April 28, 2016
AIBusiness recently interviewed one of the leading researchers in Artificial Intelligence, Dr Andrew J.R. Simpson, Research Fellow in the Centre for Vision, Speech and Signal Processing at the University of Surrey.
Andrew will bring his AI expertise to The AI Summit in London on 5 May, where he will ask 'What's so Deep about Deep Learning?', developing some simple intuitions about Deep Learning and what Deep Neural Networks can and can’t be expected to do.
AIBusiness caught up with Andrew to find out his views on AI’s impact on business, the challenges we face in adopting AI and more, looking ahead to his presentation at The AI Summit.
How do you believe AI and Deep Learning will impact business overall and in what ways?
Right now, we are living out the final moments of the quantitative computing age. As a result, we are confused about how to reconcile AI with computing. Deep Learning provides the first step towards redefining computing itself: we are entering the age of qualitative computing.
We are opening the door to the most disruptive technology since we learned to make fire. At first, fire kept us warm, but we changed the course of our evolution when we learned to cook. First we cooked food and our brains got bigger. Then we cooked metals and we entered the bronze and then iron ages.
Exploitation of new technology requires a combination of vision, infrastructure investment and monopoly. There are a number of private organisations investing very heavily in the race for monopoly at this early stage.
What do you think are the main challenges in adopting AI technologies, from machine learning through to deep learning and image recognition, in business?
With the drive for Open Source AI platforms - and Deep Learning in particular - entry into the world of AI couldn’t be easier. However, as any good economist can tell you, wealth is created by new technologies and leveraged by monopolising the new technologies.
A ‘long-view’ AI business model is emerging. AI tech IP is the core asset and acquisitions are being driven by future patents rather than by tangible products. Google are the market leader in this new model and have demonstrated remarkable foresight in terms of aggressive AI portfolio acquisition.
So, investors are starting to ask AI start-ups: How many patent applications have you filed? Who owns the IP that underpins the Open Source AI platform you rely on?
Your talk at The AI Summit focuses on deep learning. Which business sectors in particular do you foresee deep learning impacting? What is the long-term potential for deep learning (and AI in general) in business?
The take-home message of my AI Summit talk is that Deep Learning is about qualitative computing. It’s about quality in the philosophical sense of the word.
Ask an Engineer: How do you herd sheep? He will say “How do you define a flock?”. What you need is a dog. The dog understands the inherently qualitative nature of flocking. That is intelligence. When we do that with AI, that is qualitative computing. We will interact with AI on our own terms.
If we think back to the time before computers, we might have asked: “What business sectors will benefit from computing?”. The ultimate answer turned out to be “all of them, plus some that don’t yet exist”. The same will be true of qualitative computing.
Do you see the relationships between universities, product developers and businesses changing with the more widespread use of AI in business?
Academia is being completely stripped of AI researchers, particularly in the Deep Learning field, by migration to industry. If you have any profile at all, the Deep Learning job offers are relentless and industry appears insatiable.
Beyond this, the gulf between academic and private research in AI is widening at an alarming rate. When I publish a deep learning paper, it often gets implemented by industry within a week but might take a year to get picked up by an academic lab. In addition, the fundamental work I am able to do in my private neural network lab is several years ahead of what is done in the public-funded university lab, so my private lab generates a stream of IP and the university lab doesn’t.
At The AI Summit, Dr Andrew J.R. Simpson will be delivering his keynote: What's so Deep about Deep Learning?
The AI Summit is the world’s first event dedicated to Artificial Intelligence for the business world. For more information, and to join us on 5 May at the Four Seasons Hotel, London, visit: theaisummit.com