Artificial intelligence isn’t something that we hope to witness one day, it’s already here.
Apple's Siri is AI, and it’s been on iPhones for 12 years. The autofill on your Microsoft Word program is a form of AI.
But a big leap in the technology came about a year ago when generative AI burst onto the scene. Generative AI allows the AI to use information to not just repeat, but generate and formulate new patterns, text, music and a lot more.
Some of this you’ve heard of: ChatGPT is one. Then at the newly opened Sphere near the Las Vegas Strip, the Aura robot is another.
At the same time, countless AI programs and applications are already in place, unseen but working. But working to do what? What exactly can AI do in the workplace?
Interview highlights have been edited for clarity. Hear the full interview above.
LOSE JOBS TO AI?
BRIGHT: We haven't really seen that so far. How come? I think it's because the people who are doing the stuff that a computer should be doing also have the human stuff as part of their jobs. So a lot of people who are managing spreadsheets will now be managing AI agents, likely. So, you know, it's possible that there will be some reduction in employment, but it's not. It's not what we've seen so far.
For managing a [casino] player's overall experience, and making sure that your guests are having a good time, certainly the humans much better at that. There's always going to be context and information about a hospitality environment that a computer just doesn't have, right? It could be that your algorithm might determine that, hey, the floor is not performing very well. And you should do something about that. But the algorithm doesn't know that there may be a vent dripping water onto machine or there's a bad smell, because there's an open source somewhere. So we really view the AI as being a compliment to the human, not so much as a as a replacement. We really view it as helping the person do their job better.
WHAT CAN AI REPLACE?
HEIN-PETERS: There was a lot of administration and healthcare, you know, we scheduled appointments by calling an office, there is no actual person responding to our phone call, then we fill a lot of paperwork when we come to doctor's office. You know, even when I want to move from one physician to another, now there are places where I can send data electronically, but most of the time, they print out my medical history for me, then I can take this printout and I bring it to another office. I think it's all very mundane, kind of low. These are not jobs that require a lot of skills, and they can be easily automated and they can be easily improved by generative AI.
HOW IS AI USED IN MEDICINE?
HEIN-PETERS: Actually, in most of the applications that have been approved by FDA are in radiology, because AI is very good and reading images. And I think that there are some other in neurology, in in oncology in dermatology. But I think imaging was the first place where AI in medicine actually took off, because AI is so good at reading images.
In general, our current ChatGPTs are the large language models. They are basically monomodal, meaning they either read and generate text, or they read and generate images. Medicine is multimodal, right? When you go to a physician's office, they hear what you have to say; this is the text, then they listen to your heart and lungs; this is the sound. And then they send you to some, let's say imaging test X-ray ultrasound, so that they look at images. So we really need a generative AI for medicine that combines all these inputs, and can generate the output that based on all of these. And these models are under development right now; they are in the testing phase. But this will be the next frontier, so to speak, in generative AI in medicine.
AI DEVELOPING TOO QUICKLY?
LEAVITT: I don't have an answer. All I can say is that you learn from the past. I think sometimes laws are reactive, and policies are reactive. I'm one that's really anxious and excited about AI. I mean, even talking in here and what Steve [Bright] was saying, and what his technology is doing, you know, there has been a problem, at least as I understood it in the gaming industry, where you say, 'Hey, we have guests, and we only have them for two days window, we have a very, very short relationship with those folks. How do we get as much money as we can, through those right? Through those relationships? Can we build a sort of an understanding or a trend and then be able to craft an environment that is susceptible to what their habits are?' And so on. So knowing that you're using technology like ... regenerative AI to make those decision making systems quicker and faster, so we can react or our gaming industry can react or research industry can react or our guesses, I think is going to make a lot better experience for us. And for the state of Nevada, obviously, hopefully make more money out of out of the revenue generated.
Guests: Steve Bright, vice president of data science, OPTX; Joshua Leavitt, CEO, Tech Alley; Kasia Hein-Peters, MD, advisor and board member, Abante Scientific
