Andrej Karpathy: "AI Is Still Slop"
Échec de l'ajout au panier.
Échec de l'ajout à la liste d'envies.
Échec de la suppression de la liste d’envies.
Échec du suivi du balado
Ne plus suivre le balado a échoué
-
Narrateur(s):
-
Auteur(s):
À propos de cet audio
The evolution of AI agents over the next decade is expected to be transformative, moving past current limitations such as continual learning, although achieving true Artificial General Intelligence (AGI) is estimated to take approximately a decade (circa 2035).
Andrej Karpathy describes this period as the decade of agents, suggesting that while early agents (like Claude and Codex) are impressive, significant work remains to be done to make them function reliably as an employee or intern.
I have this older (actually veery old) related video about Andrej Karpathy:
Ep. 21: AI Discusses Andrej Karpathy's Twitter
https://youtu.be/yGBqnHTK0cM
Addressing Continual Learning and Memory Limitations
One of the most critical current limitations highlighted is the lack of continual learning. Current LLMs typically restart from scratch in each session, as they lack the ability to assimilate long-term information efficiently into their core structure.
#andrejkarpathy #artificialintelligence #agi #ai #claude #chatgpt
___
What do you think?
PS, make sure to follow my:
Main channel: https://www.youtube.com/@swetlanaAI
Music channel: https://www.youtube.com/@Swetlana-AI-Music
Hosted on Acast. See acast.com/privacy for more information.