Blog Software Development Chat GPT by OpenAI - how can it be used? Use cases based on our experience Chat GPT by OpenAI - how can it be used? Use cases based on our experience
Software developers have already embraced the power of ChatGPT, with many citing the tool’s ability to quickly generate code as a major benefit. But the tool is widely used by many other professionals who work in the product development industry. We asked some of the Boldare team to give us some real-life examples of how they have used ChatGPT models. Read on to see how we are embracing this topic.
Table of contents
Disclaimer: every single input written into ChatGPT “feeds” the engine. The tool’s owner, OpenAI, warns potential users:
Please don’t share any sensitive information in your conversations.
This is why we discourage the testing or use of the tool on sensitive data or bigger parts of code that are (or will be) parts of functioning products.
At Boldare, we do not use ChatGPT, or any related tools, to work directly on our clients’ products.
ChatGPT Use Cases from Boldare
Dawid Żurowski, Junior Product Designer
In the world of design, it is crucial to stay inspired and be able to find a balance between design based on established, good practice and creating something unique. ChatGPT is a tool that truly helps by providing designers with sets of rules and good practices, helping with initial research and aiding in tasks, which when done solely by hand, take long periods of time and don’t always yield the most optimal results.
Use cases which I see the most value in are conducting analyses, preparing interviews or usability tests, as well as generating placeholder text. When limiting the use to internal investigations and research, we can truly make this tool an everyday necessity, helping to minimize extended manual labor while allocating more time to the design phase aimed at creating the best user experience.
Kamil Szydłowski, Software Engineer
ChatGPT has proven itself to be a highly useful tool in my day-to-day work as a developer.
It is particularly helpful when dealing with code that I’m not familiar with yet. Instead of going through it line by line, I can ask ChatGPT arbitrary questions about its contents and it does a surprisingly good job of explaining the logic. It can even deduce the context of a particular module and suggest what role it may play in the application as a whole.
Another stellar use case is generating code comments and documentation. It may require some touch-ups every now and then, and it can get somewhat lost in very large or complex files, but it still cuts the time required to complete those tasks by at least ~80%.
When it comes to generating new code, ChatGPT can bring mixed results. Especially with frontend code, where the complexities of module systems and bundling can be overwhelming even to human developers, you shouldn’t expect the code to work out of the box.
Mateusz Rosiek, Distributed CTO Lead
I use it mainly to turn my thoughts into well-written notes. They read much better than my own, and I can also share them with others without worrying that someone will have difficulty understanding them. So, while it’s probably quite surprising for a technical person that deals with coding challenges every day, I prefer to use the tool as an ML-powered, advanced, intelligent notebook.
Paweł Kański, Content Lead
We use one of the tools based on the OpenAI engine to speed up some of our tasks. It won’t create unique content or replace content writers (yet!), but it helps to work with written texts and copy. It makes it easier to experiment with voice and tone and helps to automate simple tasks that support SEO. At the moment, it’s too early to claim it as a revolutionary tool, especially in areas where knowledge is a key medium for the users. Nonetheless, I’m really looking forward to seeing how it will look in a year’s time.
It’s just the beginning
The ChatGPT engine has caused quite a stir across all technology-related industries, but it is worth remembering that this is only the beginning of the development of this particular technology. As Kamil Szydłowski points out:
It’s important to keep in mind that GPT is a language model, not a search engine or an omniscient AI. Its main objective is to generate paragraphs of text that look convincing to the human eye - it does not care how correct or incorrect its answers are. This means that ChatGPT will be perfectly happy to make up fictional information, events, scientific publications, code variables & libraries. You always need to correct for that and verify the answers yourself.
Want to learn more about how to use ChatGPT?
Join us on 23 February at 2pm CET to learn about various use cases of ChatGPT and how it will change the business. Krzysztof Osinski, Senior Vice President of Research & Development from DTIQ, is our special guest speaker and offers a unique inside view into this advanced technology. Sign up here.
Share this article:
