The Royal Society is currently running a series of lectures called ‘You and AI’, creating a public dialogue around what Machine learning and AI is, how the technology works and the ways it may affect our lives.

As part of this conversation a number of schools groups were invited to attend the lectures to see if and how the series changes their understanding and expectations of the technology. Two students write about their experience after the first two lectures below.

Demis Hassabis, co-founder of DeepMind

Demis Hassabis, co-founder of DeepMind

“A Royal Society survey found that just one in ten people in the UK recognised the term ‘machine learning’, even though the technology is already part of our everyday lives through things like Siri on our phones and Alexa in our homes. Listening to the first lecture given by Demis Hassabis, co-founder of Deep Mind, I became better informed about rapidly developing research in machine learning and how its application has the ability to revolutionise society as we currently know it.

I learnt that there are limitations to systems utilising AI or machine learning at the moment, like Siri, Deep Blue and AlphaGo, because they can only perform one task, and one task only, unlike the multiple tasks a human can do and swap between. Hassabis’s company’s best-known machine learning program, AlphaGo, uses the strategy game Go as a way to test the capabilities of computers using machine learning principles. AlphaGo’s successfully beat a Go world champion and this has led to Hassabis’s team developing an updated system called Alphazero which can learn in an unsupervised way allowing it to develop skills for multiple games .

The lecture made me realise that we have to ask ourselves “what is intelligence and could it be created inorganically?” I felt that there would be a need for governments to develop a deeper understanding of the social and economical impact of Machine learning and AI in the future, to help them tackle the ethical implications when formulating guidelines and laws.

There is a new world coming very fast. AI was about board games, it’s now in our phones, homes and driver-less cars are very close to being in our streets, but the future could look even more different” Ollie

 

Exterior image of the Royal Society building“Before I attended the first in the lecture series at the Royal Society, I did not really know a lot about AI and its huge potential for the future. Demis Hassabis’ talk not only provided clear explanation of what Machine Learning and AI is, he described how his company Deepmind had a collection of engineers, mathematicians as well as neuroscientists that showed me that AI was a combination of lots of different sectors, something I had not realised.

The highlight of the lecture was when he talked about AlphaGo playing the world champion Go player and described DeepMind’s new AI system, Alphazero, was also incredibly exciting because it has the ability to learn and compete in any type of game. Stockfish, which was thought to be the most powerful chess system possible, was beaten within just 4 hours of Alphazero practicing against itself. This showed me that machine learning is so much more powerful than standard coding and computing because in some sense it allows the computer to have a mind of its own and develop ideas that human minds could not do.

In the second lecture Suzanne Aigrain, Professor of Astrophysics at Oxford, talked about how AI can have an impact on science and discovery by helping to solve problems like the cosmological parameter estimation and discover new things like exoplanets. AI can be used to intelligently sort through the huge amounts of data produced that would be impossible for humans to do in a similar timescale. This is incredibly useful but one of the problems she mentioned was that it was very hard to teach the computers to not throw away data because it may hold unexpected new discoveries that the computer cannot predict.

Antonio Ciminisi, principle researcher at Microsoft, talked about AI could be applied to medicine. Its impact on the public is a lot more direct compared to the discovery of exoplanets and it is an example of how AI can currently be used in our lives.

Before I went to the lecture I had very little intent of studying computer science at a higher level because I did not know a lot about its potential. I thought it would not be that exciting as a career, but the examples of the future presented at the lectures demonstrated to me that whatever sector I go into in the future, AI will likely play a huge role. The lectured have inspired me to look into AI more deeply and made me consider it as an option for university.” Simone