A lot of data that we work with everyday is messy, unstructured and hard to work with programmatically. Tasks that are easy for humans, like finding the invoice number on an arbitrary pdf are very hard to solve with ordinary code in the real world. New deep learning models based on so-called transformer architectures, like GPT-3 from OpenAI or Bert from Google, are creating a buzz by showing a level of text understanding that was only recently way beyond the reach of machines.
Time to look at this new technology that will soon be a standard machine learning tool in many software projects.
In this session, the following questions will be tackled:
– How does a neural network learn to understand language?
– How do systems like GPT-3 work?
– What is the current state of the art?
– What works and what doesn’t?
And most importantly: What can you do with that technology and how do you do it?
Note: This session does not require any prior knowledge of artificial intelligence or machine learning.
CTO at DIVISIO.
Christoph Henkelmann holds a degree in Computer Science from the University of Bonn. He is currently working at DIVISIO, an AI company from Cologne, where he is CTO and co-founder. At DIVISIO, he combines practical knowledge from two decades of server and mobile development with proven AI and ML technology. In his pastime he grows cacti, practices the piano and plays video games.
