Andy Burgin
- keynote
Large Language Models for Humans
Learn from Andy"Is this a machine thinking? Or is it just doing very clever, glorified, predictive text based on what it was trained on?"
"We want things to look human. We want variety. We want variability in there. So, we use probability to choose the next word, which doesn't guarantee us to pick the most popular one..."
Large Language Models are the brains behind AI services and applications. They appear to “know” and “understand” almost everything, but how can such breadth of knowledge be condensed down to a file that can comfortably fit on a memory card? In this talk I’ll explain how LLMs work, what is actually in them, why they seem to “know” everything and dispel some of the misconceptions and fears. All without using mathematical or technical jargon, by breaking the concepts down into a series of understandable bite size nuggets.
