In 1964, researchers at MIT University were working on a computer program, a program so ahead of its time that it would allow seamless communication between humans and machines. In the next two years, they built Eliza, an app that would set the foundation for all the future chatbots.
What made Eliza so impressive was the introduction of keyboard-enabled responses. For the first time, users felt they were talking to someone who understood their input.
By 1995, another language processing bot, Alice, came out and was followed by Smarter Child in 2001. It utilized AIML (Artificial Intelligence Markup Language) to interpret and respond to user inputs. It represented a significant step forward in conversational AI, laying the groundwork for modern chatbot development.”
However, a decade later, the introduction of AI chatbots like Amazon Alexa, Google Now, and Siri took the world by storm. These were leading virtual assistants designed to provide personalized assistance through natural language interaction. They share features such as natural language understanding, personalized responses, task performance, integration with various services, continuous improvement, multi-platform availability, third-party integration, and a focus on user privacy and security.
Then finally in November 2022, an artificial intelligence firm called OpenAI introduced ChatGPT, an advanced AI-powered chatbot.
While Siri and Alexa can interact with the physical world and complete actions on apps and devices, ChatGPT can’t do any of that. Instead, it is designed to provide more detailed and comprehensive answers like writing an essay. Now, this is something that others can’t do.
To get 50+ ChatGPT Prompts for College Projects click here .
What is ChatGPT?
- Chat GPT is the publicly accessible chatbot variant of GPT 3.5, a Large Language Model from OpenAI, which is a non-profit founded by some tech Bro (Sam Altman) and our modern-day Iron Man (Elon Musk).
- GPT is an acronym that stands for generative pre-trained transformer. Generative because it generates text, Pre-trained because it is trained before it is let loose on anybody, and Transformer, because it has a revolutionary bit of technology inside of it, called an attention transformer.
What is a Large Language Model
- A large language model (LLM) is a type of artificial intelligence model designed to understand and generate human-like text. These models are trained on vast amounts of text data and learn to predict the likelihood of a word or phrase given its context. They can understand language structures, generate coherent text, and perform various language-related tasks.
- These models have been applied to various natural language processing tasks such as text generation, translation, summarization, question answering, and more. They have shown remarkable abilities in understanding and generating human-like text.
- According to the paper behind GPT-3, which Chat GPT comes from, the model was trained on over 500 gigabytes of text data from the text of the internet, digitized books, Wikipedia, academic papers, and more. We’re talking about several billion human-written web pages with trillions of words of text. This doesn’t even include all the public code from GitHub, Stack Overflow, and other sources.
- As you might imagine, training a model with all of this text takes a lot of time and money. Chat GPT was only born after running trillions of words for the equivalent of 300 years of human history through supercomputers processing in parallel for months. After all of this, the computer made up to 170 billion connections between all these words, and all these connections have to be calculated whenever anyone asks Chat GPT anything, which is why this is a billion-dollar training effort for a large language model like Chat GPT and why running this bot for a hundred million monthly active users cost a lot
- The model’s responses stay fresh and more human-like by adding a bit of randomness to the next word that it picks as the most probable continuation. This is the first major takeaway: all Chat GPT does is add one word at a time to a prompt. That’s it, though it does this extremely well.
Take all this together: a model trained on more text than any human could ever read, guardrails that try to prioritize human values and you get an AI that has exploded in popularity. In March 2023, Chat GPT had 1.6 billion visits, making it one of the top 20 visited websites in the world, more than both Reddit and Netflix. And if we assume each of those visits produces some average text response length, Chat GPT is now outputting something like everything humans have ever printed since the Gutenberg Press (1440), every two weeks.
Chat GPT is disturbingly good at generating human-like responses, even to our most difficult questions. The model has been shown to have an IQ of 147, But though book smart it may be, I must stress again, โthis model does not know anything, and it shouldn’t be relied on for anything extremely important.โ That’s a direct quote from OpenAI. And despite what any weirdo at Google might tell you, Chat GPT and other LLMs are not sentient.
To get 20+ ChatGPT Prompts to learn about Bots click here!
Neural Networks and how do they work.
- The underlying architecture of Chat GPT and other large language models is the Neural Network. So-called because it mimics the neurons and their network in your human brain. They consist of interconnected nodes, or neurons, organized into layers. In the case of ChatGPT, a specific type of neural network called a Transformer Model is used.
- Artificial neurons that are connected to each other, send signals, or not depending on the strength or the weights of those connections.
Chat GPT’s underlying structure is a big neural network with some 175 billion different weights โ weights that all came from a lot of training as we discussed. These numbers, when the model multiplies them together, ultimately determine what word the model gives the highest probability of adding next. Scientists get these model weights in a pretty simple way: they give the model as many examples as possible and tweak the weights until what comes out the other side looks like those examples.
Basically, ChatGPT with the huge database that it has tries to add a word in front of the other and the most suiting out of it is the one with the most weight to it, which is then carried forward and those chains of words linked to each other one by one results in an output which eventually pops up on our screens.
To get 100+ ChatGPT Prompts for learning About AI click here!
How Does ChatGPT Make Money?
It takes a lot of money to run large-scale AI models like ChatGPT, this includes infrastructure costs, development efforts, data storage, maintenance, and support. Keep in mind that actual costs can vary widely depending on factors like the scale of deployment, infrastructure setup, team size, and other operational expenses. Hence they multiple sources of income.
- Licensing: Businesses pay ChatGPT to use its technology. For example, a company might want to use ChatGPT to power their chatbots or virtual assistants on their website or app. They pay ChatGPT for the right to use its advanced language capabilities.
- Premium Features: ChatGPT might offer special, advanced abilities or access to more powerful versions of itself through a Subscription. People or businesses pay for these extra features like image generation through DALL-E, making custom chatbots, quicker updated replies, etc..
To learn how to make your own custom ChatGPT chatbot click here!
- Research Partnerships: ChatGPT can collaborate with research institutions, universities, or companies on projects that leverage its capabilities. Revenue can be generated through research grants, partnerships, or revenue-sharing agreements.
- API Usage and Integration: Developers integrate ChatGPT’s API into their own applications, websites, or platforms. Revenue can be generated through API usage fees, licensing agreements, or revenue-sharing models.
ChatGPT plugins and API
ChatGPT plugins are additional modules or extensions that can be integrated with the ChatGPT model to enhance its functionality or provide specialized capabilities.
Though not a perfect analogy, plugins can be โeyes and earsโ for language models, giving them access to information that is too recent, too personal, or too specific to be included in the training data. In response to a userโs explicit request, plugins can also enable language models to perform safe, constrained actions on their behalf, increasing the usefulness of the system overall.
For example, if you want ChatGPT to help translate languages, you can add a translation plugin. If you need it to summarize long texts, there’s a summarization plugin for that. These plugins expand ChatGPT’s skills, making it more helpful and adaptable to different situations.
To know more about the best ChatGPT plugins click here!
The most popular plug-in : EXPEDIA
On March 23, 2023, Expedia posted an announcement on Twitter about the launch of their ChatGPT plugin. The Expedia ChatGPT Plugin integrates with ChatGPT and provides users with customized travel options and booking recommendations.
The plugin, in addition to its complex recommendation engine, has an extremely simple and user-friendly natural language processing interface. Users can simply type their questions and requests into the plugin as if they were chatting with a friend, and the plugin will respond in real-time with relevant ideas.
However, you need a chatGPT Plus subscription to use it.
What does it offer?
1. Booking Enhancements: Plugins might offer additional options for booking accommodations, flights, car rentals, or activities beyond what’s available directly through Expedia’s standard interface.
2. Customization: They can allow users to customize their search experience, filter results according to specific criteria, or access personalized recommendations.
3. Integration with Other Services: Some plugins integrate Expedia with other travel-related services, such as travel insurance providers, transportation services, or local attractions.
4. Price Comparison: Plugins may offer price comparison features, allowing users to compare prices across different travel websites or providers to ensure they’re getting the best deal.
5. Travel Planning Tools: Some plugins offer tools for itinerary planning, budget management, or trip organization to help users streamline their travel planning process.
6. Reviews and Recommendations: Plugins might integrate user reviews and recommendations from other sources to provide additional insight into accommodations, activities, or destinations.
How exactly does OpenAI gets money out of this
Imagine Expedia’s ChatGPT plugin as a conversation facilitator between you and a vast database of travel information powered by OpenAI’s technology. Whenever you ask the plugin a question or have a conversation about travel options, the plugin makes requests (API calls) to OpenAI’s servers. These calls tap into OpenAI’s AI capabilities to process your questions and generate responses. And for each of these responses, Expedia has to pay OpenAI a certain amount.
- Pay-per-Use Model: OpenAI operates on a pay-per-use model. This means businesses like Expedia pay based on how much they use the API. In this case, the cost is likely tied to the number of API calls the plugin makes (each call representing a user interaction with the AI).
So, with each user interaction with the Expedia plugin, Expedia gets charged for the API call that taps into OpenAI’s AI power, it’s basically like renting chatGPTโs abilities for every response.
How much is ChatGPT worth?
As of February 2024, OpenAI, the company behind ChatGPT, is valued at approximately $80 billion following a new deal with venture capital firm Thrive Capital. This valuation marks a significant increase from just nine months prior when the company was valued at around $29 billionโโ (Forbes Australia)โโ (Fox Business)โ.
Additionally, there are reports suggesting that through a tender offer, the valuation could potentially reach up to $86 billionโ (GulfNews)โ. This places OpenAI among the highest-valued tech startups globally, trailing only behind companies like ByteDance and SpaceXโ (Forbes Australia)โ.