IBM Builds Project Sapphire With University of Michigan

IBM and University of Michigan are collaborating to make a computer for advising undergraduate engineering and computer science students.

IBM and University of Michigan have announced to manufacture a computer capable of responding to conversation in the same manner as humans communicate with each other.

Project Sapphire is an undertaking worth £3.14 million ($4.5m) that would first see computers advising undergraduate engineering and computer science students on academic issues at the university. The research would find out how smart machineries interact with people and apply the same to more situations to conduct a successful conversation.

The venture would be initiated by collating human-to-human conversation between pupils and their advisors, which would then be used to teach problem solving and natural conversation to computers.

"Natural conversations bring in so many different aspects of human intelligence—knowledge, context, goals and emotion, for instance. In many ways, to build a versatile conversational system is a grand challenge for artificial intelligence," stated University of Michigan professor of engineering and computer science, Satinder Singh Baveja.

Computers would learn by using probability as a basis to reasoning rather than context. Learning, knowledge, emotional analysis and understanding of language could do much well in accuracy, IBM stated.

David Nahamoo at IBM told, "Human-to-machine interactions, similar to human-to-human conversations, are rarely confined to one question and one answer. They involve multiple turns of a conversation with responses that can be imprecise and unclear, making it difficult to simulate the human experience."

Mr. Nahamoo further elaborated that by collaborating with the American university, IBM has achieved the best opportunity to use artificial intelligence technologies beyond the educational sector to transform human to machinery communication.

In other news, the company is launching its cloud-based IBM Dynamic service for retailers to track their rival’s prices and latest market data to adjust their prices accordingly. This would be including factors, such as financial objectives, conversion rates, time of day, availability of inventory, social sentiment, market demand, and weather – all of these are known for affecting shopping habits of consumers.

Dynamic pricing would let organizations track and follow changes in the marketplace and adjust their strategy to make sure they are not left behind, as a large number of online shoppers now expect to get the best concessions by searching around many retailers.

The facility would automatically recommend prices in real-time when it ‘thinks’ that the action needs to be taken. Such predictions can be beneficial not only for retailers but for the entire e-commerce industry if the scope and capacity is enhanced.