You’ve undoubtedly heard about OpenAI’s ChatGPT, a powerful chatbot built on natural language processing and artificial intelligence technologies. In its first two months of release, the free tool attracted more than 100 million monthly active users, making it the fastest-growing consumer application of all time, according to UBS.
By comparison, it took TikTok nine months and Instagram 2-1/2 years to reach that milestone.
According to ChatGPT, “ChatGPT is an AI language model developed by OpenAI, which is capable of generating human-like text based on the input it is given. The model is trained on a large corpus of text data and can generate responses to questions, summarize long texts, write stories and much more. It is often used in conversational AI applications to simulate a human-like conversation with users.”
Breaking that down, ChatGPT is a natural-language AI processing tool that allows for engagement with a chatbot, much like you’d converse with a human. ChatGPT can answer pretty much any question you have, as well as write code, content, compose emails, research medical questions, write essays and papers, and much more.
Why’s it so hot? Because possibly for the first time, the average individual has access to a generative AI program that can generate seemingly new content like text, images, and audio due to the fact its algorithms have been trained on huge amounts of data.
What’s more, it can quickly write articles, essays, and software code that come frightfully close to what many human counterparts might produce.
So it’s no surprise that companies are exploring ways of using ChatGPT across all parts of the business, from finance and human resources to marketing and public relations.
But experts worry some companies might be forging ahead without fully understanding ChatGPT’s capabilities. This is early-stage stuff, and even OpenAI CEO Sam Altman admits its progress will be more gradual than people think.
What do businesses need to know about this wildly popular tool? Here, we separate fact from faction.
For better or worse, chatbot technology is one of the most important tools in the belts of modern marketers. And while it's advanced quite a bit lately, it’s likely to undergo even more dramatic changes in the next few years.
Its seemingly sudden popularity could make ChatGPT seem like brand-new technology. But while the tool is new, the underlying technology isn’t.
OpenAI has been around since 2015. It released GPT-1 in 2018. GPT-2 came along in February 2019. GPT-3 made its debut in June 2020 and was quickly licensed to Microsoft in September 2020.
The ChatGPT tool itself was launched in November along with ChatGPT 3.5. And GPT-4 launched in March of this year.
In a customer support model, how does artificial intelligence stack up to human intelligence? Here, we explore how AI can benefit customer service, but also where it falls short.
ChatGPT may seem extraordinary, but its uniqueness is relative.
Some say what sets ChatGPT apart from other emerging AI tools its use of Reinforcement Learning from Human Feedback (RLHF), which allows the tool to interact with users in a more conversational way.
In other words, it can answer follow up questions, admit and learn from mistakes, challenge incorrect assumptions, reject inappropriate requests, and even offer helpful, more appropriate request options.
On the other hand, ChatGPT has plenty of competition. For instance, Anthropic, a San Francisco startup, raised hundreds of millions of dollars from Google and recently attracted an additional $300 million of investment for its AI chatbot, Claude. Another conversational AI startup, Amelia, garnered $175 million.
Meta and Amazon, among others, are also said to have competitive generative AI offerings cooking.
The machines are here. AI is boosting e-commerce in countless ways – discover some of the best examples.
While some marketing organizations have started using ChatGPT to automatically write generic emails, brochures, articles, and press releases, it’s not happening at scale.
ChatGPT still lacks the human touch. It relies on cobbling information together from multiple sources. While it does that fairly well, the content it generates tends to be vanilla and probably won’t be as persuasive as something produced by real people. It’s also only as good as the information it relies upon. So, if that data is riddled with grammatical and factual errors, those could carry over to the final copy.
What’s more, ChatGPT rides a hazy thin line between repurposing publicly available content and plagiarism.
Indeed, brands could face legal issues if it turns out they trained their AI algorithms using someone else’s content or disseminated communications based on it without an author’s permission.This isn’t likely to deter companies from using ChatGPT, however. Its time-saving research and writing capabilities are being integrated into many productivity tools. Microsoft, for example, is reportedly integrating ChatGPT into Outlook and PowerPoint. And it’s already added it to Microsoft’s Bing search engine and Edge browser.
In a digital world, content marketing ROI matters more than ever. Without good content, a business can't drive sales. Here are some tips for creating content that sells.
Does ChatGPT pose a major security risk for businesses? It’s kind of a mixed bag.
Most of us are pretty good at identifying phishing attempts where a hacker sends an email aimed at tricking us into clicking a link that will let them seize control of our PCs or networks. We know to avoid emails containing hazy graphics, misspellings, and nonsensical grammar.
But hackers in those far-off lands could start using ChatGPT to clean up their communications and make phishing attacks more difficult to detect. Similarly, it could help hackers send larger quantities of phishing emails. And in some cases, it could speed creation of malware by hackers who prefer to develop their own programs as opposed to purchasing them on the Dark Web.
Still, security experts aren’t freaking out about ChatGPT. Krystal Jackson, a junior AI fellow in Georgetown University’s Center for Security and Emerging Technology, says it’s just another tool in the belts of hackers.
“Ultimately, it’s not presenting a new challenge but just reemphasizing the need for organizations to have really robust security practices, to prioritize filtering messages based on content, and to educate people on the types of phishing messages they’re most likely to receive at any given time,” she says.
Jackson and other experts also note that cybersecurity professionals could also use ChatGPT to accelerate development of anti-hacking tools. So, its impact on the threat landscape could be a wash.
Phishing attacks are stealthier than ever, putting brands and their reputations at risk. Find out four ways to fend off fraudsters.
Look, there are ways to use it to save time right now. If you don’t particularly care about communications looking like they’re written by a chatbot, have at it. It could save some time and, if you’re using the basic service, won’t cost you anything. Plus, it’s fun to toy with.
But ChatGPT is likely years away from achieving its promise. Case in point: the highly anticipated release of GPT-4 didn’t catapult the AI chatbot to new heights.
In fact, as Andrea Guzman points out in Forbes, “GPT-4 continues to suffer from some of the same key limitations that have raised concerns about AI, particularly its tendency to ‘hallucinate,’ or invent facts and present them as truth.”
OpenAI even says that people should take “great care” when using GPT 4, saying its limitations create significant safety challenges.
The vision for AI-based chatbots is their communications will someday be indistinguishable from those of human beings. But we’re not there yet, and won’t be for a while.
So, for now, the best thing businesses can do with ChatGPT is to monitor and possibly experiment with it. But don’t set expectations too high right now because it will take many years to reach its full potential.
