Trends in Natural Language Processing with Sameer Singh - #445
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Limits of Language Models in NLU
This chapter examines the inherent limitations of language models in achieving true natural language understanding, contrasting form with meaning in NLP. It discusses a thought-provoking scenario involving an octopus that highlights the gap between text prediction and genuine comprehension. By critiquing recent research and exploring the interplay between linguistic theory, visual data, and machine learning, the chapter underscores the complex challenges in enhancing NLP capabilities.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
