X @Herbert Ong

RT phil beisel (@pbeisel)FSD Version 14Maybe this week we get version 14.0, as foretold by the profit [sic].Key/expected features:1. Version 14 10x larger model than 13, trained on more data and with many reinforcement learning (RL) use cases (see article).2. Version 14 is multi-mode: it can operate in supervised, unsupervised, and unsupervised Robotaxi modes, all software-controlled*.3. Version 14.1 incorporates more fine-tuning based on an expanded range of RL use cases. Use cases likely fielded from Robo ...