There is a lot of concern being expressed in the press nowadays about the rise of artificial intelligence and how it could be the beginning of the end for humankind. The two best-known Cassandra’s when it comes to the dangers of artificial intelligence are Stephen Hawking and Elon Musk. Hawking and Musk are well-known and respected visionaries who are as concerned about the future as they are about the present. In fact, it’s the future of AI not its present that they are warning about. Hawking recently told the BBC that “the state of artificial intelligence (AI) today holds no threat, but he is concerned about scientists in the future creating technology that can surpass humans in terms of both intelligence and physical strength. … Hawking’s comments closely follow those made by high-tech entrepreneur Musk, who raised controversy in late October when he warned an audience at MIT about the dangers behind AI research.” [“Stephen Hawking Says AI Could ‘End Human Race’,” by Sharon Gaudin (@sgaudin), CIO, 3 December 2014] I agree with Hawking and Musk that caution should be the watchword as we decide how to apply AI in the future. At the heart of the debate is machine learning (ML). After all, only machines that learn will ever be able to decide if it’s worth keeping humans around. There is nothing inherently sinister about artificial intelligence and/or machine learning, even Musk agrees that today AI “is more about robotic vacuum cleaners than Terminator-like robots that shoot people and take over the world.” So what is machine learning good for?
“Machine Learning,” according to an article published by the Etisalat BT Innovation Center (EBTIC), “is now quite a matured discipline and captures in general any activity that involves automated learning from data or experience.” [“Machine Learning on Big Data: What is machine learning (ML) useful for?” EBTIC, 19 August 2014] The article goes on to note, “At the core of ML is the ability of a software or machine to improve the performance of certain tasks through being exposed to data and experiences. [A] typical machine learning model first learns the knowledge from the data it is exposed to and then applies this knowledge to deliver predictions about the new, same type but previously unseen data.” The machine learning that seems to be drawing the most attention is deep learning. Julian Green (@juliangreensf), CEO of Jetpac Co., notes, “Pioneered in the 1980s, the [deep learning] approach allows computing devices to teach themselves to recognize patterns by analyzing massive quantities of data.” [“Deep Learning’s Role in the Age of Robots,” Wired Innovation Insights, 2 May 2014] The question is: Are there good reasons to pursue deep learning technologies? A lot of scientists, and the organizations they work for, believe there are. “Using an artificial intelligence technique inspired by theories about how the brain recognizes patterns,” writes John Markoff, “technology companies are reporting startling gains in fields as diverse as computer vision, speech recognition and the identification of promising new molecules for designing drugs.” [“Scientists See Promise in Deep-Learning Programs,” New York Times, 23 November 2012]
Explaining his comment about deep learning being used to help design new drugs, Markoff reports that one of the achievements by a deep learning program was accomplished by “a team of graduate students studying with the University of Toronto computer scientist Geoffrey E. Hinton.” At the last minute, the team entered “a contest sponsored by Merck to design software to help find molecules that might lead to new drugs.” The team won the top prize even though its members had “no specific knowledge about how the molecules bind to their targets. The students were also working with a relatively small set of data; neural nets typically perform well only with very large ones.” Anthony Goldbloom (@antgoldbloom), chief executive and founder of Kaggle, a company that organizes data science competitions, told Markoff, “This is a really breathtaking result because it is the first time that deep learning won, and more significantly it won on a data set that it wouldn’t have been expected to win at all.” Markoff continues:
“Advances in pattern recognition hold implications not just for drug development but for an array of applications, including marketing and law enforcement. With greater accuracy, for example, marketers can comb large databases of consumer behavior to get more precise information on buying habits. And improvements in facial recognition are likely to make surveillance technology cheaper and more commonplace.”
As president of a company involved in the development of cognitive reasoning solutions for businesses, one of my primary interests in deep learning programs is how they can help businesses use their resources more efficiently and effectively. Jeff Rajeck (@JRajeck), Director of Digital Marketing and Analytics at Maachu, remarks, “Machine learning sounds like something that computer nerds do, but not marketers.” [“How to use machine learning to enhance your marketing campaigns,” Econsultancy, 5 August 2014] Rajeck goes on to provide an excellent primer on how and why machine learning can be used to improve marketing efforts. He concludes, “Though it’s unlikely that machine learning experts will take our marketing jobs any time soon, it’s important for us to be familiar with new technology and know what is possible.” What’s possible using machine learning goes far beyond the two examples provided so far. Back in 2008 in the introduction to a class on Theoretical Machine Learning, Robert Schapire, a Professor at Princeton University currently on leave at Microsoft Research in NYC, provided a few examples of problems machine learning can help address. The list included:
- Optical character recognition
- Face detection
- Spam filtering
- Topic spotting
- Spoken language understanding
- Medical diagnosis
- Customer segmentation
- Fraud detection
- Weather prediction
Analysts from the consulting group OPS Rules add price forecasting and supply chain optimization to that list of use cases. “Analytics is of course a very wide area,” they write, “we would like to focus on one technology that has not been implemented widely in supply chain until recently called Machine Learning and in particular how it can be combined with optimization to produce breakthrough results.” [“Combining Machine Learning and Optimization in Supply Chain Analytics,” Business2Community (B2C), 1 July 2014] As useful as traditional machine learning can be, it still faces challenges when it comes understanding nuances of the human language. Catherine Havasi, a research at MIT’s Media Lab, writes, “Common-sense reasoning is a field of artificial intelligence that aims to help computers understand and interact with people in a more naturally by finding ways to collect these assumptions and teach them to computers.” [“Who’s Doing Common-Sense Reasoning And Why It Matters,” TechCrunch, 9 August 2014] She continues:
“Common Sense Reasoning has been most successful in the field of natural language processing (NLP), though notable work has been done in other areas. This area of machine learning, with its strange name, is starting to quietly infiltrate different applications ranging from text understanding to processing and comprehending what’s in a photo. Without common sense, it will be difficult to build adaptable and unsupervised NLP systems in an increasingly digital and mobile world. When we talk to each other and talk online, we try to be as interesting as possible and take advantage of new ways to express things. It’s important to create computers that can keep pace with us.”
At Enterra Solutions®, our Cognitive Reasoning Platform™ (CRP) uses both reasoning (which involves the world’s largest common sense ontology) and computation to achieve insights. We call this a Sense, Think, Act, and Learn® process. In today’s world, where most new data is unstructured, common sense reasoning is becoming more important. Havasi explains:
“The power of common sense systems is that they are highly adaptive, adjusting to topics as varied as restaurant reviews, hiking boot surveys, and clinical trials, and doing so with speed and accuracy. This is because we understand new words from the context they are used in. We use common sense to make guesses at word meanings and then refine those guesses and we’ve built a system that works similarly. Additionally, when we understand complex or abstract concepts, it’s possible we do so by making an analogy to a simple concept, a theory described by George Lakoff in his book, ‘Metaphors We Live By.’ The simple concepts are common sense. There are two major schools of thought in common-sense reasoning. One side works with more logic-like or rule-based representations, while the other uses more associative and analogy-based reasoning or ‘language-based’ common sense — the latter of which draws conclusions that are fuzzier but closer to the way that natural language works. Whether you realize it or not, you interact with both of these kinds of systems on a daily basis.”
Using both mathematical and semantic reasoning means that insights gleaned from data analysis are less likely to contain meaningless conclusions — like, “dead people don’t go shopping”. As a result, insights provided by common-sense reasoning machines will provide a much better basis for taking action and much better understanding of new relationships. Havasi concludes:
“The last few years have seen great steps forward in particular types of machine learning: vector-based machine learning and deep learning. They have been instrumental in advancing language-based common sense, thus bringing computers one step closer to processing language the way humans do. NLP is where common-sense reasoning excels, and the technology is starting to find its way into commercial products. Though there is still a long way to go, common-sense reasoning will continue to evolve rapidly in the coming years and the technology is stable enough to be in business use today. It holds significant advantages over existing ontology and rule-based systems, or systems based simply on machine learning.”
I believe that common-sense reasoning platforms will be part of a human/machine team that will help advance human knowledge and understanding at a breathtaking pace. As businesses catch the vision of what these systems can do for them — from improving operations to increasing sales — they, too, will jump on the bandwagon. The narrow uses of AI and machine learning discussed above are unlikely to lead to systems capable of destroying mankind; but, we would be foolish to ignore the possibilities being raised by Hawking and Musk.
