On GPT-4 self-improvement
On GPT-4 self-improvement, as also mentioned by Andrej Karpathy, we are not seeing max GPT-4 capability yet. Through “prompt design, tool use, meta cognition strategies (e.g., idea of attempt, critique, retry, capabilities model, etc.) this technology is very likely to go a long way”.
See also “AI Explained” video https://lnkd.in/dFgxNqr3 and the following recent papers as background.
✅ Reflexion: an autonomous agent with dynamic memory and self-reflection (20 Mar 2023)
https://lnkd.in/dHS76uYs
https://lnkd.in/dJiPHstp
✅ HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face (2 Apr 2023)
https://lnkd.in/dwb7KUJD
✅ DERA: Enhancing Large Language Model Completions with Dialog-Enabled Resolving Agents (30 Mar 2023)
https://lnkd.in/devvDs_2
✅ Language Models can Solve Computer Tasks (30 Mar 2023)
https://lnkd.in/dDED_ikv
✅ Language Models can Solve Computer Tasks (30 Mar 2023)
https://lnkd.in/dDED_ikv
✅ TaskMatrix.AI: Completing Tasks by Connecting Foundation Models with Millions of APIs (29 Mar 2023)
https://lnkd.in/dJc_Wndw
✅ Sparks of Artificial General Intelligence: Early experiments with GPT-4 (27 Mar 2023)
https://lnkd.in/dZTrY2ay
✅ GPT-4 Technical Report (27 Mar 2023)
https://lnkd.in/dmSEHesx
✅ Alpaca: A Strong, Replicable Instruction-Following Model (13 Mar 2023)
https://lnkd.in/dC66CjwM
✅ LARGE LANGUAGE MODELS CAN SELF-IMPROVE (25 Oct 2022)
https://lnkd.in/dNVbQCGC
✅ ALFWORLD: ALIGNING TEXT AND EMBODIED ENVIRONMENTS FOR INTERACTIVE LEARNING (14 Mar 2021)
https://lnkd.in/dp6SjPuj
NVIDIA shows new research on using AI to improve chip designs
https://lnkd.in/dN2ZFRA5
Google C.E.O. Sundar Pichai on the A.I. Moment: ‘You Will See Us Be Bold’
https://lnkd.in/dgWBd56m
🌍 The Power of Generative AI: Exploring its Impact, Applications, Limitations, and Future
https://lnkd.in/dp2e6c5P
https://lnkd.in/dMRdCJv4
🌍 “Democratizing Artificial Intelligence to Benefit Everyone: Shaping a Better Future in the Smart Technology Era” jacquesludik.com
OpenAI, SwissCognitive, World-Leading AI Network, Machine Intelligence Institute of Africa | MIIA, Cortex Logic, Cortex Group, DSNai – Data Science Nigeria, AICE Africa, Google, Meta, Meta AI, DeepMind, Stanford Institute for Human-Centered Artificial Intelligence (HAI), OECD.AI, World Economic Forum, VERSES, University of Washington, Microsoft, Carnegie Mellon University, Microsoft Research, Curai Health, UC Irvine, Zhejiang University, University of Illinois Urbana-Champaign
#artificialintelligence#technology#datascience#africa#network
#ai#design#research#learning#google#intelligence#language#video#cognition#cognition#microsoft#university#health
