We at AIBusiness are officially launching our new series ‘AI Innovators’ this January. This is a series dedicated to featuring interviews with relevant spokespersons across a range of businesses that have entered the world of artificial intelligence. AI Innovators aims to provide our readers with a broad insight into how artificial intelligence is implemented within different industries, anything from finance to medical research.
Second in line is BenevolentAI, a British technology company harnessing the power of AI to enhance and accelerate scientific discovery and innovation. We spoke to their CEO, Jérôme Pesenti, who is a world leading pioneer in AI focused on big data and machine learning for the past 16 years.
Pesenti joined BenevolentAI from IBM Watson where he created and led the development of the Watson Platform – a technology that uses natural language processing and machine learning to give insight into large amounts of unstructured data. At IBM, Pesenti also held the role of CTO for the core IBM Big Data product portfolio. Prior to joining IBM he co-founded search and text analytics company Vivisimo which was acquired by IBM in 2012.
How is Benevolent currently involved in the AI space?
“BenevolentAI is a British company applying AI to enhance and accelerate global scientific discovery. It does this by transforming the world’s mass of highly fragmented scientific information into new insight and useable knowledge that ultimately benefits society”, Pesenti explains.
“Initially we are focussed on the area of drug discovery – applying AI to create better medicines more quickly for patients”.
Which AI technologies are you specifically focusing on developing, and how does this fit into your existing solutions?
“We use Machine Learning (in particular Deep Learning) and Natural Language Processing to process and understand structured and unstructured scientific information (like scientific papers, patents, and clinical trials), create a large repository of life science knowledge, and predict from it new interesting links that can help our scientists come up with new discoveries”, Pesenti explains.
He mentions that Benevolent also applies novel Deep Learning algorithms to do predictive and generative biochemistry, to come up with better therapeutic molecules.
As the implementing of AI in healtchare and medicine is not new, we wanted to know where Benevolent stands out from the crowd. What is it that sets you apart from others in the market?
What is it that sets you apart from others in the market?
The way we bring bioscience and computer technology together is unique. We develop and apply AI technology together with our own drug discovery scientists – so we combine the power of machine brains with expert human brains.
Pesenti explains that he does not believe any other AI companies are doing this – having AI technology and drug discovery experts sitting together everyday working towards a common purpose.
“It means our sole focus is on a specific discovery use case: end to end drug discovery and development with direct and constant access to the BenevolentAI science team”, Pesenti explains. So we can understand very precisely the domain & language that can extract and represent existing knowledge, use the domain knowledge to make predictions and improve and learn quickly based on direct user feedback.
“So we can understand very precisely the domain & language that can extract and represent existing knowledge, use the domain knowledge to make predictions and improve and learn quickly based on direct user feedback”.
Given the rapid involvement of AI in this industry, we wanted to know what Benevolent’s 5-year plan was to keep up with the development and still stay ahead of the game?
“In five years we will have a large number of successful drug programs, demonstrating the effectiveness of our approach”, Pelenti says. “We also plan to diversify beyond human drug discovery to other scientific industries both in bioscience (e.g. animal health, nutraceuticals) and outside bioscience e.g. energy, materials science, agriculture etc”.
“Looking back, 2016, was the year when AI hit the mainstream. But as it grew in prominence, the conversation has been ambiguous about what AI could do. 2017 will see AI really ‘grow up’ and we’ll start to see it being used for more real-world applications”.
“With advanced technology like deep learning available, the challenges for enterprises will be identifying the areas they feel it can help and having the training data in place to take advantage of it”, Pesenti says when asked what he believes are the biggest challenges faced by enterprises looking to adopt AI.
“In particular, teaching computers to learn from industry data will require innovative new semi-supervised learning techniques, more economical and practical than the wholly supervised techniques seen today”.
So where do you see the biggest opportunity for the enterprise over the next 5 years?
“AI has the opportunity to transform every industry, some sooner than others, which will cause a disruption not seen since the Internet was created”.
Pesenti look at the example, of the car company of the future who will be the first to deploy and master a functioning self-driving car, while successful future retailers will provide seamless checkout experiences and drug companies will be creating AI-assisted discoveries.
“Enterprises that adapt the best will have the opportunity to become the major players in their industry”.