Part 3/12:
The advent of advanced language models like GPT-4 and GPT-3.5 turbo sparked his interest in refining inference hyperparameter tuning and optimizing prompt strategies. Recognizing the transformative potential of agents working in concert, Chi shifted focus towards building intelligent multi-agent systems capable of tackling complex tasks—culminating in the development of Autogen.
The Genesis of Autogen: From AutoML to Multi-Agent Collaboration
Chi explains that Autogen emerged from a desire to harness the full power of massive language models in creating scalable, automated AI applications. Initially driven by opportunities to automate model selection and inferencing, the project evolved in response to the capabilities unveiled by ChatGPT and GPT-4.