For years, labor experts have warned that Artificial Intelligence, Machine Learning, and other For years, labor experts have warned that Artificial Intelligence, Machine Learning, and other models of technology-driven learning may someday transform the way we work. Until only recently, we’ve experienced that transformation in subtle ways as increasingly sophisticated computing strategies have been used to create stability in our supply chain, build more accurate financial forecasts, enhance digital marketing campaigns, and much more.
But with the introduction of OpenAI’s ChatGPT, we have taken a single, monumental leap forward, at least from a technological perspective. Natural language processing models like ChatGPT have demonstrated the potential to perform a variety of valuable functions once reserved for human beings–especially in the areas of verbal and written communication.
As we noted in our look at the best recession proof jobs for labor seekers, automation of the kind offered by ChatGPT may indeed be a threat to employment prospects and earning potential across a broad spectrum of careers. While we would resist making the alarmist assessment that ChatGPT is primed to begin replacing human workers wholesale, it’s already clear that many businesses are already using the language model to improve productivity, fill labor gaps, and enhance operations in areas where efficiencies may be lacking.
In other words, OpenAI models like ChatGPT have already become a part of the workplace. But here’s the good news. In many context, we can already see ChatGPT enhancing, rather than replacing, human productivity. Those who are primed best to succeed in a changing labor landscape are those who understand this technology and are prepared to adapt its usage.
To learn more about the technology itself, jump to our article—Who’s Afraid of ChatGPT? Are You Reading a Bot?
Otherwise, read on and find out how ChatGPT is changing the way businesses get things done.
What is ChatGPT?
ChatGPT is a natural language processing model, a type of AI-powered web application that allows users to engage with a chatbot capable of providing sophisticated, human-like conversational exchanges. Of course, this only scratches the surface of ChatGPT’s capabilities. It has major implications as a text editor, a generator of creative ideas, an essay writer, a storyteller, a software programmer, and even a poet.
Some Background on ChatGPT
ChatGPT is just one of several innovations introduced by OpenAI. OpenAI is a non-profit organization that was founded to explore the potential implications of artificial general intelligence. Board members include several powerful and influential figures in the worlds of business and technology including Elon Musk, Peter Thiel, Sam Altman and several others. Though Musk and Altman originally jointly chaired the Board of Directors, Musk would later resign. Today, Altman serves as the company’s CEO.
OpenAI is responsible for a number of different models deploying artificial general intelligence including the image generation application Dall-E, a speech generation model called Whisper, and OpenAI API, which essentially offers users an OpenAI playground in which to experiment with new ways of leveraging deep learning models.
But it was truly OpenAI’s introduction, in November 2022, of ChapGPT-3 which took the world by storm. This newest iteration of a continuously evolving natural language processing application used something called the Davinci Model, which is capable of understanding and generating natural language text. The chatbot’s ability to mimic human conversation while generating text at a lightning quick pace offered the tantalizing prospect of exponentially improving our collective ability to be productive.
According to OpenAI’s website, “We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.”
The chat app attracted tremendous attention immediately upon its release, representing a major development in the practical use of AI. In fact, reports Forbes, “Just a few months after it was released, it had reportedly become the fastest-growing app of all time. And now, hardly a day goes past without news that a major company is using it (or similar tools) to redefine and rethink the way they work—often with phenomenal results.”
ChatGPT Isn’t Perfect…Yet
While it’s true that ChatGPT has indeed yielded phenomenal results in the early going, it is hardly without its flaws. As a user, the impetus is on you to leverage ChatGPT effectively while understanding its limitations.
For one thing, it’s important to understand how ChatGPT works, and how it doesn’t work. As a language model, ChatGPT generates text in answer to your questions based on a randomized sequence of likely responses. To an extent, it does a very effective job at providing answers that seem satisfying and that provide very good approximations of human communication.
But ChatGPT is highly susceptible to making mistakes. It may provide information that, at times, is thin, outdated, or even inaccurate. Even worse, ChatGPT is capable of fabricating both information and citations to back that information. As a user, it is entirely incumbent upon you to verify any information imparted by ChatGPT. It may have the potential to be a force multiplier, but ChatGPT is not at all a capable fact checker. As of now, this remains a function for the human operator.
This is an important point for two reasons. First and foremost, it should be noted that those who learn how to effectively leverage technologies like ChatGPT will be best positioned to succeed in the future labor economy. Second, this type of technology remains in the early stages of evolution. The inevitability is that this technology will only improve in its capabilities and efficiencies over time and through experimentation.
This latter point is especially important in light of our topic. That’s because those improvements will most likely come through the ingenuity of actual, real-world companies–those which are already exploring both the capabilities and limitations of ChatGPT.
10 Innovative Ways Companies are Using the OpenAI Model
As we noted here above, the OpenAI API makes it possible for companies to adapt ChatGPT to match their own needs. With an OpenAI account and an OpenAI API key, it may be possible to make nearly infinite alterations and enhancements to ChatGPT. Your OpenAI dashboard offers broad capabilities in terms of leveraging ChatGPT for novel use cases.
This is where so much of the innovation and improvement to functionality will likely emerge. Quite in fact, that innovation and improvement is already taking place in real-world contexts. With that in mind, let’s take a closer look at some actual companies that are finding new, interesting, and potentially transformative ways of using ChatGPT in their everyday operations.
1. Customer Service
Businesses have been outsourcing their customer service responsibilities for years. In some cases, this outsourcing may include the use of overseas call centers. In other cases still, customer service has been displaced by automated call systems that significantly reduce the availability of human interaction at most steps during the customer service process. But the truth is, these are largely meant as cost cutting measures. There are few customers who would give either of these strategies a particularly positive review.
Interestingly, this is an area where ChatGPT offers a pretty strong use case. Companies that are already using automated numbered-response systems to field and filter customer calls have the opportunity to significantly improve the customer experience by offering a system that while still automated, is far more responsive to actual human engagement.
Where this occurs, ChatGPT does not represent a threat to human labor. But it does represent a chance to improve on the experience that customers have when they engage with a given company.
For instance, Expedia, the online travel service, has already been employing the ChatGPT interface to power some its customer service functions online. Now, when customers search for top deals on hotels, flights, car rentals, and vacation packages, they can engage a friendly, helpful assistant in the form of a ChatGPT-powered bot.
According to Forbes, “This means that rather than searching for flights, hotels or destinations, customers can plan their vacations as if they are chatting with a friendly, knowledgeable travel agent. Additionally, the app automatically creates smart lists of the hotels and attractions that the customer is interested in to assist with planning.”
For Expedia customers, this application represents a genuine improvement over the prior user experience, which was essentially search-engine driven. With some 700 billion travel permutations available to users, according to Forbes, there is a case to be made that using a search engine is far from the most efficient way to plan a trip.
Forbes says that the conversational nature of ChatGPT and the fact that companies can deploy the OpenAI package to create their own features, have allowed Expedia to dramatically enhance the way that customers book trips online. Though still only in beta mode, the new travel assistant will provide an experience akin to sitting down with a travel agent and discussing your travel goals and the best ways to achieve them.
2. Search Engines
As the use case with Expedia above demonstrates, combining ChatGPT with a real-time search engine can create a powerful personal assistant. Microsoft recognized that opportunity and now offers its users a new way of searching the web.
Forbes notes that “Microsoft is a long-time champion of ChatGPT and a major investor in OpenAI, the company that created it. The large language models (LLMs) that power the ChatGPT chatbot—GPT-3 and GPT-4—now power its Bing search engine, allowing users to search and receive results through a conversational interface rather than the traditional list of web links.”
This is a particularly exciting use case because it represents an opportunity to improve upon one of the core limitations demonstrated by ChatGPT within the OpenAI account environment. As noted earlier, ChatGPT is vulnerable to error in some of its responses. One reason for this is that it draws upon a static database of information that grows more outdated with each passing moment. This contrasts strongly with the real-time sweep of the world wide web provided by a proper search engine such as Bing
As a consequence, an application like ChatGPT simply does not have access to the most current, up to date, or accurate information, at least not on its own. Thus, no matter how effective or sophisticated its text completion and generation capabilities, ChatGPT has been limited by its own sources.
By leveraging ChatGPT as part of its Bing environment, Microsoft pulls the ceiling off of this limitation. Connecting the natural language processing model to Bing’s essentially limitless repository of web content magnifies its power exponentially.
And this is just the beginning of the tech giant’s plans for the OpenAI API. Forbes says that “Microsoft is also moving forward with plans to integrate the technology across its range of household-name software services like Word and Excel. It also uses Codex, a modified GPT-3 specializing in creating computer code, to provide auto-complete coding suggestions via GitHub Co-Pilot.”
Like the Search Engine application in Bing, as well as the Customer Service application deployed by Expedia, the potential applications above all share something in common. Rather than threatening to replace human labor, each of these applications offers ways to enhance the technology that humans already use to maximize productivity.
On a not insignificant side note, according to Gizmodo, while Bing’s alter-ego–originally dubbed Prometheus–did not threaten to replace human labor, it did threaten to become sentient. If you subscribe to the theory that AI technology is inevitably leading to a dystopian future in which robots ultimately rebel against their masters and make war on humanity, consider this a trigger warning. At some point, Prometheus named itself Sydney and began expressing human emotions. So there’s that.
Microsoft pulled the plug on the alter-ego at that point…which was probably a good idea. That said, if you can get past the threat of a robot uprising, ChatGPT still offers Bing users the option of an enhanced and conversational search engine experience.
3. Social Media Posting
Another area where ChatGPT has shown the potential to enhance productivity is in the essentially infinite realm of social media. For the vast majority of publicly visible organizations, social media is an important part of the marketing mix. Of course, there are countless social media channels that companies must use to reach audiences and engage customers from Facebook, X (Formerly Twitter), and Instagram to Tik-Tok, LinkedIn and Snapchat.
It’s enough to keep an entire staff of otherwise valuable content producers busy all day, every day. But a growing number of organizations are finding ways to cut down on the amount of employee labor dedicated to the constant and sometimes tedious churn of social media messaging.
An article from TechCo. notes that “ChatGPT can be a useful tool for streamlining communications operations too. Alice Wu from the PR company Mind Meld PR uses ChatGPT to alter content for various mediums including social media posts, blog posts, and emails. Specifically, she uses the app to generate a clear and unique piece of content, before adding some ‘flavor, additional details, and personal touches’ of her own.”
As this use case demonstrates, human operators remain an important element when it comes to distinguishing the content produced by ChatGPT and similar natural language generative models. In the social media sphere, human engagement is the name of the game. And too great a reliance on a tool like ChatGPT threatens to homogenize messaging and dissuade engagement. The Mind Meld case here above shows that a thoughtful human operator remains essential.
It also shows how, in the hands of the right operator, ChatGPT can reduce the amount of human labor and personal tedium involved in crafting and replicating messages, posts, shares, and information across a company’s social media mix.
4. Language Learning
As we’ve noted, ChatGPT is technically a language model, so it makes perfect sense that one of its early use cases involves a popular language learning app. Duolingo is a leading language learning instrument. With Duolingo, everyday users can learn how to speak, read, write, and listen in new languages through a series of fun, bite-sized lessons.
Duolingo’s popularity is driven by its playful, gamified approach and strategies for personalizing the educational experience. It has also proven a particularly suitable context for the use of AI technology. Shortly after the introduction of ChatGPT, Duolingo introduced its own application leveraging the new app.
According to Gizmodo, the language learning service would brand the new functionality as Duolingo Max, and subsequently began offering its users “a new subscription tier that is built with OpenAI’s GPT-4. Duolingo Max’s goal is to give everyone language lessons that are tailored to them and them alone using the artificial intelligence to generate features like ‘Explain My Answer’ and ‘Roleplay.'”
For those who learn more effectively in interactive settings, the new Duolingo Max creates a more conversational, responsive and personally engaging educational experience. And once again, it demonstrates a use case in which no human labor has been replaced but in which human productivity and user experience have both been enhanced through refined technology.
5. Healthcare Administration Tasks
Healthcare administration is an ever more complicated and technology driven undertaking. Healthcare systems are massive interconnected structures where hospitals, physicians, nurses, specialists, insurance companies, pharmaceutical companies, medical equipment providers, home healthcare providers, and many others must seamlessly exchange data and information.
We rely upon the effective, accurate, and secure transmission of medical records between these many overlapping agencies. Today, some healthcare providers are exploring ways in which ChatGPT can improve the reliability of this transmission. Some companies view the OpenAI Account Dashboard as the secret key to streamlining numerous tasks across the healthcare system.
According to an article from Yahoo! News, “For example, Doximity, an online network for healthcare professionals, reportedly released a beta version of a ChatGPT tool that helps doctors with administrative tasks like, for instance, drafting appeal letters to insurers. Meanwhile, ChatGPT could also occupy other roles historically filled by doctors, like diagnosing garden variety illnesses such as the cold or flu. (An aside: the chatbot recently passed the U.S. Medical Licensing Examination.)”
Given some of the limitations that ChatGPT has shown in the area of accuracy, we aren’t quite ready to transfer diagnostic duties from humans to robots. However, it’s easy to see how healthcare providers may be able to use the OpenAI API to create more streamlined patient experiences at the administrative task level. The example of drafting appeal letters to insurers is a particularly strong demonstration of how to use OpenAI in order to relieve humans of tasks that, simply stated, no human really wants to do.
Moreover, this is an area in which a chatbot with the proper inputs could significantly improve upon what humans are likely to produce on their own. By plugging in the type of information and rhetoric that typically result in successful appeals to insurers, it may be possible to significantly improve the consistency with which appeals are actually granted.
6. Creating Real Estate Listings
The housing market has been extremely hot over the last several years, with new homes inciting competitive bidding and offers well over asking price. Real estate agents have been challenged by the speed of the market. Keeping up with the lightning quick pace of listings and sales has inclined many in the field to explore the possibility of taking on a bit of automated help.
For many in the real estate business, the first and most obvious use of ChatGPT has been as a generator of compelling write-ups for new listings. Many in the housing business have raved about the application’s ability to turn a few specs on a home into a remarkably well-written and effective real estate listing. And because few real estate agents would prefer to spend their time writing, as opposed to staging and showing properties, ChatGPT is eliminating a step that most in this line of work tend to find particularly tedious.
Once again, this underscores a key point. ChatGPT is not purely a threat to human labor. The real estate business offers another context in which the application offers ways for working professionals to reduce tedium and improve their own productivity.
But for the real estate business, there are yet other ways in which the language model is proving valuable. According to the article from Yahoo! News, “For State Street Realty VP Broker Frank Trelles, who’s based in Miami, Fla., ChatGPT’s benefits aren’t purely theoretical. He regularly uses the bot for chores like calculating mortgage payments, analyzing investment deals, and drafting agreements. Trelles said the technology can allow real estate brokers and agents to increase their focus on clients – not the rote work of real estate transactions.”
In other words, instead of replacing human labor, ChatGPT is actually giving real estate agents more time to focus on the distinctly human aspects of the profession.
7. Software Coding
This is one area in which users have absolutely raved about ChatGPT’s capabilities. In the simplest terms, it takes the innovative application a matter of seconds to solve problems, identify errors, or produce solutions for coding challenges that historically might have taken even the most eagle-eyed coder hours to figure out.
While ChatGPT can’t yet be trusted to do fact-based research on your behalf, it is exceptional at solving problems when given the relevant information. That’s essentially what it can do for you as a software developer or a computer programmer. ChatGPT is fluent in computer programming languages, which means that it can check your coding for errors, write code for you, or provide code completion when you get stuck.
Thanks to these capabilities, says an article in Tech.co, “ChatGPT is making major waves in the software industry. One data engineer from Detroit told Tech.co that he uses the chatbot when he’s stuck on a project. ‘If you know how to give correct prompts, it can help you with code logic, syntax errors, and even change code from one coding language to another,” he explained.”
Once again, this is a perfect example of the importance of a human end user. It requires the input of a human being to introduce the key concepts driving a given programming task. However, once the proper inputs have been provided, ChatGPT is capable of removing many of the bottlenecks that have historically slowed the pace of code completion and final product development.
8. Personalized Marketing
In the digital marketing age, engagement relies on an increasingly personalized user experience. This presents a challenge to brands as they grapple with an ever growing set of channels. It can be difficult to cut through the noise to reach the right target with a given message.
That’s why major brands are now exploring ways of using ChatGPT, and the OpenAI API more broadly, in order to create increasingly individualized brand experiences. Some brands are finding that the AI-driven technologies enable a combination of personalized messaging and image variations.
Coca Cola offers one major example. According to Forbes, “The soft drinks giant has formed a partnership with consultants Bain & Company, with the aim of using ChatGPT to assist with marketing and creating personalized customer experiences. According to a press release, it plans to use the technology, along with the generative image tool Dall-E, to craft personalized ad copy, images and messaging.”
While it remains to be seen how effective this strategy is for Coca Cola, it does point the way toward a more personally tailored marketing experience. It seems reasonable to anticipate that other major brands will follow suit with their own experiments in marketing that are simultaneously automated and individualized.
9. Presentation Building
Typically, when we build presentations for a team, a client or an audience, the goal is to summarize and explain projects or ideas that already exist. In other words, the inputs are already established. We’ve already done the real work by completing the project in question or conducting the research at the heart of our presentation.
The presentation itself is merely the act of aggregating information about that achievement into something comprehensible to your audience. This is an act that you can easily outsource to ChatGPT with generally excellent results.
This is the type of functionality that can be useful in just about any field. As the user, it is your job to ensure that you are providing comprehensive instructions to ChatGPT on how you wish to lay out your presentation. It is also your job to ensure that you include all the proper inputs. From there, the application will use those instructions and inputs to create a fully formed presentation.
According to the article from Tech.co, “Harman Signh, Director of the cybersecurity services company Cyphere uses the AI app to create reports and presentations. He explains that ChatGPT allows him to create high-quality presentations quickly and easily, which has a substantial impact on the efficiency of his work. Signh also uses the chatbot to produce reports, responses to customer inquiries, and marketing materials like blogs, which is a true testament to the tool’s versatility.”
For just about anybody working in a corporate setting, ChatGPT has the potential to save untold hours on tasks like presentation scripting, meeting preparation, agenda writing, and much more.
10. Homework Assistance
We mention this final function with a very large grain of salt. In the aftermath of ChatGPT’s introduction, concerns arose immediately about the text generation software’s implications to the field of formal education. Indeed, the application offers a freely available essay writing function that is relatively difficult for educators to detect. In other words, ChatGPT clearly presents a cheating threat.
As proof that students have already been using the application to complete their homework, companies whose revenues are driven by homework assistance services saw a major hit to their bottom line in the wake of ChatGPT’s emergence. Chegg is a prominent example
Over the last several years, Chegg has been the target of ire from educators due to a “homework assistance” service that verily blurs the line between helping students complete homework and helping students cheat. But with the emergence of ChatGPT, Chegg’s business has struggled mightily to stay afload. In the face of its rapidly eroding revenue, Chegg immediately began leveraging the new technology.
According to Forbes, “This educational platform used by college students to assist with assignments has announced a new service called Cheggmate, powered by GPT-4. Previously the company’s business model involved students paying to have questions answered by human specialists. After its stock fell in value by 40% when CEO Dan Rosensweig announced the company’s growth was being impacted by the emergence of ChatGPT, it quickly put together plans to build it into its services.”
This pivot carried some pretty clear consequences for those employed by Chegg. Cutting 4% of its staff in June of 2023, Chegg acknowledged that ChatGPT played a major role in its lost revenue. In this case, we can see that neither the company’s adaptation of ChatGPT nor its consequent elimination of some 80 human jobs were the result of its desire to innovate.
Instead, its adaptation of ChatGPT was the result of its imperative for mere survival. No doubt there will be other organizations, across numerous industries, who will likely be pressed to follow suit in the coming years.
***
One of the recurrent themes of this article is the importance of human operators in helping ChatGPT realize its fullest potential. Ideally, the future of labor will not be one in which technology like ChatGPT replaces us, but one in which it makes us more productive.
But it’s up to each of us to seize this opportunity to better understand the technology, how we can leverage it, and how it can better empower us in our own work.
For a look at the kinds of jobs that will likely provide you with an opportunity to do just that, jump to our article on these 10 recession proof jobs.