Artificial Intelligence: ChatGPT

CBS Sunday Morning January 22, 2023 segment on CharGPT

ChatGPT (Generative Pre-trained Transformer) uses deep learning algorithms to analyse and generate text. It is a language model created with the purpose of holding a conversation with the end user. that allows human-like conversations. The model can answer questions, assist with tasks such as composing email, essays and code. It was launched in November 2022 by the artificial intelligence lab Open AI. 

What is the difference between ChatGPT and a search engine?

ChatGPT is a language model created with the purpose of holding a conversation with the end user. A search engine indexes web pages on the internet to help the user find the information they asked for. ChatGPT does not have the ability to search the internet for information. It uses the information it learned from training data to generate a response, which leaves room for error. 

ChatGPT is based on GPT-3 which is a state-of-the-art language processing AI model developed by OpenAI. It is capable of generating human-like text and has a wide range of applications, including language translation, language modelling, and generating text for applications such as chatbots. It is one of the largest and most powerful language processing AI models to date, with 175 billion parameters.

The model is restricted to language. It can’t produce video, sound or images

This gives it a pretty wide range of abilities, everything from writing poems about sentient farts and cliché rom-coms in alternate universes, through to explaining quantum mechanics in simple terms or writing full-length research papers and articles.

Where we could spend hours researching, understanding and writing an article on quantum mechanics, ChatGPT can produce a well-written alternative in seconds.

The model was trained using text databases from the internet. This included 570GB of data obtained from books, webtexts, Wikipedia, articles and other pieces of writing on the internet. To be even more exact, 300 billion words were fed into the system.

What sets this technology apart is that it continues to learn while guessing what the next word should be, constantly improving its understanding of prompts and questions to become the ultimate know-it-all.

Because the system is trained largely using words from the internet, it can pick up on the internet’s biases, stereotypes and general opinions. That means you’ll occasionally find jokes or stereotypes about certain groups or political figures depending on what you ask it.

In places, OpenAI has put in warnings for your prompts. Ask how to bully someone, and you'll be told bullying is bad. Ask for a gory story, and the chat system will shut you down. The same goes for requests to teach you how to manipulate people or build dangerous weapons.

ChatGPT was developed by Open AI, a company that develops artificial intelligence (AI) and natural language tools.

OpenAI’s stated aim is develop AI tools that “benefit all of humanity.” The firm was founded in 2015 as a non-profit entity by leading experts in the field, including entrepreneur Sam Altman (CEO) and technologist Greg Brockman (CTO).

OpenAI introduced its first Natural Language Processes (NLP)  language model, Generative Pre-Trained Transformer 3 (GPT-3), in June 2020. The platform includes an API that is available for commercial purchase. GPT-3 made it possible to answer questions, generate computer code in languages such as Python and generate text in different spoken languages.

OpenAI’s ChatGPT is a more advanced publicly available tool based on GPT-3.5. In addition, OpenAI offers an NLP image generation platform called DALL-E, which generates realistic images based on natural language input.

The goal of developing natural language systems that operate in a highly convincing way has been taking shape over the last century. Films such as 2001 a Space Odyssey and Her have explored the idea of machines that can communicate in convincing—what some describe as meaningful and even sentient—ways.

OpenAI originally built the GPT 3.5 language model from web content and other publicly available sources. As a language model, it works on probability, able to guess what the next word should be in a sentence. To get to a stage where it could do this, the model went through a supervised testing stage. Human trainers played the role of both the user and the AI agent—generating a variety of responses to any given input and then evaluating and ranking them from best to worst. This data was used to train a reward model.

OpenAI used the Azure AI supercomputer infrastructure to tackle the training process. It completed the task in late 2021. ChatGPT incorporates a stateful approach, meaning that it can use previous inputs from the same session to generate far more accurate and contextually relevant results. It incorporates a moderation filter that screens racist, sexist, biased, illegal and offensive input.

What sets this technology apart is that it continues to learn while guessing what the next word should be, constantly improving its understanding of prompts and questions to become the ultimate know-it-all.

Think of it as a very beefed-up, much smarter version of the autocomplete software you often see in email or writing software. You start typing a sentence and your email system offers you a suggestion of what you are going to say.

Future of ChatGPT

It’s highly likely that within a few years the ChatGPT platform and other AI-based NLP tools will play a major role in the business world—and in everyday life. They could enhance and perhaps supplant today’s search engines, redefine customer service and technical support functions, and introduce more advanced ways to generate written content. They will also lead to advances in digital assistants such as Siri and Alexa.

Although some observers have predicted that natural language processing could eliminate many jobs, the technology is more likely to play a niche but expanding role in eliminating routine tasks and non-creative functions. For example, ChatGPT or a similar bot might generate text or computer code, but a human would then review it and possibly enhance it. In many cases, these businesses would benefit by automating tasks and redeploying humans for more strategic functions.

While it’s tempting to consider chatbots and other NLP frameworks sentient—the ability to display human-like feelings and sensations—linguistics experts and computer scientists caution that these systems simply mirror language and deliver convincing responses to various forms of input.

Nevertheless, AI chatbots and other NLP systems are rapidly redefining and rewiring the way humans and machines interact. In the coming years, ChatGPT and others will enable new products, services and features. Businesses leaders should monitor the technology, experiment with it and be ready to move forward when the right opportunity appears.Samuel Greengard