As we continue to integrate Gen AI into our work lives, we should also be cognizant of the next wave—conversational AI.
By now, the secret is out on generative AI (Gen AI). The technology can consume massive amounts of information in seconds, and then provide insights or create new content to users who submit queries. Gen AI finds patterns in the data it consumes to create its responses.
It’s already being used across industries to distill large loads of information into digestible bits, to spur creativity, and to aid productivity by taking away repetitive tasks.
But as its use becomes more pervasive, I challenge people to stretch their understanding of artificial intelligence (AI) beyond generative. Because when we solely focus on Gen AI, we miss an opportunity to understand where AI is headed—and how it can positively impact our lives and businesses in the future.
TOWARD CONVERSATIONAL AI
As we continue to integrate Gen AI into our work lives, we should also be cognizant of the next wave—conversational AI.
Whereas Gen AI focuses on recreating the patterns it notices in the data it consumes, conversational AI is much closer to what the general public thinks about when they hear the term AI—a machine that can use inference and reasoning to “think” and respond similarly to a human being. It can learn an individual’s preferences over time and personalize its responses based on that knowledge. That creates powerful opportunities for businesses to create a dynamic and seamless experience for their customers.
AN AI THAT UNDERSTANDS YOU
It’s helpful to look at one illustrative example.
Imagine this: You’re traveling to a new city, you’ve just touched down, and you’re hungry. You decide to turn to AI for a recommendation. A generative AI engine offers you a list of 5-10 highly rated restaurants. But much of it does not apply to you. You can’t eat sushi because of a fish allergy, let’s say, and lately, you’ve been steering toward veganism. Gen AI does not know these things. You could tell (or “prompt”) it, of course, but you’re hurrying between baggage and the stand for an Uber. You’d also have to tell it again next time you’re ready to order.
Conversational AI, on the other hand, does not need this added context. You’ve been having an ongoing conversation with it about restaurants in each city you’ve visited, and now it has dialed in its view of your eating habits. So it pops up three vegan restaurants within a short radius of your hotel. As you revisit similar spots during your travels, it learns even more about what you like.
That’s the power of conversational AI. It makes inferences. It reasons. And it evolves with you.
BEYOND THE FEARS ASSOCIATED WITH AI
As we sit today, a lot of people are still very much on the fence about AI, generative or otherwise. The reasons are many, and there’s a highly justified concern regarding data privacy concerns.
Frankly, most of the fears are warranted. This is a new technology, and we’re working out the implications. There are also experts and scholars pointing out the proverbial Pandora’s box with respect to algorithmic fairness, a term Dr. Joy Buolomwini describes in explicit detail in her recent book Unmasking AI: My Mission to Protect What is Human in a World of Machines. I also see parallels between AI deployment and the cycle that the Internet of Things (IoT) went through a few years ago. IoT had all the hype of AI, which meant companies moved so quickly, we forgot to put in place the appropriate privacy measures to secure our businesses, our personal data, and in the case of AI, our personal safety.
However, we have an opportunity to get things right this time around, and the first step is to lean in on education. When we don’t have a clear understanding of how our data is being used and stored, we’re susceptible to manipulation and abuse. There are early AI tools, for instance, that were built on data that did not belong to the developers. We should instead require that AI companies are transparent about how they’re using customer data, while allowing customers to opt in or out. The time to press political leaders on these safeguards is now; it’s much easier to build data privacy measures into the development of new AIs than to go back and patch the holes.
I’m optimistic we’ll figure it out. The potential of AI is just too big not to. As we work past security hurdles, we can recast our AI discussions around the ways the technology will help us grow more productive.
That might mean helping you pick out the right restaurant. It could also mean so much more. Your AI-powered in-car navigation knows you detest driving on the highway during rush hour, so it automatically sets your path along the side streets when things are backed up. Your AI scheduling tool learns your meeting preferences and optimizes your schedule for productivity. The use cases are infinite and exciting.
https://www.fastcompany.com/91000869/conversational-ai-will-be-the-powerful-successor-to-generative-ai
