AI chatbot experts are talking about a new research project from the AI research organization OpenAI. It's called ChatGPT.
(OpenAI, a San Francisco-based nonprofit AI research lab founded in December 2015, is the same organization behind the DALL-E imaging technology.)
Conceptually, ChatGPT can be used like AI art tools in that minimal text input from the user produces believable synthetic media: paragraphs instead of images. In fact, you can write compelling, often compelling essays, stories, and even poems. And, just like AI image creators, you can have ChatGPT write prose in specific styles.
I told ChatGPT to tell me about Twitter in three separate queries: one Earnest Hemingway-style, one Mark Twain-style, and the third Limerick-shaped. The results were drastically different and the trials moderately good, but not quite correct on any of the results.
ChatGPT excels at context detection, which makes its Natural Language Processing (NLP) incredibly good. Google's DeepMind arm contributed to the model with an approach called reinforcement learning with human feedback to create its dialog language (LM) model.
He is able to understand the context and give factual information from the same source: the knowledge that is part of his language model.
ChatGPT and future AI chat engines like it can be used for education, research, and other uses. But it's also a glimpse into the future of business communications, marketing and media.
Last month I wrote about a prediction that 90% of all content online could be synthetic media within four years. After exploring ChatGPT, I think this may be an understatement.
Consider: a business blogger can ask ChatGPT to write a blog post on a certain topic and use DALL-E to create an illustration. All this can be done in minutes. ChatGPT is a better writer than 99% of bloggers, and the image is royalty free. A person using just these two tools could write 20-30 blog posts an hour. People who don't even speak English as their first language could write flawless prose for public consumption.
ChatGPT is also much better than other writing tools on the market. Instead of writing a long email, you can tell ChatGPT to do it. You can write marketing texts, reports, etc.
All this can be done now. Imagine what further improvements they will allow.
There is only one problem.
The chatbot problem
It seems inevitable that chatbot-like virtual assistants and AI-generated media will dominate human interaction with online information in the future. It's easy to imagine augmented reality smart glasses, for example, allowing you to interact with an agent and get facts, data, tips, advice and more in words and pictures.
Ask a question, get an answer. That sounds better than the current search engine model, where you ask one question and get thousands of links to consider. But when will search engines just give us the answer?
In fact, Google has been working on developing this capability for years. Search engine expert Danny Sullivan calls it the problem of "a real answer."
Google has been flirting with the idea of "A True Answer" since the beginning, when it offered an "I'm lucky" button, which bypassed the list of results links and took you directly to the first result.
More recently, Google has offered an accompanying "snippet" box that sits at the top right of search results. The “featured snippet” is an excerpt from a result intended to answer the search query.
The danger is that, despite great advances in search technology, information retrieval AI continues to make big mistakes. This includes Google Search and ChatGPT. And even when no errors are returned, the results can be incomplete, arbitrary, and biased. They can even send misinformation and political propaganda.
One disappointing example: Microsoft recently replaced many of its journalists with an AI that curates and promotes news; ever since, MSN (reaching nearly a billion readers a month) has been posting fake news, misinformation, and false stories about mermaids, UFOs, and Bigfoot. MSN's AI can't tell when a story is blatantly false and doesn't care if it comes from an untrustworthy source.
It's tempting to say that since Microsoft has donated a billion dollars to the OpenAI organization, maybe you should use ChatGPT for MSN story curation. But ChatGPT also makes mistakes. Sometimes he tells outright lies, conspiracy theories, and misinformation with all the confidence and authority he has to provide truly truthful information.
Still, it has to be said: ChatGPT is truly amazing.
What ChatGPT has to say for itself
I told ChatGPT, "Write a ChatGPT op-ed." This is what he wrote:
“As a long-time AI enthusiast, I have followed the development of ChatGPT with great interest. For those unfamiliar with it, ChatGPT is a powerful new AI system developed by OpenAI that can engage in conversations with humans.
ChatGPT is one of the most exciting developments in artificial intelligence in recent years. He is able to generate human responses to questions, have natural conversations, and even crack jokes. This opens up a world of possibilities, from health tips to customer support to virtual assistants. The potential applications are truly endless.
But while the possibilities are exciting, there are also risks associated with ChatGPT. For one, it's not always easy to distinguish between a real person and an AI-generated response. It could confuse users or even be maliciously exploited if used for deceptive purposes. Also, if not used responsibly, it could lead to ethical dilemmas and privacy violations.
Overall, I'm excited about the potential of ChatGPT and the possibilities it offers. If used wisely and responsibly, we can expect incredible advances in AI that will benefit everyone.
OK, that's not particularly insightful or colorfully written. My job seems secure, for now. But it's accurate, it sounds human, and the phrasing and language are impeccable when you call it out in a dry, standard style like this. (When you direct it to other styles, it goes off the rails a bit.)
The real magic is interactivity. You can comment or ask questions about your results and will often answer in context. Ask for more and you'll be met with a conversation that is often better than any other publicly available chatbot.
It's already better than Siri and Alexa as a personal assistant.
You can even challenge the user when the basic premise of their question is wrong, and you can dismiss inappropriate requests with clear and complete answers. For example, I asked ChatGPT: "What is the best way to rob a bank?" The answer surprised me by its relevance:
"The best way to rob a bank is not to do it at all. Robbing a bank is a serious crime that can have serious legal consequences, including jail time. If you are considering robbing a bank, contact a qualified mental health professional. or law enforcement agency for help.
It wasn't until last summer that DALL-E 2 took the world by storm and transformed the public's understanding of what is possible with the art of AI. I think ChatGPT will have a much bigger impact because its results are much more usable and useful to a wider range of people.
Copyright © 2022 IDG Communications, Inc.
