Augmenting GTP-3 with Additional Information Prompts New Coding Businesses

IBL News | New York

Customer support Q&A AI chatbots based on ChatGPT are becoming an unexpectedly profitable business for developers.

The GTP models have picked up a lot of general knowledge in training but personalized chatbots often need to ingest and use a large library of more specific information taken from Wikipedia or other sources, as OpenAI explains in a notebook in GitHub.

The company that has created ChatGPT is providing docs for fine-tuning these new virtual agents, along with a myriad of bloggers like machine learning engineer Peter Foy.

These custom bots are being offered as a service. They can offer multiple applications, such as investment research (crypto, IPO, earning calls…) assistants, legal or medical Q&A bots that answer users’ questions, along with a code documentation bot that provides answers to developer questions.

Experts quote Sam Altman, OpenAI’s CEO, when he said that there will likely only be a few companies that have the budget to build and manage Large Language Models (LLMs) like GPT-3, but there will be many “layer two” service companies in the business of fine-training these models in domain-specific industries.

“The fundamental interface will be natural language models,” he said. “They will be commoditized” and “economical impact will be huge.”