When we swipe a credit card to pay for a product or service, we expect a response from the financial institution, which in turn needs to assess whether the transaction is fraudulent or not before accepting it. All in less than a millisecond.

This is a great example of how artificial intelligence can be applied at the edge to process data quickly and efficiently to enable business decisions, according to Seth Dobrin (pictured), global chief AI officer at IBM. And this use of AI tends to grow.

“I often get asked: ‘What do you see as the future of AI?’ And my response is the future of AI is edge,” he said. “The reason for that is, if I can solve for an edge use case, I can solve for every use case between the edge and the data center.”

Dobrin spoke with Dave Vellante, host of theCUBE, SiliconANGLE Media’s livestreaming studio, during IBM Think. They discussed AI applications at the edge, the bets on augmenting intelligence to solve critical problems, the essential characteristics for AI models, and AI use cases. (* Disclosure below.)

The need to trust AI

The quick response in the case of the credit card example cited above is possible because the AI ​​models are deployed on the mainframe where transactions are taking place. That way, the mainframe does not have to call out elsewhere to score the model and then wait for a response; everything is done directly at the edge.

“This gets back to the power of hybrid cloud,” Dobrin stated. “We’ve actually done this with a couple of banks where we’ve trained models in the cloud on GPUs and done the inferencing and scoring on the mainframe for just exactly that, for fraud.”

While AI has the promise of providing valuable insights across multiple applications, its wide adoption will largely depend on the ability of businesses to trust its outputs, according to Dobrin. It is necessary to understand how and why the systems reached a decision and that, ultimately, that decision is protecting the data of the end users.

“Trusting the AI really means five things … the AI needs to be transparent, the AI needs to be fair, it needs to be explainable, it needs to be robust, and it needs to ensure privacy,” he said. “And without those five things all combined, you don’t really have the ability to trust your AI either as a consumer or even as an end user.”

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of IBM Think. (* Disclosure: TheCUBE is a paid media partner for IBM Think. Neither IBM, the sponsor for theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE