Wednesday, May 9, 2007

Conversational machines - Is it a possibility?

There are many researches going around on Artificial Intelligence and Business Intelligence but Fair Isaac's News Bulletin ViewPoints surprised me. It talks about the research going on conversational Machines - a machine that can not only understand on what you say but also act on it, learn from past actions/conversations and predicts your future need.

Well, there is no doubt about the business success of this machine once it comes to market. The only doubt is, whether it is possible?

Dr. Robert Hecht-Nielsen, vice president in Fair Isaac’s Research & Development group is leading this team and is very optimistic of its success. He claims that through confabulation theory, Fair Isaac has developed a system for making sense of speech based on an understanding of the structure of language and the context of what is being said.

Dr. Hecht-Nielsen started doing research in this field as early as 1968. Confabulation theory is fundamentally a neuro-science which offers a comprehensive explanation of the mechanism of thought in humans and animals.

More about the product, here.

I actually started not to explain this product but to discuss how the world would be once this gets successful. It will open a new window to look at the business. There are always a lot of decision points where human involvement is required in today's business environment, especially for strategic and tactical business decision making. But I see this too is replaced by machines.

I go to a bank and apply for a loan. The credit check is done and a credit report is sent (which is automated), which is seen by a executive and processed. I may be given a loan or rejected. This can actually happen today by replacing an executive by decision optiser and letting it contact directly with the customer. But after conversational machines, it is possible to move further. It will move the ladder up replacing the managers to take not only tactical decisions but also for making strategies. This is possible because it learns from past experiences as humans do.

But a big but comes to my mind, how does this machine handle sudden change in business environment? Due to highly competitive state of the present world, this type of change comes quite often. And each of these changes may be very different from the earlier ones, thus feeding past data may not be the solution. Fair Isaac's R&D team may have the answer for this or they are exploring. I am going to wait and see, it is going to be a interesting wait this time. The product coming is really exciting.

All the best Dr. Neilson and Team.

No comments:

Page Views from May 2007


© New Blogger Templates | Webtalks