- What it is. ChatGPT is a form of generative AI that creates new content in various forms, including text, audio and images by understanding context and relevancy. It has potential applications in education, marketing, ecommerce, legal and software development.
- Concerns about generative AI. While ChatGPT and other forms of generative AI have significant potential benefits, there are also several concerns to consider, including the potential for inaccurate information, dangerous propaganda, cybersecurity threats and the replacement of human workers.
- A new transformation with serious caveats. Despite the concerns, the rise of generative AI represents a significant transformation in how we discover, search, create and work. It has the potential to enable a new level of productivity, but it will require careful consideration of ethical and regulatory issues.
I’ll spare you one more article that begins with “This article was generated using ChatGPT,” since in actuality it was created entirely by a human. ChatGPT debuted only a few short months ago, and it has already taken the world by storm. But as with any shiny new object, it’s important to take a step back from the hype, assess the benefits, weigh the challenges and slowly carve out its role in how we work and create.
ChatGPT, created by OpenAI, is a form of generative AI, that unlike traditional AI systems that are trained only to recognize patterns, creates new content in the form of images, text, audio and other content by being programmed to understand context and relevancy. It’s essentially a chatbot on steroids, generating brand-new content that is humanlike.
Some of the industries where ChatGPT can have great benefits, include:
- Education: ChatGPT can help students develop essays or complete homework assignments. For example, a high school student can ask it to create a haiku poem about nature for a poetry class or develop a report on the first settlers in Puerto Rico. Within minutes pretty good drafts are created.
- Marketing: These systems can help jumpstart marketing campaigns, creating taglines, developing blog posts or other types of content. Simply provide the goal and key messages and it’s on it, providing a pretty good place to start.
- Ecommerce: Much like other chatbots, ChatGPT can play a role in customer service, but taking it to a whole other level, answering customer questions, providing more information on the company’s products and services and anticipating new questions.
- Legal: ChatGPT can be used by lawyers to gather data on prior court proceedings and develop oral arguments or even by judges to decide on cases. It also could be used by individuals to assist in the creation of a living will.
- Software Development: Software programmers are already using AI to generate code using Microsoft’s GitHub Copilot, which employs OpenAI. It helps developers write and fix code, and Microsoft estimates that Copilot generates 40% of the code, according to an article in the Wall Street Journal.
Related Article: ChatGPT Is Already Replacing Humans in the Workplace
ChatGPT Gets Some Competition
While Microsoft, which recently announced a $10 billion investment in OpenAI, is working to integrate ChatGPT into its solutions, including its search engine Bing, other major players are exploring new forms of generative AI. For example, Baidu a Chinese internet search engine company is developing a generative AI-driven solution, which is expected to launch in March 2023. Other competitors with solutions in the works include Sparrow, developed by DeepMind, a British unit of Google’s parent company Alphabet. Google is also testing a ChatGPT rival called “Apprentice Bard,” which uses LaMDA, a model developed by Google.
There’s no doubt that more and more players will be entering the ring, refining and expanding upon the generative AI capabilities of their solutions. As more and more users play around with these types of bots and share feedback, they will continually improve them. Today, however, there are precautions and challenges to be overcome given the nascence of the industry.
In fact, according to the New York Times, the CEO of OpenAI, Sam Altman himself, has tried to set realistic expectations, offering a disclaimer that ChatGPT’s capabilities are still relatively limited and warning users about the potential for inaccurate information. To that end, JP Morgan is restricting how employees use ChatGPT, and Verizon and other companies have banned its use altogether because of concerns about inaccuracy, as well as sharing customer data.
Related Article: Humans Put ChatGPT Customer Experience Outcomes to the Test
The Caveats to Consider With ChatGPT
While inaccurate information certainly is cause for concern when, for example, you’re using a bot to convict a defendant or to explain to a patient the common causes of a medical condition, there are other concerns that need to be addressed as well, to ensure the safe use of ChatGPT.
- Dangerous propaganda: Taken to the extreme, millions of ChatGPT bots could be programmed and deployed to appear as humans or deep fakes, having conversations designed to convince unwitting people of a particular point of view or to share misinformation.
- Cybersecurity: Given its ability to sort through massive amount of data, generative AI has the potential to make it easier for bad actors to more easily create malware.
- Replacing workers: Especially for industries, such as marketing, customer service and software development, there are serious concerns about ChatGPT’s ability to do the work of many humans. This is not a new concern, given the level of automation already occurring in the industry, yet the sheer size, ease of use and power of ChatGPT, means that the man-versus-machine debate can go into overdrive.
- Copyright concerns: When someone works off of content prepared by ChatGPT, how can they be sure they are properly citing the source? According to Bern Elliot of Gartner, “the model is trained on a corpus of created works and it is still unclear what the legal precedent may be for reuse of this content, if it was derived from the intellectual property of others.”
A Transformational Era Driven by Artificial Intelligence
Yet, aside from the more concrete concerns about ChatGPT and other types of generative AI, there still remains the issue of creativity and genuineness. When students are using it to generate research and written content, are we creating a new generation that will have no idea how to conduct their own research, form their own opinions and apply it to a written piece of content? Will human exploration and discovery go the way of the encyclopedia or the library, or will it simply enable them to gain even more knowledge faster, much as the Internet has enabled in general?
Even if it provides a faster route to knowledge and content creation, will everyone’s written creations, prose and poetry start sounding the same, with one word being used because it follows the highest statistical probability from the previous word?
These are all issues that need to be addressed, and clearly regulatory guidance will need to be developed about the role transparency must play when using generative AI. But aside from government intervention, it also will require corporate, public and individual ethics.
But regardless of what we think of generative AI today, there’ no doubt that we’re on the precipice of a transformation in how we discover, search, create and work. It can most certainly enable a whole new level of productivity, but as with all automation, humans must continue to drive the bus.
Learn how you can join our contributor community.