Hi CoE,
I’ve been familiarising myself with some of the activity I wasn’t aware of on the CoE forum and checking out the current state of the three main CoE initiatives: NumerBay, Newsletter and Meetups. Some initial thoughts on these initiatives I thought I’d share:
NumerBay:
The Good:
- This is super cool and a great example of the kind of project I think the CoE should be sponsoring.
- A great way for good modellers with no NMR to earn some initial NMR, is to sell their models and at the same time allow normies with no datascience knowledge to join Numerai and start staking
The Bad:
-
Unfortunately, it doesn’t look like its being actively maintained nor does there appear to be any effort put into growing it e.g.
- the OnlyFams page has links to Shirts and Hoodies and Souvenir coins that are blank pages
- looking at the Community Apps page:
- Numerai Payouts Dashboard: 503 Service Temporarily Unavailable
- Numerai Insights Dashboard: This site can’t be reached www.numerai-insights.com’s DNS address could not be found. Diagnosing the problem.
- Power BI Numerai Dashboard: This site can’t be reached qe-intelligence.com took too long to respond.
- Numerai Streamlit Dashboard: Error running app.
- Numerai Dune Dashboard: Broken. The provided execution request is invalid: The query is on an deprecated query engine
- Numerai Earnings Dashboard: Application Error
- OpenSignals: Repository mostly redundant now with signals tournament data v1 and v2 coming
- NumerBlox: Seems to be the only true community project that is running
-
Existing incentives for data-scientists to sell their models is weak (prices are low relative to the risk of getting diluted in payouts).
Suggestions:
- Can we reboot the website to remove all the broken links and maybe put some merch on there for sale?
- Can NumerBay include Numerai Crypto models?
- Can we develop more of an incentive for data-scientists to offer their models by using an affiliate style reward programme where data-scientists can share links on social media with a easy user onboarding flow to staking on their models and then have their personal payout factor boosted by the number of new participants that are staking on their models. This would also encourage the community to be more active on social media.
Newsletter:
A newsletter is great to have and has been done well in the past. The only problem is the last newsletter was end of May 2024 and content is limited to the little Numerai specific news that has occurred recently.
Perhaps the Newsletter could also include more substantive content e.g.: a curated list of links to interesting academic papers or empirical research in machine learning and finance that is potentially relevant to the tournaments. Perhaps also an incentivised Quant Club of sorts with bounties for the community to collect for providing a tutorial or review of some interesting and relevant technique they’re using to construct their models.
Meetups and community outreach:
Always great idea to have irl meetups and it looks like CoE has been doing a great job of organising these. My feeling is irl meetups are only useful or a value add to the community at a certain scale e.g. making them side events of a big crypto conference might be a better use of resources.
Generally, my feeling is CoE should be actively aiming to drive user growth and it would be nice to see more of a top down active request for proposals driven by CoE for projects that aim to deliver on this goal rather than just have community proposals for funding on an arbitrary set of projects.
For example, publishing bounties for projects in specific categories like data and tools to bring more normies on board (which could also be CoE organised competitions or hackathons). The CrunchDAO Pi project is an awesome example of creating another funnel for non-technical folks to contribute and at the same time give the project an edge over others.
Critically, there needs to be an incentive for the CoE to do the heavy lifting to drive user growth and so perhaps the community should reconsider the poll on remuneration for CoE elders. Perhaps it could be performance based rather than a salary?
Appreciate you considering these thoughts.
Thanks,
Efficient Meerkat