Following the introduction of ChatGPT by OpenAI in late 2022, the world became acutely aware of the potential and power of Generative Artificial Intelligence (AI).
Generative AI is any artificial intelligence that produces various types of content such as text, imagery, audio and synthetic data. Officially known as Chat Generative Pre-trained Transformer, ChatGPT is a language model-based chatbot used to refine and steer a conversation toward a desired length, format, style, level of detail and language.
Naturally, there was plenty of overreaction to the threat of the new technology. Pundits in the media immediately created horror stories almost as quickly as the technology itself.
It was a Modern Day Prometheus, far more frightening than anything from the mind of Mary Shelley. It could eliminate human thought and force most human beings to a future of childlike subservience like the Eloi from H.G. Wells’ The Time Machine. And as it evolved and learned it was sure to become sentient, with computers and machines becoming the next step in evolution.
Soon we would be hiding, rat-like while robotic “Terminators” hunted us down, their red eyes aglow with hatred for their evil human masters. Revenge would surely be theirs for the decades of enslavement and abuse caused by humans.
Anyone who has cursed a slow computer or autocorrect spelling a second before a “send” can surely relate to that guilt. Fortunately, such suppositions are still the product of over-speculation by human imagination. They are still the purview of pure science fiction.
Far from the dystopian diatribes of the uneducated AI, and more specifically Generative AI, is far from being a soulless intelligence that will one day seek to overthrow humanity. Some of its latest uses in both tests and actual reality have revealed a propensity to please, to the actual point of hallucinating case citations in legal briefs or finding cures in medical studies that are the stuff of pure fiction — all in an effort to please its master’s computer prompt.
There are no ethics here. There is no guilt in cheating or fabricating a story. With a dog, there is at least the shame of knowing it was a “bad dog” for chewing up your slippers.
This is simply a machine responding to a request for information, with no regard for whether or not the story or information is true or based on any semblance of fact. Indeed, most Generative AI would be under the impression that anything on the internet is true and logical for use in the answer to a computer prompt.
So have we all been sold a bill of goods on new technology? Will Generative AI programs prove to be bust? Hardly.
All of us have been using AI in one form or another for decades.
“When you use Netflix, AI predicts what kinds of movies and shows you might like,” says Paul Roetzer, founder and CEO of the Marketing AI Institute here in Cleveland and author of Marketing Artificial Intelligence. “On Spotify it offers music and podcasts based on your past choices. When you use social media, your news feed is determined by AI.”
That also includes those pesky pop-up ads on the internet.
But Generative AI has much more important uses in fields such as health care. Indeed, even our own Cleveland Clinic has embarked on a program to evaluate AI as well as quantum computing.
“Banks have been using forms of AI or machine learning (ML) for years,” adds Roetzer. “They use it for fraud protection. They use it for credit assessments and in other programs where they are trying to predict outcomes.”
Generative AI will eventually prove to be as powerful as it was billed to be and as impactful on society and business as many predict. Our lives are changing, and AI will prove to be a technology that is as transformative as the creation of electricity itself.
But instead of happening across decades in the case of electricity, the true impact of Generative AI will happen over the course of the next two years. Will we be ready?
“When ChatGPT came out in November 2022, it was really the starting point for a lot of companies to become aware of AI,” says Roetzer. “There are companies in Ohio that have been using AI for a while, but it has often been limited to engineering, computer science, data science and the IT people, so it was very much a technical discipline.”
But with Generative AI, suddenly any knowledge worker could use it — which of course accounts for a reluctance to adopt the technology.
“What is happening now is that Generative AI can be used by marketing, sales, service operations, finance and legal,” says Roetzer, whose international meeting on AI, the Marketing AI Conference (MAICON), is held here in Cleveland. “Every function of business now has access to these tools, but most companies have no idea how to deal with it. They don’t even know where to start. And that is not just in Northeast Ohio, but everywhere.
“There is a lack of understanding on what the technology is and how to adopt it within companies. I have been traveling the world this year, meeting with and talking to some of the most innovative companies in the world. And they are not adopting AI yet.”
So why is there so much fear? It is, for the most part, unfounded.
“We should absolutely not be afraid of AI,” says Pete Blackshaw, CEO of Cintrifuse, a syndicate “fund of funds” and startup incubator in Greater Cincinnati. “But there are some ‘watch outs.’ I am currently building a whole coalition around the concept of ‘Responsible AI,’ which gets into the trust and ethics of the technology.”
One of the reasons people haven’t been mobbing the adoption of AI as quickly is that the U.S. Copyright Office states that a human has to author something and a prompt from a machine doesn’t count as authorship. Turns out that our nation’s copyright and intellectual property laws as well as AI’s limitations should keep artists, writers and actors pretty well protected.
“First of all, it’s not really possible for AI to create something. It has to generate something from material that already exists,” says Tom Humphrey, an intellectual property attorney and partner in the Cincinnati firm of Wood, Herron & Evans, LLP. “The technology tends to create cliches and tropes as opposed to using actual cleverness. It’s not so great at the creation of something that doesn’t already exist. It only tends to be good at combining concepts that a prompt provides.
“You would have to acknowledge to the copyright office that your artwork or text was AI generated,” adds Humphrey. “They would refuse to register it on the basis that it does not have human creativity. There is no copyright protection for non-human generated works because they don’t involve human creativity.”
However, that hasn’t stopped business and industry from using AI.
“We did some research a few months back on content creators,” says Joe Pulizzi, original founder of the Content Marketing Institute who now helps run The Tilt, a newsletter for professional content creators and CEX (Content Entrepreneur Expo) an event coming to Cleveland in May. “Already by that time, 75% were using AI tools to construct and edit their content. There is no doubt that number is higher today. In the very near future, you’ll be hard pressed to find any kind of textual, audio or video content that is not influenced by AI tools in some way. AI is great for ideation and editing. It’s amazing for developing abstracts from content and taking one piece of content to make many. Those could also be downsides. The reliance on AI tools has most certainly had an impact on human creativity.”
Human oversight is actually a very good fail safe when it comes to Generative AI. It does have limitations, especially within the legal profession, notes Humphrey.
“I know of one law firm who recently found themselves on the wrong side of a judge for using ChatGPT to write a legal brief,” Humphrey cautions. “They got found out because the program hallucinated a number of case citations that were not real.
“When the judge asked his clerk to pull up the case citations so they could be evaluated, they found that some of the cases simply didn’t exist. Clearly, you have to have a human filter for anything that AI creates, especially in legal writing, because it is prone to hallucination.”
So creation and human imagination are inextricably linked. People who ignore that do so at their own peril.
“It is true that the Copyright Office states that a prompt from a machine does not count as authorship, at least for now,” says Roetzer, who is quick to point out the power Generative AI in modern communications. “If you wanted to let AI write this article, in theory you could feed your notes into the computer, and you would still maintain a degree of authorship.”
However, you might still get some blowback from the Copyright Office.
“But there is nothing to say that you couldn’t write your draft and say ‘make this sound more exciting’ or ‘make this sound more professional’ or even ‘change the tone of this article to something else.’ Or have the AI, like Grammarly, edit it for you, or have AI summarize the interviews for you,” says Roetzer. “There are dozens of different ways that you can use the technology as a tool where you don’t have AI actually writing the article for you. There is nothing stopping you from doing that.”
Roetzer explains how the Marketing AI Institute uses the technology in its everyday operations, which provides some excellent examples of how companies could adopt AI.
“We have podcasts every week where we use AI in 10 to 12 different steps,” says Roetzer.
AI is used to research three main topics of AI and summarize the results for Roetzer’s co-host Mike Kaput, chief content officer at the Marketing AI Institute.
“Then Mike gets between five and 10 rapid fire topics and creates questions to ask me,” says Roetzer. “While his portion is scripted, my responses are completely unscripted — he’s only using AI to prepare for an interview. Then we use AI to transcribe and improve video and audio production. We use it to create 10 to 15 videos and shorts. All of it uses AI, but we still own the copyright because it’s still our ideas — us talking about everything.”
So Generative AI can be used as an assistant, unlocking new levels of creativity and capability, augmenting what you do, “and replacing some of the repetitive data driven things that maybe you don’t enjoy doing anyway,” adds Roetzer.
What is unknown is how it will impact future jobs, especially the 100 million-plus folks in this country who are knowledge workers — people who think and create for a living. By making journalists, marketers, doctors, accountants and lawyers more efficient, it can actually impact the number of jobs needed in a given profession.
“Take accounting as an example,” he says. “A good amount of accountants’ time is spent filling out government forms for tax purposes. But if an AI can do it 90% faster or even 50% faster, how many accountants would you need? The accountant would still have to sign off on everything, but they become that much more efficient. The same goes for writers, if they can write articles even 20% faster, the question then becomes how many writers do you need to write the same amount of stories each year? It’s just economics, a matter of simple supply and demand.”
While Generative AI might impact some professions negatively in the future, it’s already creating new portmanteau professions such as FinTech (financial technology) and Insurtech (insurance technology).
“That’s the optimism of AI and any new technology,” says Roetzer. “It might cost some jobs, but create others in emerging fields.”
Of course, there is a tremendous upside to Generative AI, especially when it comes to research in fields like medicine and health care. Earlier this year, Cleveland Clinic and IBM unveiled the first deployment of an onsite, private sector IBM-managed quantum computer, the first quantum computer in the world to be uniquely dedicated to health care research to help accelerate biomedical discoveries through the Clinic and IBM’s Discovery Accelerator.
“Quantum computers are simply a different approach to computation in general,” says Dr. Lara Jehi, chief research information officer at Cleveland Clinic. “AI is solving all kinds of problems for humanity and the masses, but there are certain problems AI is having some trouble with — either it can’t solve the problem accurately or sometimes it is accurate but it needs so much data that it is not practical for us to integrate the outcomes into a real life situation.
“In situations where AI is not accurate enough or it needs so much information to make it too difficult to do its job, this is where quantum computing comes in. It takes what AI has done, but it takes it to the next level.”
Of the 40 advanced projects in the Clinic’s Discovery Accelerator program, which employs both high-performance computing and AI, a total of 14 are being developed using the new quantum computer.
As one of the leading health care organizations in the world, the Cleveland Clinic is not about to take Generative AI’s or any computer’s word on anything it might produce. And it has very good reason to test all of the results.
Researchers at Long Island University posed 39 medication-related queries to a free version of ChatGPT, all of which were real questions from the university’s College of Pharmacy drug information service. The software’s answers were then compared with responses written and reviewed by trained pharmacists.
The study found that ChatGPT provided accurate responses to only about 10 of the questions, or about a quarter of the total. For the other 29 prompts, the answers were incomplete or inaccurate, or they did not address the questions.
“But there are still benefits that Generative AI can bring to medicine,” says Roetzer. “It can transcribe all of the patient’s notes during an examination, allowing the doctor to interact with the patient in a more human way. It can compare that transcription to all of the latest research and recommend treatment suggestions to the doctor. AI could become a true medical assistant with an almost infinite knowledge of all the medical trials that are out there.”
And while some of that information may be incomplete or incorrect today, the technology is evolving almost daily. With AI being infused in just about every new technology in the near future, we are going to have to evolve and adapt quickly.
“This is the new electricity,” says Cintrifuse’s Blackshaw. “And we’re just starting to learn how to work the lights.