AI Business recently interviewed one of the leading figures in AI, Juan Miguel Lavista. Juan is the Principal Data Scientist for Microsoft Data Science team, where he works with a team of data scientists searching for insights in petabytes of data.
The largest software developer in the world, Microsoft has been active in the AI space, particularly in 2016. At their Build conference back in March they announced Microsoft Cognitive Services, in June their CEO Satya Nadella published an editorial on building helpful AI, in July it was revealed that the latest Microsoft Office 365 include AI-powered features, and last month they acquired smart scheduling assistant startup Genee – and the list goes on.
Juan joined Microsoft in 2009 to work for the Microsoft Experimentation Platform (EXP) where he designed and ran randomize control experiments across different Microsoft properties. Juan has also worked as part of the Bing Data Mining team.
Having shared his AI and data analytics expertise at many conferences across the globe, as well as being a TedX Speaker in 2010, Juan will be speaking at The AI Summit in San Francisco on 28-29 September, where his keynote will reveal ‘All you wanted to know about Big Data and AI, but were afraid to ask…’
Juan Lavista of Microsoft
Juan believes the “biggest impact” of AI will be felt in scenarios where the technology enables companies to do things that were not previously possible. He elaborates on this point by example:
“An excellent example of this is search engines. There are trillions of web pages on the internet so it would be impossible to search and rank them without machine learning. AI will also impact business in automation, process improvement, cost reduction and an increase in agility. Because of this, many people are worried that AI will take their jobs, and organizations need to understand their employees’ concerns.
Juan acknowledges that emerging technologies have replaced jobs in the past, and explains by example once more:
“It is true that technology has historically impacted certain jobs – indeed, many jobs that existed a few decades ago have been entirely replaced by technology. For example, in the 1940s, telephone companies employed hundreds of thousands of phone operators, who were later replaced by electric switches”.
But while technology has destroyed jobs, Juan cites the conclusion from a study by Deloitte that says in the last 140 years, technology has created significantly more jobs than it has destroyed.
Juan believes the challenges for adopting AI technologies are not the same across all industry scenarios. Interestingly, he also points to the behaviour of humans, noting that our inability to act rationally is one that will impact the adoption of AI in several areas. A key example of where this is an issue, he explains, is driving:
“We are probably five years away from self-driving cars, and I’m confident that they will be safer and will reduce the number of car accidents significantly, but I’m also certain there will still be car accidents. Now, here is the dilemma: let’s say self-driving cars can reduce car accidents by half. Would you use a complete autonomous self-driving car even if you knew there was a chance of fatal accidents? From a statistics point, this is a no-brainer. However, the problem is that the majority of people think that they are good drivers, so they believe that those statistics don’t apply to them. In a complete autonomous self-driving car, the chances of accidents will be independent of the driver, so many of the drivers will think that now the chances of accidents are higher than before, and this will be a big challenge for adoption”.
“Another challenge for automakers is the liability when a self-driving car crashes. When an AI driver replaces a human one, it is not clear whether the motor vehicle owner or the person’s insurance company is liable. So I foresee a lot of changes in the insurance industry and changes in regulation before autonomous self-driving cars can go mainstream”.
The attitude of humans is key to allowing enterprises to embrace AI and enable its adoption. But true to his field, Juan also pinpoints data as another crucial facet to get right:
“Data is the fuel of AI. As a required prerequisite to embracing AI, an organization requires having good quality data, data management and be data-driven and willing to experiment. The will to experiment is easier to say than do and it requires allowing its employees to be able to fail and recognize that failing fast is a fundamental step of the innovation process”.
Microsoft have a part to play in almost every knowledge-working enterprise across the globe, let alone every industry. Juan shares his thoughts on the use of AI in various industries – how far have they come, where are they now, and are there any industries getting left behind?
“Some industries have been using machine learning (AI) for decades now”, he points out.
“In the case of the financial sector, the use of AI in production can be traced back to the 1980s, mainly in fraud detection, algorithmic trading, and risk prediction. Search engines are probably the best example of a whole industry that is mainly based on machine learning. The e-commerce industry has been leveraging machine learning for many key scenarios including product recommendation, targeting, fraud detection and ads ranking, for quite some time.
Healthcare particularly is a field with a high potential for AI adoption, Juan says:
Healthcare has been using machine learning for quite some time, but not to the point that it has become mainstream. Nevertheless, I believe that this is an area with a much higher potential and is a place where I would love to see more investment and research.
At The AI Summit in San Francisco on 28-29 September, Juan Lavista will deliver his keynote entitled ‘All you wanted to know about Big Data and AI, but were afraid to ask…’
He will be joined by other senior figures from the world’s biggest AI software developers and enterprises, all gathering to explore the huge opportunity that AI presents the business world.
To find out more, and to register to join us at the event, visit: theaisummit.com