The Disruptive ChatGPT
The Generative AI chatbot could change how we learn and work
Steve Jobs’ 1985 Playboy Interview
The Turing Test
Steve Jobs’ vision of the intelligent computer is realized last month with the release of ChatGPT, a new artificial intelligence chatbot from OpenAI. Not only it can summarize information such as Aristotle’s worldview and answer any question about him, but it can also create new content, tailored to your request, often with a startling degree of nuance, humor, and creativity. In fact, ChatGPT has worked so well that a teacher couldn’t always tell whether the essay is written by a fourth grader or ChatGPT. ChatGPT has just passed the Turing Test, a demonstration of a machine’s capability to think intelligently.
ChatGPT explains the Turing Test
Finally, a Machine That Can Finish Your Sentence
The underlying idea of GPT-3 is a way of linking an intuitive notion of understanding to something that can be measured and understood mechanistically, and that is the task of predicting the next word in text.
Other forms of artificial intelligence try to hard-code information about the world: the chess strategies of grandmasters, the principles of climatology. But GPT-3’s intelligence, if intelligence is the right word for it, comes from the bottom up: through the elemental act of next-word prediction. To train GPT-3, the model is given a ‘‘prompt’’ — a few sentences or paragraphs of text from a newspaper article, say, or a novel or a scholarly paper — and then asked to suggest a list of potential words that might complete the sequence, ranked by probability. In the early stages of training, the suggested words are nonsense. Prompt the algorithm with a sentence like ‘‘The writer has omitted the very last word of the first . . . ’’ and the guesses will be a kind of stream of nonsense: ‘‘satellite,’’ ‘‘puppy,’’ ‘‘Seattle,’’ ‘‘therefore.’’ But somewhere down the list — perhaps thousands of words down the list — the correct missing word appears: ‘‘paragraph.’’ The software then strengthens whatever random neural connections generated that particular suggestion and weakens all the connections that generated incorrect guesses. And then it moves on to the next prompt. Over time, with enough iterations, the software learns.
Ilya Sutskever, Chief Scientist at OpenAI, explains how GPT works
Powering ChatGPT is the “Generative Pre-Trained Transformers” ( GPT ), an AI language model that is practically allowing the chatbot to guess the next word in a sentence. Trained by a massive amount of data ( 45TB of data, 135 Billion of parameters ) from the internet — digital books, the length and breadth of Wikipedia, and nearly a trillion words posted to blogs, social media, and more —, GPT is able to learn skills such as language translation and question answering. By mastering these tasks, GPT would also learn how a natural language is pieced together. And it is exactly this deep understanding of the language that allows ChatGPT to write jokes (some of which are actually funny), generate working computer code, author college-level essays, guess at medical diagnoses, create text-based Harry Potter games and explain scientific concepts at multiple levels of difficulty.
ChatGPT explains why it is so good at understanding natural language.
Transformers, explained: Understand the model behind GPT, BERT, and T5
The Rise of Generative AI
The disruptive ChatGPT is one of the many apps in Generative Artificial Intelligence, an AI technology that can generate text, images, sounds, and other media in response to short prompts.
Companies have developed Generative AI for years, including tech giants like Google and Meta. But the technology has not created such a buzz until ChatGPT’s creator OpenAI unveiled a system called DALL-E, which let people generate photo-realistic images simply by describing what they wanted to see.
DALL-E generated these images by following a command for “a giant hamster blimp carrying passengers.”
By combining these generative AI tools, a designer can now publish a children’s book co-written and illustrated by AI, in a weekend.
ChatGPT is indeed capturing the public imagination.
And this is only the beginning
ChatGPT can completely disrupt how we look for information online. Instead of asking the user to click through a number of search results from Google, ChatGPT can curate an easily-understood paragraph that includes all the relevant information
Ultimately, ChatGPT could provide a new way of interacting with almost any software, letting people chat with computers and other devices as if they were chatting with another person.