Over the last several months, I have been asking CIOs about the technology topics they wanted to explore in greater detail. One topic that scored high on most CIOs list was Chatbots and RPA.
The resulting discussion should be of interest to technology leaders of all stripes because it represents innovation that can take place today.
CIO David Seidl kept pulling apart the two technologies. He said that “these are two somewhat different areas. Chatbots provide or access information and RPA is for processes and procedures. Both have a role in my mind, but RPA success really requires assessing and fixing processes instead of just using old processes. Meanwhile, chatbots are seeing a lot of adoption where information already exists for which the results can be exposed via the Chatbot. This means the real work is about putting information to use.
He continued: This opens up a conversation about the different ways of relating to technology and different uses/needs for ways for getting answers. There's a whole generation of folks who are used to saying hey Alexa/Siri/Google, what is the answer to X and having that answer provided who have had a completely different way of interfacing with the world, and they're coming to colleges and businesses soon.”
While agreeing with the above technology distinctions, CIO Stephen diFilipo, “suggests that both technologies can be used to replace repetitious tasks. Optimally, they create new models for digital experience including customer experience, employee experience, and beyond. Both incorporate today varying aspects of AI and machine learning.”
Other CIOs, however, see a bit less potential in particular for RPA. CIO Martin Davis suggests that “RPA is really more of a tool for digital optimization than for digital transformation. If you have properly digitally transformed your need for RPA should be minimized. Chatbots, however, when done right provide a scalable way to provide customer service and avoid the customer being put on hold for hours.”
CIO Milos Topics agrees with Davis and suggests, “most RPAs today is not AI powered or enabled.” On chatbots, CIO Deb Gildersleeve suggests today “they are still pretty tactical in nature and are best suited for the level 1 questions. In my experience, they are still pretty frustrating for the higher order questions.”
Seidl believes this is “only in the broadest sense. We label so many things AI and machine learning that yes...sure, in the marketing sense. But I'm not sure that many are AI, and that quite a few are mostly matching search terms which isn't really AI or Machine Learning (ML). What you do with the data to get better counts! This makes me want a Turing Scale which rates how smart the BOT/AI/ML is, with some measures of things like natural language processing, personalization, data analysis, false positive/negative, and user response satisfaction.”
Davis says again here that “the two again are very different tools. Chatbots perhaps, RPA no, as a business case prover for AI /ML. I believe RPA is just intelligent scripting, chatbots can be much more if they are done right.” Seidl adds that “it gets interesting to me, because maturity and algorithmic bias might go hand in hand sometimes.
