ChatGPT is a conversational AI interface created by OpenAI, which combines the capabilities of genera-purpose large language models, like GPT-3, and it has been fine-tuned with a model called InstructGPT, which added a supervised learning method, which is human-in-the-loop, to increase truthfulness, and reliability of the conversational interface.
How does ChatGPT work?
To grasp large language models, it’s critical to understand a few key concepts.
Current large language models, like GPT-3, have been trained on a massive amount of data and billions of parameters, which were used to pre-train those models.
Once the model has been pre-trained, it’s pretty good at general tasks. Though it can be misleading and untruthful, thus hallucinating often.
To “put some breaks” and enhance the guardrails of a general-purpose model, like GPT-3, that can be fine-tuned.
In short, it can be trained, with a supervised learning approach, on a much smaller dataset, which is human-labeled, and where humans show the general-purpose engine how to get better at specific tasks and be more truthful.
In the specific case of ChatGPT, it has been used a specific human-in-the-loop model, called InstructGPT, to smooth out some of the negative aspects related to GPT-3 to make ChatGPT viable as a conversational interface.
ChatGPT can often still be misleading, yet it can improve over time as it learns how to deal with more and more edge cases.
How does ChatGPT Make money?
ChatGPT Premium (Plus/Turbo)
ChatGPT is available in a premium version called Plus, available at $20/mo, which comprises:
- General access to ChatGPT, even during peak times
- Faster response times
- Priority access to new features and improvements
ChatGPT is now available also as API. As OpenAI explained, ChatGPT APIs use
gpt-3.5-turbo, which is the same model used by ChatGPT product and it is priced at $0.002 per 1k tokens, which is 10x cheaper than the existing GPT-3.5 models.
- ChatGPT is a conversational AI interface created by OpenAI.
- It combines the capabilities of large language models like GPT-3 and is fine-tuned with InstructGPT, a human-in-the-loop model, to increase truthfulness and reliability.
- Large language models like GPT-3 are pre-trained on massive amounts of data using unsupervised learning to predict the next token in a text sequence.
- After pre-training, fine-tuning is done with a smaller human-labeled dataset to improve the model’s performance on specific tasks.
- ChatGPT is designed as a conversational interface, but it can still be misleading and improve over time as it learns from more edge cases.
- To monetize ChatGPT, OpenAI follows a freemium model, offering a premium version called ChatGPT Plus starting at $20/mo.
- The premium version includes faster response times, priority access to new features, and general access even during peak times.
- API access is another revenue generation option, allowing developers to integrate ChatGPT into their applications at $0.002 per 1k tokens, 10x cheaper than existing GPT-3.5 models.
- ChatGPT’s meteoric rise in popularity has driven significant traffic to OpenAI’s website since its launch.
Connected Business Model Analyses
Stability AI Ecosystem