OpenAI’s Awesome GPT-3 Chatbot Wrote a Working Program For Me
In today’s world of cutting-edge advances in artificial intelligence and machine learning, OpenAI has made a name for itself as one of the foremost organizations leading the way. Their GPT-3 chatbot has created an enormous amount of speculation in its first week after release.
Founded back in 2015, this non-profit AI research company seeks to prioritize safety in its efforts toward the popularization of general artificial intelligence with an emphasis on promoting beneficial outcomes rather than conquering its competition. The company has had its computers sift through millions of conversations and websites (and followed it up with reinforcement learning from human Engineers) in order to “learn” how speech works.
Learning from human feedback and the reinforcement learning accompanying it has been important since OpenAI wants to break through the “robotic” responses that are so common in most chatbots today.
In recent years, OpenAI has become a major player in the software engineering field. It uses language models based on a massive amount of computer code to learn how to create programs on its own. In other words, it seeks to create machines that can learn from their environment and make decisions based on what they’ve learned. I forget, what are the three laws of robotics again?
OpenAI recently released chatGPT, a chatbot that gives incredibly realistic replies to your commands and is powered off the OpenAI brain. It’s based on GPT 3.5 and has been an incredible success if its first-week usage is any indication.
But what exactly does this mean for software engineers? I tested it out and the results are incredible, to say the least. Maybe even more incredible than riding a suitcase through the airport (haha).
OpenAI’s GPT-3 Chatbot (chatGPT) Stats
- Released on November 30, 2022
- Free for now, but will most likely not be free in the future
- Built using OpenAI’s GPT 3.5 technology
- Reached its one-millionth user less than a week after launch
- Knowledge is based on a massive amount of data ingested prior to 2021
OpenAI charges for using their service within an application that you write. However, using chatGPT (which is their own web-based application) is free for now.
Language Models
JavaScript and Python are common languages that chatGPT can write in. However, it can respond back in many languages since it has ingested so much code while it’s learning. We tried C#, JavaScript, Adobe ColdFusion, and more. It seems to have knowledge of many languages.
Traditional Software Engineering
Software engineers are the backbone of the technology industry, responsible for designing, building, and maintaining the complex systems and applications that power our modern world. As the demand for new and innovative software continues to grow, the role of the software engineer has become increasingly important and in demand.
One of the biggest challenges that software engineers face is the need to constantly learn and adapt to new technologies and methodologies. The fast-paced nature of the technology industry means that new tools, frameworks, and languages are constantly being introduced, and software engineers need to be able to quickly master these new technologies in order to stay relevant and competitive in their field.
Here Comes chatGPT
This is where chatGPT comes in. chatGPT is a new AI-powered tool that is designed to help software engineers learn and adapt to new technologies faster and more efficiently. chatGPT uses natural language processing and machine learning algorithms to understand and respond to questions from software engineers, providing them with the information and guidance they need to quickly master new technologies and stay on top of their game.
GPT-3 chatbot Examples
I began testing this remarkable application and here are some of my results.
chatGPT Gave Me Christmas Present Recommendations
One of the key features of chatGPT is its ability to provide personalized, context-aware recommendations and guidance. chatGPT is able to analyze a software engineer’s past experience and skills, as well as the specific task or technology they are working on, in order to provide tailored recommendations and advice. This means that software engineers can get the information and guidance they need, when they need it, without having to wade through irrelevant or outdated resources.
chatGPT Is Hitting on My Girlfriend
I might have asked it to write a poem for my girlfriend, Cindy.
Online Support Chats Are Almost Too Easy For OpenAI
Another important aspect of OpenAI’s GPT-3 chatGPT is its ability to provide real-time feedback and support. Unlike traditional learning resources, which are often static and one-sided, chatGPT allows software engineers to ask questions and engage in a back-and-forth conversation with the AI. This means that they can get answers to their specific questions, and receive immediate feedback on their work, which can help them to learn faster and avoid making mistakes.

chatGPT Wrote This Program
Additionally, chatGPT is able to provide software engineers with access to a vast knowledge base of information and expertise. chatGPT is constantly learning and updating its knowledge based on the latest research and best practices in the field, which means that software engineers can access the most up-to-date information and advice. This can be especially helpful for software engineers who are working on cutting-edge technologies or tackling complex challenges, as they can rely on chatGPT to provide them with the latest information and insights.
I did not think this was possible, but I told chatGPT to write a quick chrome extension file that will highlight the word “and” on every page. Within a second, it started writing out the javascript file to do the job. Amazing, and a bit scary. This is how the machines take over, right? Writing their own code?
Then I asked it to write out JavaScript code that will submit a form via the AJAX method and email me the contents. It pretty much got it correct on the first try.
chatGPT Can Fix My Code Errors
I wanted to test chatGPT and see if it could tell me what was wrong with a line of JavaScript. It gave me three possible reasons.
This could be used as a mentor reviewing your code and helping you debug. That can be extremely useful for any programmer.
ChatGPT Knows Python and Other Languages Too
I wanted to see what else it knew, so I asked it about Python and creating a smart app to control a lamp via Alexa.
I think the speed at which programs will be cranked out in the future will be light-years faster than today.
chatGPT Can Write Release Notes Easily
If you want to
OpenAI and Pair Programming
OpenAI has already been used in GitHub’s Copilot application, which allows for the OpenAI Codex to act as your pair programming partner. It will suggest code and entire functions to you in real-time, right within your code editor.
It is trained on billions of lines of code across many languages. GitHub Copilot turns natural language prompts into coding suggestions across dozens of languages.
The code that GitHub’s Copilot generates is not always perfect, but it’s a step forward. It should continue to improve which will be a huge benefit to Software Engineers.
chatGPT Will Write Social Content For You
This may not be strictly for Software Engineers, but if you’re an entrepreneur that struggles to put out quality content and post to multiple platforms, it can help you.
I started by asking it to write some funny Facebook posts for my business. It gave me mostly dad jokes.
So I tweaked it and removed “funny” from my request. It began to give me a better list. With more tweaking, I could probably have it write out an entire year’s worth of posts.
The Future Is Going To Be Amazing
Overall, chatGPT has the potential to revolutionize the way that software engineers learn and stay up-to-date in their field. By providing personalized, real-time guidance and access to a vast knowledge base, chatGPT can help software engineers quickly master new technologies and stay ahead of the curve. This, in turn, can help software engineers to be more productive and effective in their work, and ultimately drive the development of better, more innovative software.
I personally see the transition from Engineers writing code to letting Product Owners instruct OpenAI with functionality requirements, and pushing a big “create” button. And out comes a program doing what they want.
That doesn’t seem so remote anymore.
