Home  >  CrowdScience  >  Can Artificial Intelligence teach itself?
CrowdScience
Can Artificial Intelligence teach itself?
CrowdScience
Jun 9, 2023

Welcome to the exciting new field of generative artificial intelligence - or generative AI. We’re 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’s what got CrowdScience listener Ian wondering. He says: “Before 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?”

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 “inner child”. She took a decade’s 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 ‘generative AI’ 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

More Episodes
Apr 5, 2024
How many people have ever existed?

Today there are over eight billion people on Earth. That’s an awe-inspiring figure… but how does it compare to the vast numbers who came before us? Listener Alpha wants to know how many people have ever existed, so CrowdScience sets out to do a historical headcount.

The Population Reference Bureau in the USA estimated this number back in the 90s, and have been updating their calculations ever since. Demographer Toshiko Kaneda explains how their model works, the assumptions it makes – and the huge uncertainties around the number it comes out with.

We first need a date for when ‘humans’ first began, so Caroline travels to the Natural History Museum in London to meet human evolution expert Chris Stringer, and marvel at his collection of replica fossil skulls. Chris demonstrates how to distinguish our species, Homo sapiens, from other species like Neanderthals. When did these species first appear - and which of them count as human?

And once you know where to start the clock, how do you estimate the numbers of people alive at different points in history? For a population demographer like Walter Scheidel, it helps that some ancient civilisations kept detailed censuses, a few of which have survived to the present day. Caroline and Walter pour over one of these census fragments, and learn how to combine them with other archaeological clues to get some very rough numbers.

And finally: what does the future of our population look like? Poonam Muttreja from the Population Foundation of India discusses developments in the world’s most populous country, as well as the big demographic trends ahead for humanity.

Presenter: Caroline Steel Producer: Phil Sansom Additional Recording: Umaru Fofana Editor: Cathy Edwards Production Co-ordinator: Connor Morgans Studio Manager: Sue Maillot

Featuring: Toshiko Kaneda, Technical Director of Demographic Research, Population Reference Bureau Chris Stringer, Research Leader in Human Evolution, Natural History Museum London Walter Scheidel, Professor of Classics and History, Stanford University Poonam Muttreja, Executive Director, Population Foundation of India


26min 28sec


Mar 22, 2024
Do animals have anxious habits like us?

Many of us have habits that calm us down in times of stress. Things we find deeply comforting, like sucking our thumb or biting our nails. We might not even be aware we’re doing them, but they play a fundamental role in helping us regulate our emotions.

Our question this week comes from CrowdScience listener and nail-biter, Ash. He wants to know where these habits come from. And since his pet dog is also a nail-biter: do we share these traits with other animals?

Recently, a video of a mouse cleaning up a man’s shed took the internet by storm. Was this a house-proud mouse, or was it the animal's way of making sense of a frenetic environment? An emerging field of scientists focusing on animal behaviour and emotions help us shed some light on such questions.

Along the way we meet a dog training specialist, learn what a sniffari is, go for playtime with a thumb-sucking otter, and visit an OCD clinic. We’ll also be getting tips on how to give your pets the best home environment, and meet an animal enrichment officer in South Africa, who knows how to spot the signs of an unhelpful habit developing.

Contributors: Karolina Westlund, Ethologist, Stockholm University and ILLIS Ben Terry, CBT Therapist, Priory Hospital North London Karin Pienaar, Animal Behaviourist, COAPE International Candice Ward, Animal Behaviourist, Johannesburg Zoo

Jaak Panksepp clip: The science of emotions: Jaak Panksepp at TEDxRainier

Producer: Robbie Wojciechowski Presenter: Alex Lathbridge Editor: Cathy Edwards Production co-ordinator: Connor Morgans Additional recording by Elna Schutz

(Photo: Portrait of border collie puppy biting a curtain. Credit: Rawlstock/Getty Images)


27min 29sec


Can Artificial Intelligence teach itself?

--:--
--:--