OpenAI's ChatGPT has caught the globe by storm with its ability to produce replies that seem human. The artificial intelligence (AI) software aids experts in conducting research, writing articles and writing code in many different languages. Let's test it out to see whether it aids in the study of electronics for engineers.
Asking research questions
Let's begin with an example from semiconductor devices. An engineer might find it useful to read up on recent developments in transistors. New studies in this area may be of interest to a researcher. So, it may give ChatGPT prompt of:
“Tell me about the latest research papers related to carbon nanotube transistors.”
In response, ChatGPT stated that it is unable to recommend any articles published after September 2021 due to a lack of relevant data. However, it can share some paper titles before that and unfortunately, those research papers are also made up and don’t exist on the internet. For example, to the above prompt, ChatGPT replied with:
"As an AI language model with a knowledge cutoff in September 2021, I don't have access to real-time information or the ability to browse the internet. Therefore, I can't provide you with the latest research papers on carbon nanotube transistors. However, I can provide you with some prominent research papers published prior to my knowledge cutoff that may be relevant to your interest:
"Carbon Nanotube Transistors on Flexible Substrates" by Cao et al. (Nature, 2018): This paper discusses the development of carbon nanotube transistors on flexible substrates, enabling applications in flexible electronics and wearable devices.
"Scalable Fabrication of High-Mobility Semiconducting Carbon Nanotube Thin Films" by Zhang et al. (Nature, 2018): The researchers demonstrate a scalable approach to fabricate high-mobility semiconducting carbon nanotube thin films, which could be used in the production of large-area transistors and displays.”
But if you try to find these articles online, you won't find them since the titles, authors, journals and publication years are all fabricated.
[See also: ChatGPT prompts for electronics engineering and PCB design]
Asking general questions
Now, let’s see if ChatGPT can give accurate answers to general technical questions. When asked about carbon nanotubes, it replied with:
“Carbon nanotubes (CNTs) are cylindrical nanostructures made of carbon atoms arranged in a hexagonal lattice. They are essentially rolled-up sheets of graphene, which is a two-dimensional form of carbon consisting of a single layer of atoms arranged in a hexagonal pattern. CNTs can have diameters ranging from a few nanometers to tens of nanometers, while their lengths can extend up to several centimeters.”
This information is correct so it can be concluded that it gives accurate responses to general straightforward questions. As a result, it may serve as an excellent resource for teaching electronics to newcomers, enthusiasts and students. The program can produce a large amount of information in a matter of minutes and can also help in giving latest research ideas. However, in order to get the greatest results, the information sometimes requires substantial editing and fact-checking.
Solving technical problems
ChatGPT may be asked by an engineer to solve an issue using a specific programming language, such as MATLAB or Python. Though the ChatGPT does provide a code, it is likely to be too general to actually fix the issue. For instance, it may create code that works well for a forecasting problem but requires tweaks before it can be used for an optimization problem; thus, ChatGPT is of little assistance. The engineer or programmer must engage in independent creative thinking.
ChatGPT may also analyze the input code to locate and repair a bug. However, it is unable to process lengthy lines of code and provide appropriate solutions. It has been observed that it frequently responds with false information and creates reference connections to support data that do not exist. ChatGPT can help engineers or students with their homework by providing the essential procedures for solving difficult electronic-related mathematical equations. Boolean expressions, mathematics and common signal-processing methods like Fast Fourier transform solutions are all covered.
Email and technical support for engineers
Email support is a means used by many companies and organizations to communicate with their customers and provide answers to their inquiries, details about their products, and help with any issues they may be having. Engineers working in such industries may benefit from ChatGPT's ability to automate email responses, create brand-new emails and enhance existing ones. Moreover, a database of frequently asked questions and their solutions may be used to teach ChatGPT. Because of this, it can deliver prompt responses to frequent requests without requiring support personnel to continuously type out the same answers by hand.
Conclusion
ChatGPT helps electronics engineers save time on research despite a number of limitations to its use. ChatGPT can provide answers to the kind of inquiries an engineer might type into Google before doing some light reading to find the best potential solution; nonetheless, a quick internet search is still recommended. It can conduct basic research, examine facts, code, design and generate content. To sum up, ChatGPT is a helpful AI tool for electronics engineers.
