AI Meets… Research
Jul 30, 2015
At Frost & Sullivan, we are undertaking a large scale research project exploring the applications of artificial intelligence (AI) across vertical industries including; healthcare, financial services, transportation and energy. AI Meets…. will be a regular blog that highlights some key insights from this research project. The first part of the series is apt, AI Meets…Research. Otherwise known as why I will not have a job in 5 years.
One of the critical errors made when discussing AI is imagining robots like HAL9000 or a computer with a human voice trying to end the world like in War Games. All great fun, sure. But not a useful guide to better understand how businesses are changing. AI will be incorporated into existing software as new functionality like IBM Watson or Microsoft Azure. It will also be driven by new startups offering some form of machine learning, machine vision or natural language processing. The language used will differ but broadly we are talking about AI-based software. The introduction will go unnoticed and unlike discussion about artificial general intelligence with human-like abilities, there will be no ethical debate. The software will be differentiated by its ability to complete increasingly complex tasks like identifying objects and human emotions in videos or to summarising millions of articles in an infographic. Advances in machine vision and natural language processing are most noticeable in driverless cars and speaking robots, but it is their impact in the hum-drum world of research that could be just as revolutionary.
Ellipse, the first product from Thoughtly, is one of the first, but certainly not the last to apply artificial intelligence to research. Using natural language processing to analyse, visualise and summarise large volumes of text in real-time. This is the first of many products that will use AI allowing researchers to search greater numbers of documents and sources and pull out greater insight more quickly. This will make existing researchers more productive reducing the overall number required for any task, and in the aggregate reduce the number of research-based jobs.
Many people in the economy, including me, are paid for their ability to analyse and provide insight from data. Until very recently, it was only numbers that could be analysed, and these numbers would have to be in a database readable format. Products are now coming to market that can analyse text, voice, photos and soon video. Voice analysis is fairly standard in many call centres and video analysis in sports for example. However, these solutions are often expensive, bespoke and unable to analyse across multiple data formats. For now, Ellipse is a canary down the mine for researchers. In a very short space of time, a more advanced version will collate all online published research, blogs, podcasts, YouTube video and press releases. Insight could be delivered in a cloud-based dashboard allowing any member of the organisation to instantly find answers to their business questions. Insight derived in the same time it would take a human researcher to finish the morning emails. I am calling this new market Insight-as-a-service or the Insight-on-Demand Economy. Neither are particularly catchy, however.
The initial market for Ellipse is researchers. This strategy of clearly defining the market and attacking a niche is the right strategy for a start-up. But don’t be fooled. Millions of people are employed across almost all industries to pull data from various sources, analyse this data and provide insights that help executives make better business decisions. A piece of software that is able to analyse, visualise and summarise vast amounts of text-based data is just the start. Whether it is Thoughtly, or a competitor, more and more data formats will be added. For businesses, this is simply the next layer of automation. Just as Microsoft Excel replaced hundreds of thousands of people using electro-mechanical calculating machines, AI enterprise software will do the same to hundreds of millions of people using Excel, Powerpoint and an Internet browser.
For more #hottakes on AI and emerging technology including virtual reality, the blockchain and robotics follow me on Twitter and Medium @lawrencelundy.
Please add your bio info through your member profile page, or through your dashboard.