AI progress on track, test is to translate it into action: Cherubal

In a panel discussion among experts in the city recently, Cherubal pointed out the problems in introducing AI to help in large-scale operations, such as governance. “One of the major problems that we face is lack of data. Even if data exists, it is in bits and pieces. Secretaries and commissioners need insights to govern, not data,” he said.

“Now the interesting thing is, because of smart cities – we have been installing a lot of CCTV cameras and smart meters among other devices – and adoption of technology, we are going to be flooded with data without knowing what to do with it,” said Cherubal adding that it would be easy to get overwhelmed with it. “The situation will be equivalent to drinking from a hosepipe. And no human brain can make sense of all this data,” he said.

According to him, that’s when AI can step in to process the data and come up with insights. But, this also poses a challenge as every AI expert would have different ideas about how to solve these problems with the bigger struggle being implementing these ideas.

“Consider this, the city is trying to increase its revenue. With regard to property tax, there is a section of taxpayers who will always try to shirk paying up. Now, we do not need AI to figure out that people owning commercial buildings are not paying property tax properly. The scenario is the same when it comes to paying for electricity. But, the problem is, electricity board has its own divisions and parameters which is different from other departments such as the property tax department,” Cherubal said.

“If you ask them for information on customers who are cheating, you have a lot of data, which does not make any sense and in different database management systems such as FoxPro and Microsoft Excel. Nobody has the time to sift through the data and compile a formative chart. This is where AI —in terms of finding out what is needed — can come handy. By analysing the data, it can figure out who is violating property tax. The process is similar to Alexa trying to find out our favourite songs,” he added.

“The possibilities are endless. The problem is the capacity of work that has been going on in the field of AI inside universities or boot camps is so low that it is absolutely frustrating to translate it into action,” said Cherubal.

“We can, however, be inspired by the fact that it is only a matter of time before we reach the tipping point when we figure out the scope to address our challenges,” he added.