Google DeepMind: The Podcast

Speaking of intelligence

Jan 25, 2022

38:11

Creator website

AI Summary

Highlights

AI Chapters

Episode notes

Podcast summary created with Snipd AI

Quick takeaways

Language models have the potential to achieve artificial general intelligence by instilling AI with language capabilities.

While language is a powerful tool, combining it with other capabilities like perception and reasoning may be necessary to achieve a well-rounded AGI.

Deep dives

The Power of Language Models

Language models, particularly large ones like GPT-3, have the ability to generate human-like text and have widespread applications in our daily lives, from auto-completing sentences to chatbots and translation services. Researchers at DeepMind believe that language models not only have the potential to write better scripts but also play a pivotal role in achieving artificial general intelligence (AGI). AGI refers to a system's ability to succeed in a wide range of problems and situations. By instilling AI with language capabilities, researchers hope to create a versatile and powerful intelligence that can contribute to various fields such as science, arts, literature, and exploration.

Can You Detoxify Language Models?

03:08

Are Language Models Just Clever Parrots?

03:55

Language Models Could Be Used to Spread Misinformation

03:50

Introduction

5min

Is Language an Important Capabilities for Artificial Intelligence?

3min

Language Is Important for Communication, Memory and Thought

4min

What's My Worst Model?

2min

The Impact of Large Language Models on the Environment

3min

How to Write a Link Model for a Black Hole?

2min

Language Models Could Be Used to Spread Misinformation

4min

Language Models - Toxicity

4min

Is Your Toxicity Classifier Biased?

3min

10.

Can You Detoxify Language Models?

3min

11.

Are Language Models Just Clever Parrots?

4min

12.

Is Language the Key to Artificial General Intelligence?

3min

Hannah explores the potential of language models, the questions they raise, and if teaching a computer about language is enough to create artificial general intelligence (AGI). Beyond helping us communicate ideas, language plays a crucial role in memory, cooperation, and thinking – which is why AI researchers have long aimed to communicate with computers using natural language. Recently, there has been extraordinary progress using large-language models (LLM), which learn how to speak by processing huge amounts of data from the internet. The results can be very convincing, but pose significant ethical challenges.

For questions or feedback on the series, message us on Twitter @DeepMind or email podcast@deepmind.com.

Interviewees: DeepMind’s Geoffrey Irving, Chris Dyer, Angeliki Lazaridou, Lisa-Anne Hendriks & Laura Weidinger

Credits

Presenter: Hannah Fry

Series Producer: Dan Hardoon

Production support: Jill Achineku

Sounds design: Emma Barnaby

Music composition: Eleni Shaw

Sound Engineer: Nigel Appleton

Editor: David Prest

Commissioned by DeepMind

Thank you to everyone who made this season possible!

Further reading:

GPT-3 Powers the Next Generation of Apps, OpenAI: https://openai.com/blog/gpt-3-apps/

https://web.stanford.edu/class/linguist238/p36-weizenabaum.pdf

Never Mind the Computer 1983 about the ELIZA program, BBC: https://www.bbc.co.uk/programmes/p023kpf8

How Large Language Models Will Transform Science, Society, and AI, Stanford University: https://hai.stanford.edu/news/how-large-language-models-will-transform-science-society-and-ai

Challenges in Detoxifying Language Models, DeepMind: https://deepmind.com/research/publications/2021/Challenges-in-Detoxifying-Language-Models

Extending Machine Language Models toward Human-Level Language Understanding, DeepMind: https://deepmind.com/research/publications/2020/Extending-Machine-Language-Models-toward-Human-Level-Language-Understanding

Language modelling at scale, DeepMind: https://deepmind.com/blog/article/language-modelling-at-scale

Artificial general intelligence, Technology Review: https://www.technologyreview.com/2020/10/15/1010461/artificial-general-intelligence-robots-ai-agi-deepmind-google-openai/

A Definition of Machine Intelligence by Shane Legg, arXiv: https://arxiv.org/abs/0712.3329

Stuart Russell - Living With Artificial Intelligence, BBC: https://www.bbc.co.uk/programmes/m001216k/episodes/player

Please like and subscribe on your preferred podcast platform. Want to share feedback? Or have a suggestion for a guest that we should have on next? Leave us a comment on YouTube and stay tuned for future episodes.

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.

Google DeepMind: The Podcast

Speaking of intelligence

Podcast summary created with Snipd AI

Quick takeaways

Deep dives

The Power of Language Models

The Evolution of Conversational AI

Challenges and Considerations of Language Models

Beyond Language: The Need for Grounded Intelligence

Can You Detoxify Language Models?

Are Language Models Just Clever Parrots?

Language Models Could Be Used to Spread Misinformation

Remember Everything You Learn from Podcasts