In Episode 64 of the Enterprise AI Minute, Aaron explains the growing investment major companies are making in their own large language models (LLMs).
00:00 — Aaron opens with a preview of the Cloud Wars Expo on-demand pass, available now.
00:39 — Tech-speak is littered with acronyms. AI (artificial intelligence) and natural-language processing (NLP) have been around a while, but LLM may be lesser-known. LLM, or large language models because they work off of huge language datasets, are used as a foundation for NLP and AI.
01:45 — Massive LLMs are surfacing among companies like Microsoft, OpenAI, and Google Cloud.
02:04 — NVIDIA jumped into the LLM pool last year by introducing “NeMo Megatron” to help scale and optimize the GPUs tasked with processing massive datasets.
03:09 — As LLMs grow in size and scale, they require more resources; this includes cost, energy resources, and sustainability considerations.
04:00 — Many large LLMs are funded by and located within company clouds, which means limited accessibility for smaller tech companies.