I am working on a LEGO-like News Feed that allows you to combine different AI models and Filter/Ranking Blocks to create your own For You feeds. You can take control of the algorithm and iteratively train your algorithm on the data you specify. No more black box feeds that try to optimize you into mindless scrolling.
If you have diverse interests (say Tech vs Games vs Traveling), they can each be contained within their own feeds and your recommendation signals will not be polluted between the different feeds.
Agreed. I think dislikes are so important for tutorial / seminar / presentation videos in sieving out the good from the bad. The part about positive comments being upranked is also adversarial for learning.
I think you should open source the core part of the HTML visualization algos and if ppl like it, they may consider paying for the premium version. I don’t feel like ppl want to move their analysis workflow to yet another platform without trying it enough in their existing workflows (e.g Jupyter/Colab/Databricks notebooks)
Thanks so much for writing this article and pointing readers to the resources. I am planning to build my own personal search engine to reduce SEO spam in my searches, and the resources and concepts here were very helpful.
Regarding DNN-based semantic search, what would you say are the top notable works that one should dive into? What search experiences have you built that you feel was strongly enhanced by DNN-based methods?
If you have diverse interests (say Tech vs Games vs Traveling), they can each be contained within their own feeds and your recommendation signals will not be polluted between the different feeds.
https://personamo.com