The ChatGPT sign displayed on the OpenAI website on a laptop screen and the OpenAI logo displayed on a phone screen are seen in this illustration taken in Krakow, Poland, on February 2, 2023.
Yakub Pazhytsky | Nurphoto | Getty Images
What is ChatGPT? I asked the AI chatbot that has ignited conversations in schools, corporate boardrooms and social media to explain itself.
In its own description, ChatGPT is “an AI-powered chatbot developed by OpenAI and based on the Generative Pretrained Transformer (GPT) language model. It uses deep learning techniques to generate human-like responses to text input in a conversational manner.”
The tool is the talk of the business world. It was mentioned in earnings calls from the management of a number of companies, including oil giants, banks – and even industrial behemoth Caterpillar.
It also raised concerns about possible abuse. In classrooms, students used ChatGPT to create entire essays, and hackers began testing it to write malicious code.
So what is ChatGPT? Here’s a simple guide to everything you need to know about the popular AI chatbot.
What is ChatGPT?
ChatGPT is an AI chatbot developed by San Francisco startup OpenAI. OpenAI was founded in 2015 by Elon Musk and Sam Altman and is backed by notable investors – primarily Microsoft.
This is one of several examples of generative AI. These are tools that allow users to enter written prompts and receive new human text or AI-generated images and videos.
ChatGPT provides an AI-generated response to the “Tell me about ChatGPT” query.
Leon Neal | Getty Images
Previous examples include Dall-E, a text-to-image conversion program from OpenAI that has caught the attention of people fascinated by its ability to create realistic, often absurd images that match people’s textual descriptions.
Lensa, based on the open source AI project Stable Diffusion, has been used to turn selfies into beautiful self-portraits inspired by everything from sci-fi to anime.
In the case of ChatGPT, the service is a text-based tool that can generate human responses to users’ queries, from William Shakespeare-style poems to advice on what to do at a child’s birthday party.
What is special here?
ChatGPT is based on a large language model, or LLM, which means it is programmed to understand human language and generate responses based on large corpora of data.
ChatGPT’s LLM is called GPT-3.5. This is an update to the OpenAI GPT-3 language model.
With a whopping 175 billion parameters, GPT-3 is one of the largest and most powerful AI language processing models to date.
What makes ChatGPT so impressive is its ability to provide human-like responses, thanks in no small part to the vast amounts of data it’s trained on.
“What’s exciting is that the responses are becoming more and more human-like, so you’re seeing things that we didn’t think computers could do before,” Jeffrey Wong, global chief innovation officer at the professional services firm, told CNBC. EY.
Another thing that sets ChatGPT apart is its ability to log context from earlier messages from users in the stream and use it to shape responses later in the conversation.
Why is it so popular?
No generative AI application has managed to achieve the impact and virality that ChatGPT has.
It has been the subject of countless memes and talk about the business community at the World Economic Forum in Davos, Switzerland, last month. Chinese tech giant Baidu has made its own version called Ernie Bot.
The chatbot registered 1 million people five days after its release, according to a Tweet on December 5 from Altman. By January, ChatGPT had amassed 100 million monthly active users just two months after its launch, making it the fastest-growing consumer app in history, according to a UBS note published last week.
TikTok took nine months to reach 100 million users, while Instagram took two and a half years.
January 31 was ChatGPT’s biggest day yet, with its website hitting a record 28 million daily hits, according to Similarweb. This was 165% more than a month ago.
One of the reasons for ChatGPT’s popularity is its accessibility. The service is open to anyone via the OpenAI website, and its potential applications range from school homework to legal reports.
According to Wong, timing also played a role.
“When we come out of a pandemic, you usually see this burst of creativity,” he said. “The biggest example of that after the Black Death was this Renaissance, this surge of creativity in all directions.”
Why the tech giants are fighting
Microsoft is betting billions on ChatGPT’s owner, OpenAI. In late January, the tech titan announced a multi-year, multi-billion dollar investment deal with OpenAI.
Microsoft declined to disclose a specific dollar amount. A report from Semafor says the Redmond, Washington-based tech giant is in talks to invest up to $10 billion in the company. Microsoft previously invested $1 billion in OpenAI.
Microsoft held a press event on Tuesday to announce new AI-powered updates to its Bing search engine and Edge browser. Altman confirmed that Microsoft has incorporated some OpenAI GPT-3.5 language technologies into Bing.
It was a day later Google introduced its own answer to ChatGPT called Bard AI. The company plans to begin rolling out Bard to Google Search in the coming weeks.
ChatGPT is considered a threat to Google. Instead of turning to the pioneer of web search to ask their most pressing questions, people can rely on ChatGPT.
Google was actually early in the advanced conversational AI game, launching its own large language model called Lamda in 2021. It missed the mark with the launch of its own Lamda-based consumer product — and hopes to change that with Bard, which supports the language model.
How good is ChatGPT?
ChatGPT has its limitations. Chatbot responses may contain factual inaccuracies. For example, he may come up with fictitious historical names and books that do not exist, or fail to solve some mathematical problems.
“The open-ended aspects of these models are a double-edged sword,” Will Williams, vice president of machine learning at UK AI startup Speechmatics, told CNBC.
“On the one hand, there is a high level of flexibility and fluidity in the interaction, and an interesting conversation is possible on almost any topic. On the other hand, you never know when the model comes into contact with reality, and not an overly confident hallucination.”
ChatGPT’s knowledge is still limited to 2021 data, but may improve over time. Looking ahead, ChatGPT is expected to be a precursor to much more advanced AI systems.
At the moment, experts say that generative artificial intelligence is not yet capable of reaching human, “general” intelligence.
Artificial General Intelligence, or AGI, is often considered the holy grail of the AI community. It most commonly refers to the ability of an intelligent agent to understand or learn any intellectual task that a human can do.
A host of companies are hoping to achieve this, from OpenAI to Google’s DeepMind.
The capabilities of GPT-3 have already generated excitement for OpenAI’s next generation LLM model GPT-4.
Tempering expectations, OpenAI’s Altman dismissed the hype surrounding GPT-4, saying in a recent interview with StrictlyVC that people were “asking to be disappointed”.
WATCH: What is Web3? We ask the person who invented this word