Lauren Biscaldi, editor-in-chief of Topics on drugs, sat down with Lori Schafer, CEO of Digital Wave Technology, to discuss Generative Artificial Intelligence (AI): the ins, outs, and importance of using these technology tools to improve everything from product marketing to patient advice.
This interview has been lightly edited for clarity.
Lauren Biscaldi, Editor, Drug Topics: Can you explain the difference between AI and generative AI?
Lori Schaefer: Excellent question. AI… has actually been around for many years. There are different types of artificial intelligence. When we think about human intelligence, artificial intelligence solves very difficult problems. This could include, for example, looking at drug development; it may involve synthesizing proteins and making this process much faster. This could be about optimizing price levels. It can be any very complex problem that a computer can solve faster than a human, if you will.
Today’s hot topic is Generative AI, or Gen AI, which became very popular earlier this year. The difference is that there are now large models. AI is usually numbers… it deals with a lot of data and numbers. Generative AI looks at text, images, videos and audio, all these different ways of communicating. It can generate new data very quickly; it solves problems by literally coming up with new text, creating its own images, and creating its own videos.
So, is generative AI more like, say, Chat GPT?
So, ChatGPT is a brand, if you will. It is a type of generative AI that works on text. AI is a big umbrella; Generative AI is one of them. It has to do with natural language, so it puts everything into the natural way that you speak or visualize or taste – that sort of thing.
Despite the growing prevalence of AI and generative AI, the technology can still be intimidating to people who are unfamiliar with it or how to use it. In your experience, what is the current role of generative AI in healthcare specifically?
Right now, generative AI is relatively new; it’s the fastest growing technology of all time, in the sense that adoption expectations are, say, 10 years faster than those of the World Wide Web. This was the last big jump. They compare generative AI to this kind of impact on society.
In healthcare, most companies are just beginning to use AI. What we recommend as a generative AI software provider is to start with experimentation: you want the human to be in charge; you don’t want to have generative AI or AI as autopilot, which for some people is where it gets scary. You want it to be the co-pilot, and being the co-pilot makes you much more effective.
Some of the specific areas where we apply generative AI: Think about a website or product content. You have different medications, they may be over-the-counter or over-the-counter. And you have to describe these products; you have to tell a story. You have product descriptions, website content, marketing content, etc. Instead of having to write all that copy, you can simply tell the generative AI: “Tell the product story in our own brand voice.” From there, by literally typing a little prompt, it will come back and create a new copy.
A pharmacist can have thousands of products in a store, like retailers and even B2B websites, and all that product content. Generative AI can automatically generate all of this content in seconds or minutes, whereas a copywriter might have to take half a day to write copy for a few of these products.
Editors are trained for this; pharmacists are not necessarily inclined to write.
Not exactly. Think about applications: we’re just starting to do this, aren’t we? A pharmacist’s best skill is not necessarily writing. But a good editor will use generative AI as an aid, so that they can now start thinking about more important things; let him do the legwork. You create the style and content you want to talk about, then the AI does the writing.
If you think about it, in healthcare, words aren’t necessarily user-friendly, they’re not patient-friendly. I think of a friend who was diagnosed with cancer; the pharmacist has this first line of contact. And so the patient wants to understand: “What is wrong with me?” Why am I taking this medicine? What does it do for me?” We’re currently testing this with pharmacies, telling Generative AI, ‘Write this in plain English.’ That way, the person who receives it… You know, when you get a medication, and this long paper, and the font is size 6, and no one will ever read it, right? Say it in simple English, and then you can educate the person to want to take the medicine .
When I was going into this interview, I was thinking more about the use of generative AI in terms of product and front-end in pharmacy, but it seems like an incredible tool that you can use for patient education opportunities .
Exactly. For marketing at Digital Wave Technology, where I run the company, we have software that will automatically produce content, such as an image; tell it: “Create a stock photo containing XYZ, ABC” and it creates a new one. And now it’s your copyright.
Is there a real use case for generative AI that you can share with us?
Of course. Those that are successfully applied today in mass production are typically in the content/product space. For example, creating a website and determining which product attributes (and thinking of an attribute as a keyword) you need for each product, so that when you search, you find it instantly. This is all done manually on each website; there are a lot of people who need to grasp this. With what we’re doing, it’s now fully automated, which has reduced time and created better content and search score, with a 97% increase in efficiency.
This probably helps on the SEO side as well.
This is a huge SEO help.
I feel like no one understands how SEO works; it’s a bit of a mystery.
Yes, but very few people do it, and I can tell you that it gives your product SEO a whole new perspective, in a way where a human wouldn’t necessarily think that way.
Likewise, consider blogging. As a healthcare business, blog more, not take time to blog. All you have to do is say, “I want a blog about XYZ” and it creates it. You can edit it a little bit and then send it out, and it has all the right SEO words, and it really helps your content.
Another example in pharmacy is all that paperwork. This can automate a lot of this in the future. Look at chatbots: When the pharmacist is busy, you can have a kiosk in the store (or an app on your phone) where a patient can say, “I really need to understand XYZ about my medications.” » The pharmacist can say, “Hello, I’m so-and-so, and this is what we recommend,” and it’s all done through generative AI.
It’s really exciting. It’s a little scary and it’s a little intimidating, and you have to be safe. You have to ensure security and everything else, but that’s the kind of thing we focus on in our business.
If a pharmacist is looking to work with a company, for example your company, Digital Wave Technology, should they come back to AI every time they want to use it. Do you have to retrain it to write with your voice every time?
You train it once. You train it once in your brand voice, and it sticks with it. After training it once, you can use it again and again; It’s a bit like a template in Microsoft Word. You can record a template prompt. Some of these companies have millions of products and are set up to do batch processing: “I want this voice and I want to create one copy for 10,000 products” – boom, done.
And of course, it’s secure; working with, for example, a GPT Chat… it’s not secure, unless you really know how to use it. For the average person, it’s not secure. As a business, you want to benefit from this security.
What practical advice do you have for pharmacists who want to upskill with generative AI, or integrate generative AI to improve their digital presence or in-store offerings, but are nervous about it. What should they do?
I would say a few things. The first is to work with someone – a small or large company – who you trust and who understands. You need the partner at the beginning. It’s almost like a crash course; Generative AI is designed for the average human who is not necessarily a great writer or marketer.
Start very small: install Chat GPT on your phone as an app to try it out. When it comes to business, work with a consultant or software company (someone who does this for a living) so you can at least test it and see how it would work in your particular use case. For example, the idea that a pharmacist could create a chatbot to write this human, friendly description of what the patient is experiencing. These are the kinds of things that are available now, you just need to integrate this software into practice.
It’s scary at first; Some people fear the end result will be job losses. But it’s not. What’s interesting about generative AI… it’s not intelligent artificial intelligence. When you understand how it works, you realize it’s simply about generating more content from the content it’s learned from. There still has to be a human there; this will make humans think more and type less manually.
I always find that as a writer, the first draft is the hardest. It’s so much easier to go back and edit, or modify something someone else created to use for my own purposes.
With generative AI, you think first, and then you say, “I want to write about so-and-so, and I want it tailored this way, and I want it in outline form.” I want it in 1000 words or less. Then you let it go; copy it into a word processor and edit it. This gives a writer the opportunity to be an editor; This is how you should think about it. This will increase your ability to think and do better, not spending all your time with syntax and worrying about commas and grammar. It can create a photo, it can create a video of you talking that looks like you and talks like you, but it’s not you.