So, fund every content proposal until, when, how? There is going to be a bottom to the well. There has to be some consideration by the CoE of what gets funded and for how long. Blasting all the treasury out with no thought isn’t much of a financial plan. We still have no idea what Numerai’s long term participation is here.
1 Like
@liz any chance you’d follow @arbitrage lead with a poll so CoE can gauge interest from the community?
1 Like
Obviously it doesn’t make sense to fund every proposal and no one has suggested anything like that. Let’s cut the nonsense arguments. If CoE is concerned about funds they should communicate that instead of doing nothing
2 Likes
You’re a Numerai employee, right?
If the company is unhappy with how the CoE is proceeding, they can certainly jump in and directly fund any project they like. If you think @liz proposal needs immediate funding, by all means, take it to @richai and see if it can get fast-tracked internally.
For the record I’m not arguing, you sure as shit would know it if I was. I have no dog in this fight. I’m not part of the CoE and I have no proposal on the table. I do have 30 years of business experience and I’ve been in the competition since January of 2016 and I know stuff takes time to get moving and the team knows it as well, i.e. Stake Management. I think we’re all on board with the concept of rapid development but this is brand new. Let’s take a deep breath and let these folks get their processes in place before we start crying.
Going forward I’ll stay out of the CoE discussion so I don’t further waste anyone’s time.
1 Like
Yep I’m also gonna step out. Last thing I’ll say is Signals growth super important 
3 Likes
Defiantly was never saying to fund every project @objectscience but at the same time we cant just keep letting stuff go dead in the water either. Mostly saying it needs to get a bit more streamlined eg my post above “We should be 1) making proposals 2) awaiting feedback from CoE members and community 3) make final edits 4) the person who made the proposal start a community vote 5) CoE votes on community passed votes to make final say 6) get paid 7)Numerai gets more outside exposure/ community stays tight.” Also why I stated that the CoE will still be able to filter projects even if they pass through community with overall yes votes. If we had a certain member of the CoE that could facilitate better flow eg, commenting on edits to proposals, creating CoE meetings to discuss current proposals and getting proposals to a final community vote to gauge interest, things will move smoother and get to the CoE vote faster.
4 Likes
@aventurine I rethought this last night, and I’m jumping on the other side of the fence. I have a pretty concrete bias based on my own experience, but in hindsight, it doesn’t really fit here. I think right now the entire process should be condensed and streamlined to speed up approval/rejection. At that point, we can see how Numerai responds and the CoE can add filters, checks, and balances once the initial process has played out.
4 Likes
FUNDED 7/21/2021 TX: https://etherscan.io/tx/0xde131addf5889076427111348094c7e1fdfcc0e3bd0c9eb22a435265961292d9
7 Likes
THANK YOU, thank you, thank you!
5 Likes
Glad we could get this proposal through. @liz I hope you also take some of the feedback into account, maybe publishing what can be the canonical R example script as you complete it after the livestreams? Maybe also doing a shorter walkthrough video of that example script? My two cents at least
4 Likes
I like those ideas! I’ll definitely keep them in mind as I move forward
2 Likes
Some other thoughts I’ve had and wanted to add since this proposal was funded:
Despite calls to the contrary, I think it’s incredibly important that we converge on some standardized value and have concern for the overall solvency of the CoE. To date, I think we’ve already been funding at an hourly rate that is probably too high. I would suggest we aim for somewhere in the $40 - $60 per hour for content (paid post-hoc) with the NMR payments marked to market at the time of payment.
In addition to the fact that I think we need more Signals development, I was one of the four that voted/approved this initial funding transaction simply to get things moving and people motivated. That said, I think we need to be more scrutinous on the rates that people are proposing because it’s simply not sustainable (or wise) to approve all proposals as soon as they come in at whatever price is initially suggested. So, I hope if we continue to fund content proposals in the future, including this one, we would use a standardized rate and pay out post-hoc.
5 Likes
youtube links from stream # 1
stream summary from twitch email (stats)
this last one was split up because the stream crashed after an hour.
next stream Monday 09:30AM - 01:30PM Mountain time Twitch
1 Like
google drive updated : lizDoesIt - Google Drive
github updated : GitHub - completelyAbsorbed/lizDoesIt: code and more from twitch.tv/lizDoesIt and beyond
1 Like
github updated GitHub - completelyAbsorbed/lizDoesIt: code and more from twitch.tv/lizDoesIt and beyond
started youtube export, will link
This Monday I did a half-stream (> 2 hours) as I was losing my focus, and getting deep into debugging. Debugging is important, and I covered a bit of important info, but this was a good time to jump because testing the fixes requires long wait times. Given the partial stream, I consider the promised streams to be 2.5 / 4.0 complete. I will attempt my next stream this coming Monday around the same time I’ve been going. All remaining streams will be 2 hours or longer, and counted as I described above (so, in total, I have 3 development streams remaining, of 2 hours each). I firmly believe this is a better value than the 4 hour streams, but if the council has issue with this, please reach out to me directly.
Great progress is being made and I expect to be able to provide the following cleaned up R scripts : scraping yahoo for pricing/volume data, example model training script, example model prediction script.
Additionally, I am planning, as a bonus gift to the Numerai R community, to produce and release a short walk-through video of my scripts resulting from this project. Given that this is outside the scope of the funded proposal, and relies on the finished output of said proposal, this will be delivered after the conclusion of the project.
Here is the most recent Twitch stream summary, and link to the youtube export.
2 Likes
