out here:
@mariopasquato Yeah kinda. My platform would be more focused on optimization and also reward good contributions dynamically based on their actual expected value. On wikihow you basically get one random approach that seems good, and may have worked for someone, and there aren't that strong incentives to optimize it further.
Here anyone can just come, submit an insightful approach to an already existing problem and get immediately rewarded for it.
@patrik Are you planning to make your own prediction market platform or to use manifold? Do you think you could automate translating your couple (question, metric) into a new market to be posted on manifold?
@mariopasquato I already have a functional platform with betting and everything.
I guess there is a way. Maybe multiple option questions where everyone has to denormalize the probability in their head into appropriate range and each option could just be referring to a comment since I don't think you can write enough text there.
I'd also have to distribute rewards somehow manually.
But maybe I could somehow interconnect my platform with manifold, such that when someone creates approach on Approdict it creates an option with a comment here on Manifold.
Maybe you could save a lot of time building out potentially wrong features by instead running a prototype on top of Manifold? You'll have a baked in audience (with the needed network effects), and you would learn a ton about what your potential users would be looking for, so you can be much more confident that you're building something people want.
I definitely think there is a product for Personal Goals <> Prediction Markets, and have built a bit in the space, so I'm happy to share more thoughts if you're curious.
You'll learn a ton working on anything, so it's hard to discourage anyone from working on anything without comparing it to the other ideas they're excited about. But as you're probably a fairly technical minded person, my guess is the best learning will come not from building a product but from trying to grow something that's useful to a lot of people (which might require less building than you think).
From firsthand experience:
"First time founders are obsessed with product, second time founders are obsessed with distribution."
If you can get yourself to start thinking more like a second time founder, you can save a ton of time!
Good luck!