Main content

Can Artificial Intelligence teach itself?

Welcome to the world of generative artificial intelligence. How reliable are these algorithms? And what happens if they start learning from their own material?

Welcome to the exciting new field of generative artificial intelligence - or generative AI. We鈥檙e not talking about robots or spaceships: instead these are image generators and chatbots that are already revolutionising the way people write, research and interact in the virtual world. Their capabilities are extraordinary. But before they can produce results they first need to be trained on vast amounts of data. And that鈥檚 what got CrowdScience listener Ian wondering. He says: 鈥淏efore long, much of the material on the internet will have been written, or at least co-written, by AIs. What will happen when AIs are being trained on texts they have written themselves?鈥

To answer this question Anand Jagatia first meets artist and scientist Michelle Huang. Michelle recently embarked on a therapy project to try and connect with her own 鈥渋nner child鈥. She took a decade鈥檚 worth of her diaries from her childhood and fed them into an AI to try and simulate a conversation between her and her younger self. Can a machine convincingly recreate the voice of a human being? You can be the judge!

Then, data scientist Briana Brownell joins Anand to explain a little about how 鈥榞enerative AI鈥 like this actually works. Together they open up the popular chatbot, ChatGPT, to put it to the test: but can it help Anand solve his personal conundrum?

Now that we understand what AI is capable of we also need to know its limits. AI researcher Pablo Villalobos describes the process of training generative AI and where it gets its material from. The amount of data consumed in this way keeps going up and up, Pablo and his colleagues have been calculating exactly when that data is likely to run out. Pablo also discusses recent research exploring a unique effect that occurs if AI is repeatedly trained on its own material. Finally, Briana Brownell returns to cover a range of wider issues related to artificial intelligence and what effects it might have on our future.

Presenter: Anand Jagatia
Producer: Phil Sansom
Editor: Richard Collings

Available now

26 minutes

Last on

Mon 12 Jun 2023 19:32GMT

Artificial Intelligence Poem

Artificial intelligence language models like ChatGPT are capable of incredible things. We asked it to compose a poem as an ode to our show, and here is the result:


Upon the airwaves, voices tell a tale,

Of CrowdScience, knowledge's holy grail.

From the heart of 大象传媒, it emerges,

An oracle of wisdom, where curiosity surges.

With a world as its stage, questions in hand,

Seeking the mysteries of sea, air, and land.

The cosmos above, or the earth's inner core,

Each episode a journey to explore.

Through time zones and cultures, it does span,

Revealing the secrets of nature to woman and man.

In the language of science, truths it unfolds,

In a symphony of stories, brilliantly told.

The whispers of atoms, the roar of space,

The complexities of life in every place.

Invisible forces that pull, push, and twine,

The cosmic dance, in their hands, they define.

Listeners tuned in, from east to west,

In the quest for knowledge, no time for rest.

From dawn's early light to the midnight hour,

Drinking from the fountain of science's power.

In the melody of the World Service song,

To CrowdScience, we all belong.

A beacon in the static, a voice in the night,

Guiding us all in the quest for light.

Broadcasts

  • Fri 9 Jun 2023 19:32GMT
  • Mon 12 Jun 2023 01:32GMT
  • Mon 12 Jun 2023 04:32GMT
  • Mon 12 Jun 2023 08:32GMT
  • Mon 12 Jun 2023 12:32GMT
  • Mon 12 Jun 2023 19:32GMT

Podcast