🗣️ What is ChatGPT
Issue 26 — Technology
DALL-E, ChatGPT, OpenAI- what is all this? You may have heard a lot of headlines lately about OpenAI and Microsoft making a massive 10-billion-dollar investment into the company, but what is all this hype about? In this article, we’ll get to all of this plus more! Make sure to like, share, and subscribe!
What is OpenAI
Elon Musk is known for his impulsiveness to start companies and his crazy new endeavors, and OpenAI is one of them. The company was founded in 2015 by Elon Musk, Sam Altman, Greg Brockman, Ilya Sutskever, and Wojciech Zaremba. However, Elon left the company’s board of directors in 2018. The company was founded based on the idea that Artificial Intelligence should be used in a good, safe way, and it began as a nonprofit organization, but that was changed in 2019 to a profited organization. The company’s mission is to develop and promote friendly AI in a way that benefits all people. They conduct research in machine learning and deep learning and have made significant contributions to the field, including the development of ChatGPT, one of the most advanced language models in existence. They have also made an AI image creator called DALL-E. OpenAI is also developing AI systems that can be used in various apps like natural language processing, computer vision, and robotics.
The company also wants to popularize artificial intelligence by making it more mainstream by letting developers be able to implement OpenAI’s technologies into their applications.
DALL-E
One of the systems that OpenAI has developed is called DALL-E. It’s an artificial intelligence system that can generate diverse and highly imaginative images from any textual description. A textual prompt, such as “a 3d render of a green astronaut walking in the desert,” could be used. The system can generate an associated image showcasing its creativity and imagination.
Image generated by DALL-E
I’ve written another article about Artificial Intelligence on a more fundamental perspective which you can check out here. It may help you better understand this next part! DALL-E is based on state-of-the-art deep learning algorithms, specifically a transformer architecture, which allows it to process and understand the meaning of text and translate that into visual information. The system has been trained on a diverse range of images and textual descriptions, allowing it to generate highly imaginative images that go beyond the limited scope of its training data.
The system has the potential to revolutionize the way images are generated and used in various applications, such as product design, marketing, and entertainment. Its ability to create unique and highly imaginative images from textual descriptions opens up new possibilities for creativity and innovation in the field of AI.
So there are 2 main parts to generating an image like this. In my example DALL-E knows what the concepts of green, astronaut, 3D-render, and walking are (this is the first part). The second part is that it knows how to create an aesthetically pleasing image with all those qualities above. When DALL-E takes its ten seconds to create an image, you can most likely assure that you will have a nice image.
Now Artificial Intelligence isn’t perfect, and there are some limitations. For example, when you try to ask for specific text, it starts to fall apart. Here’s what happens when I ask for “a road sign that says please prepare to stop”:
Image generated by DALL-E
By the way, if you are enjoying this article, please be sure to share and subscribe. I have a promotion going on which allows new and existing subscribers to get a 1-month free trial to Skillshare (the best place to learn any new skill)
ChatGPT
Now we come to the system that’s been talked about all over the place these days: ChatGPT. It was released on November 30, 2022, so pretty recent, and it’s a conversational AI model. The “GPT” in ChatGPT stands for Generative Pre-training Transformer. It is based on the transformer architecture (hence the “T”) and is fine-tuned on a large corpus of text data to generate human-like responses to text-based prompts.
The system has been trained on a diverse range of texts, allowing it to respond to a wide variety of questions and generate text in a natural and coherent manner. It is designed to be used as a conversational AI agent and can be integrated into chatbots, customer service systems, and other applications where text-based interactions are involved. The bot doesn’t really have a personality so if you ask it “what’s your favorite color” it will say something like “I don’t have a preference”.
ChatGPT works as follows: You give it a prompt and the chatbot writes something for you. For example, if I ask it “what is the superbowl?” it says:
The Super Bowl is the championship game of the National Football League (NFL), the highest level of professional American football in the United States. The game is played annually on the first Sunday in February and is considered one of the biggest sporting events in the world. The Super Bowl is known for its high-level of competition, celebrity halftime shows, and the excitement surrounding the event. Millions of people tune in to watch the game on television and it is also a major event for advertisers, with commercials during the broadcast being some of the most expensive and highly anticipated of the year.
Text Generated by ChatGPT
Now just like DALL-E there are some limitations. As I mentioned before it doesn’t have a personality, but that’s okay. A current limit is that ChatGPT can’t tell you very recent events since it can only answer things that happened before 2021, however this can be changed soon.
So overall, ChatGPT is a powerful tool for generating human-like text and can be used for a variety of purposes, such as answering questions, generating creative writing, or even generating code snippets. With its ability to understand and generate text, ChatGPT represents a significant advance in the field of AI and natural language processing.
What’s Microsoft Have to do With This?
You may have been seeing headlines all over the place about Microsoft making a passive investment into OpenAI which is true. Microsoft placed a 10 billion dollar investment into the company with Microsoft receiving 75% of OpenAI’s profits until it recoups its investment, which means that Microsoft would own 49% of the company (according to VOX).
Microsoft of all companies is doing this because they want to bring more competition to the Google search engine by integrating OpenAI’s services into their own products. We’ll have to see where this will take them!
Thanks for reading this week’s article and I hope you learned something new about OpenAI and what it does. Talk to you next week!
-Luke Rapaka
Originally published at https://lukearapaka.substack.com on February 15, 2023.