You can check out this session on


About this Session

We can train autoregressive language models such as GPT-3 that use deep learning to produce human-like text, and this brings tremendous advantages to SEOs and content marketers. Still, there are limits and pitfalls that we should avoid (limited generalizability, unwanted biases, low accuracy, practical efficiency and environmental costs). This session is about learning how to generate text (and quickly walk you through the evolution of machine-generated text from Markov chains to neural networks) and focuses on how to “hack” text-to-text transformer models to combine human intuition with artificial intelligence.

Sign up to the mailing list for more info about the event
How we use the data in this form.