When Kenneth Wayne Jennings, noted for holding the record for the longest winning streak of 74 games on the U.S. syndicated game show, bowed to IBM’s Watson as the new “Jeopardy!” champ in 2011, he quoted an episode of “The Simpsons” and wrote “I for one welcome our new computer overlords,” on his video screen. That was probably one small step for a computer but a giant leap for computing.
It’s ironic to say that Watson’s dominance on the game show didn’t come out of the blue. The result was a culmination of over a decade of IBM’s research. The victory, however, wasn’t the end but the beginning of a new era. “It opened up a new chapter in information technology called cognitive computing—based on the idea of a natural interaction between systems and people,” says Zachary (Zach) Lemnios, vice president of strategy for IBM Research.
And this evolving relationship between humans and machines was also the key theme of Gartner’s recent “Hype Cycle for Emerging Technologies, 2013.” The analyst and consulting firm says it chose to feature the relationship between humans and machines due to the increased hype around smart machines, cognitive computing, and the Internet of Things.
Systems augment humans
It’s no secret that organizations today, across industry, are overwhelmed with so much data that they are unable to make time-critical decisions. Not only is the data growing by leaps and bounds it is also coming in multiple shapes and forms. “These increasingly challenging times need organizations to make tighter decisions in tighter timelines with the consequence of each decision going up,” Lemnios said.
Such a scenario probably calls for a closer look at the interaction between humans and computers.
“We encourage enterprises to look beyond the narrow perspective that only sees a future in which machines and computers replace humans. In fact, by observing how emerging technologies are being used by early adopters, there are actually three main trends at work,” said Jackie Fenn, vice president and Gartner fellow.
These trends, in Fenn’s view, are “augmenting humans with technology—for example, an employee with a wearable computing device; machines replacing humans—for example, a cognitive virtual assistant acting as an automated customer representative; and humans and machines working alongside each other—for example, a mobile robot working with a warehouse employee to move many boxes.”
Of these major trends shaping the emerging technologies domain, IBM is focused on cognitive computing which is also broadly classified under the natural-language question and answering systems. Lemnios explained the relevance of cognitive computing in the context of network security.
“Networks today are entirely ad hoc, largely mobile, and a mashup of consumer and industrial grade systems and they are vulnerable because they are so changeable that it allows access to attackers,” he said.
The current approach to tackling such vulnerabilities is a forensic one. “We look at error logs, attack files, and issue patches but it is essentially a rear view mirror approach,” Lemnios said. But a network with cognitive agents residing in it can provide real-time assessments of the vulnerabilities in the network and other meaningful insights through a continuous dialogue with the network operations center. “This could fundamentally change how we run our networks,” he added.
IBM hopes to extend what has been learned with analytics, extend that to human-system interactions using speech as the medium, and take away the need to code a software tool. “Why have structured programs to deal with information when the underlying data is not?” Lemnios asks.
Best of both worlds
Gartner's studies explain that "humans versus machines" is not a binary decision; sometimes machines working alongside humans is a better choice. IBM’s Watson does background research for doctors, just like a research assistant, to ensure they account for all the latest clinical, research and other information when making diagnoses or suggesting treatments.
At a time when humans are clearly reaching the limits of what we can absorb and understand, Gartner suggests the main benefit of having machines working alongside humans is the ability to access the best of both worlds (that is, productivity and speed from machines, emotional intelligence, and the ability to handle the unknown from humans).
As machines get smarter and start automating more human tasks, humans will need to trust the machines and feel safe. IBM’s Watson provides “confidence” scores for the answers it provides to humans.
Lemnios admits that the evaluation of the proficiency of a system and the reliability of system to gain trust in the dialogue is a challenge. “We see this issue of trust and proficiency in our human interactions. Transforming that set of ideas into a man machine computer user interaction is really what the technical challenges are about,” he said.
It’s the Watson jeopardy system that led to the further research and the announcement of the IBM Watson Engagement advisor which is delivered from the cloud and into the hands of mobile consumers, with the ability to crunch big data and provide fast, personalized advice. But a key factor with its predecessor in the game show was the ability to decide whether to answer, and building a level of confidence in its advice.
“Evidence behind the answer is a critical element of what we are exploring,” Lemnios said.
Computers can learn from us, too
Gartner's research suggests that machines and systems can only benefit from a better understanding of human context, humans, and human emotion. This understanding leads to simple context-aware interactions, such as displaying an operational report for the location closest to the user; to better understanding customers, such as gauging consumer sentiment for a new product line by analyzing Facebook postings; to complex dialoguing with customers, such as virtual assistants using natural language question and answering (NLQA) to interact on customer inquiries. NLQA is one of the technologies on Gartner’s Hype Cycle that represent these capabilities.
NLQA technology can improve a virtual customer service representative. NLQA can also be used by doctors to research huge amounts of medical journals and clinical tests to help diagnose an ailment or choose a suitable treatment plan. These supporting technologies are foundational for both humans and machines as we move forward to a digital future and enterprises should consider quantum computing, prescriptive analytics, neurobusiness, NLQA, big data, complex event processing, in-memory database management system (DBMS), cloud computing, in-memory analytics and predictive analytics.
The IT department faces the challenge of running a setup under enormous disruptions where the pace of innovation is only increasing. “Cognitive systems will assist the humans to mange these disruptions and let operate in this environment. This is like the transition from horse-drawn carriages to automobile,” Lemnios said.
This story, "Next up: Humans, systems team in cognitive computing" was originally published by Channelworld-India.