The London constituency is extremely pleased to welcome you to this Meet the Scientist event! Come to know our scientists and endulge in discussions about the boundaries of neuroscience and artificial intelligence over a pint.
Humankind is currently in the process of solving two major but related challenges: the understanding of the brain and the development of artificial intelligence. For this event, we got three experts from UCL, King’s College, and the company Cambridge Analytica to come down to the pub and share with us some of the main questions that are yet to be solved in neuroscience, machine learning and neural networks.
At the end, we will have an open discussion to talk about the interface between neuroscience and artificial intelligence.Will computers be able to outthink us in the future?
This event will take place in the Hoop & Toy pub, 34 Thurloe Pl, Kensington London, SW7 2HQ, very near the South Kensington tube station, on the 13th of June at 19.00. When entering the pub you have to take the stairs of your right, and the event is in the private room of the first floor.
Our amazing speakers on this occasion and a brief summary of their talks are:
– The language of the brain: how neurons talk to each other
by Victoria González Sabater (PhD student at the MRC Centre for Developmental Neurobiology of King’s College London)
Did you know that the human brain contains over 80 billion neurons, and they constantly talk to each other? Making sense of this chatter requires understanding how neurons process and transmit information. Investigating this process will help us understand how we think, how things might go wrong in disease and even use this powerful biological processor as an inspiration for technology. In this talk I will discuss how neurons transmit signals and explain some of the methods that we use in the lab to “listen in” on neuronal communication. Neuroscience has gone far in understanding the functioning of the brain, but in this talk I will try to show that even how a single neuron encodes information is still somewhat of a mystery.
– A brief history of Machine Learning
by Dilan Chavda (Former Junior Data Scientist at Cambridge Analytica, Masters in Machine Learning at Imperial College London)
Will talk briefly about key developments in history of machine learning and provide a basic intuition of how deep learning nets work. I will then move on to applications as well as notable use cases from my own working experience and comment on machine learning in the real world today. I’ll conclude by stating my own opinion behind the machine learning hype but most importantly, throughout all of this, will probably be stressing that I am certainly no master of this field and still have a lot to learn myself
– The blind electrician: how neurons learn from experience
by Jorge Aurelio Menendez (Computational neuroscience PhD student at UCL Gatsby Computational Neuroscience Unit)
Considering the near-total helplessness of a human infant, it is remarkable that human adults are capable of such intelligent and adaptive behavior. This fact illustrates a phenomenon ubiquitous in mammalian species and beyond: the ability to learn from experience. Such impressive acquisition of novel skills is thought to emerge from highly coordinated changes in the connections between the billions of neurons in the brain. Such coordination, however, is difficult to reconcile with the known structure of neural circuits, whereby single neurons have severely limited access to information about the state of other neurons and their connections. This makes the problem of learning in the brain incredibly difficult: how could a neuron know how to change its connections to others without knowing how the others are changing their own? I consider this problem from a statistical perspective, and suggest a solution borrowed from machine learning.