Navigating the landscape of professional use for AI tools like ChatGPT can sometimes feel like tiptoeing around a taboo. There’s a lingering perception that using such tools might be akin to taking shortcuts or even cheating. However, it’s crucial to recognise that, much like search engines and YouTube tutorials, AI serves as a valuable support system, aiding us on our learning journey.

AI tools haven’t reached a stage where they can completely replace human tasks. Mistakes happen as they learn, and they may lack that elusive ‘human’ touch in certain responses. What sets AI apart from its predecessors is its ability to present information in an easily digestible format. It offers succinct, step-by-step explanations and often includes examples, making it a powerful ally in enhancing our understanding.

My experience so far has also been that men with are much more comfortable using AI tools in the workplace than women. I think this is caused by the ‘shortcuts’ and ‘cheating’ stigma that using AI technology for work brings with it. Women often feel that they have to work harder than men to achieve the same levels of promotion and respect, particularly in STEM fields. Using AI tools would make their work easier which may make them feel like an imposter, or that they aren’t good enough at their job. They should be able to ‘do it without any help’. To people who may feel uncomfortable using AI tools because it may make them feel like they don’t work hard enough, I say this: Work smarter, not harder.

To that end, I want to share a few ways that I have used AI tools to enhance my work, and make my life a bit easier, and encourage you all to make the most of AI tools in your careers too.

Documentation

The importance of documentation when working as a software engineer cannot be understated. Having code that is poorly documented can be the difference between spending 10 minutes on a problem vs. 2 weeks. Many software teams today understand this, and generally accepted best practice in our industry is to document code with comments, docstrings and stand-alone documentation like READMEs, wikis and user guides.

Sometimes though, teams can inherit large amounts of legacy code that has little to no documentation. Teams understand that they need to edit the code to document it, but depending on the amount of code there is, this can be a hugely daunting and time-consuming task.

Other times, software engineers and developers can get carried away in their rapid prototyping and write code whilst forgetting to document it. Sometimes, people writing code just don’t want to add comments because they feel like it’s a waste of time.

To combat all these scenarios, enter AI tools. AI tools is a great tool for adding and writing documentation to and about code. It can digest a few thousand lines relatively quickly – understanding the basic principles of what the code is attempting to do, and breaking that down into comments, docstrings, type hints and other forms of documentation.

This can remove some of the burden from a task that some may find boring, and help encourage good habits when writing code.

Creative Block

I am not the most creative person out there. When I need to write an abstract for a conference, develop a talk or presentation, or sometimes write a blog post, there are times when I am stuck for inspiration. This can be known as writer’s block or creative block. There are lots of techniques people use to overcome this block, but personally – I use AI tools.

You can provide the tool with prompts to help you produce a first draft of whatever it is you’re writing. When it comes to prompts, the more specific you can be, the better. Telling the tool your audience, the desired length of the response, and the tone of what you’re writing can help you generate a more useful draft to start off with.

One thing I sometimes do when using AI tools for this is to add at the end of the prompt ‘ask me any questions you need to get enough information to generate your best response’. This means that the tool will respond to me with a few questions (normally no more than 10) which I can answer to get a fully customised and appropriate response to my original prompt.

I then use the response that the tool generates as a basis for my piece of work. I will often end up rewriting parts of it to make it sound more like me, or to convey different points – but the key is that the AI tool got me past my creative block and allowed me to start creating something that I wouldn’t have been able to otherwise. It often also writes something I may not have thought about, giving me new ideas and perspectives based off what it has learned.

Debugging

This is probably my biggest use of AI tools. When I’m working on a piece of code and I see an error message or warning that I may not have seen before, don’t fully understand and/or don’t have error logs for, I copy and paste it into an AI tool. AI tools have access to information from across the internet, and so can find causes of these errors much faster than I can!

If it can’t find an exact solution, it often provides a useful set of steps to assist with debugging – providing commands you can use in the command line to get logs and helpful metrics, identifying common causes of similar problems etc. If you are lost in a problem, AI tools can show you a possible way out.

Learning

Finally, AI tools are excellent for helping you learn about new things. If you want to learn about a new technology for example, AI tools can provide you with examples, summarised lists of key points, links to further resources on the internet and step-by-step instructions on how to use that technology.

It’s worth noting that some free AI tools, like ChatGPT 3.5, may not have had a knowledge update in some time and so some of their knowledge will be outdated and/or incomplete.