Technologique
729 subscribers
143 photos
3 videos
42 files
905 links
Deeply involved developers about various aspects, tendencies & conceptions of programming technologies, FLOSS, Linux, security, cloud infrastructures & DevOps practices, distributed systems, data warehousing & analysis, DL/ML, web3, etc.
Author: @andrcmdr
Download Telegram
One of the best explanation of the current AI space state (a state of the art) and upcoming (or already current) breakthrough in improvement of LLM with planning and self-learning (self-improvement, as an AlfaGo does) and (maybe) forming synthetic data sets by itself and for itself. Crazy stuff!!! Highly recommend to watch 'till the very end.

https://www.youtube.com/watch?v=Z6E41eXStsU

#AI
#AGI
#LLM
#Qstar
And the mentioned lecture - an intro to LLMs by Andrej Karpathy.

https://www.youtube.com/watch?v=zjkBMFhNj_g

At the time of 38:02 the self-improvement extension to LLMs explained. It can be done by the automation of gathering and preparing synthetic data sets by the model itself (or agents) and by the usage of output of one model as an input for another model (self-improvement pipelining).

#AI
#AGI
#LLM
#Qstar
TikTok Video Downloader: Download TikTok Videos without Watermark