Malg_TheViking

5,540% increase in price per token average over the last 52 tokens filtered!!! Technical Report 🧵on our DegenScope (PFB) Model

One year ago an experiment sparked a Dither. Today we launch our transformer based forecasting engine which will be free in the Dither Lounge until the end of the year

We have finished backtesting our new model. We present preliminary results on optimization, validation and test sets. We also include a couple of important statistical tests. Results: BULLISH

Unicorns can now be quickly shared in our web app dith.ai/vibecheck.html

We'll be re-arranging DITHcember a bit due to feedback from our telegram. But we want to address some things head on: Dither has an opportunity to begin turning a profit going into the new year. Right now we are "profitable" but team members are volunteer (including @0xMattness) We already had a decent pump months ago. Then we had a ton of churn because frankly - our tools were not good enough. They still aren't but will be sooner than most think. So any perceived aversion to huge marketing pushes hasn't been "build and they will come" but closer to "Let's not fucking churn 95%+ of people that come through the door" So if you were previously calling for marketing - what product do you think is good enough right now? We get the speculation angle. This is where our Lindy effect hurts us. We have been saying the same thing for 10 months. That saturates our message until we ship our newest releases. We could push to try and get in front of even more people but then they either 1. Churn or 2. See we still haven't delivered. That means we dump which for some people is fine because they want marketing for an exit. Dither isn't building for an exit. This all kind of sounds negative BUT the team is the most locked in and focused we have been all year. We have a chance to push profitability into the new year. We have the best model that ONLY needs some engineering optimizations. We are scheming a grassroots marketing push. The future is bright. We're always here for these conversations. We're not afraid of being wrong or FUD. If you read this message and believe the project is way off base then jump in our telegram and tell us! If you can outline anything that can be improved or show us where we're failing, we fucking love to hear it. DITHcember is to push updates and get us ready to blow people's minds.