Are skills like JQL and SQL still needed?

Magic?

I know it’s not, but some of the recent advances in AI tools are basically magic. They can take a question I type in and translate into something the computer easily understands. While there are certainly other query languages out there, the ones that I’ve used the most are JQL (Jira Query Language) and SQL (Structured Query Language)(I”ll leave it to the comments to argue on how to pronounce it!).

I began using both only after my job demanded I needed those specific tools.  SQL came around when I was asked to help build out reports, and JQL was forced upon me when I was expected to lead a team’s Jira initiatives. Both required me to beat my head into the keyboard repeatedly, until, finally, I got the answers I needed.

At first blush AI would seem to reduce, or eliminate the need to learn how to write queries in those languages. It’s easy enough to just type in a question and have the computer do it for me, so why bother learning how to do it myself? Looking back on my career this could have saved me a TON of effort, energy and time by giving me a way to just type in a question and get an answer.

If you don’t know…

This technology will help a LOT of people get information they need more quickly. Many folks don’t have the time (or the ability to replace keyboards broken out of frustration) to learn a query language.  This means that without this type of tool, they’re stuck with basically three options:

1. Not get any data -  Not knowing how to write queries, or not having someone who can, means you won’t be able to get information. This will result in poor decision making and a general blindness to what’s going on.  Certainly not a good option given the amount of data our systems generate these days!

2. Learn how to query - Personally I’m a fan of this option as it helps expand skillsets and ways of thinking.  It does, however, require time, effort and energy… and not everyone has those. It’s also a skill that you’ll lose without practice… I’ve learned SQL at least 5 times, and every time I have to use it I find myself further back on the learning curve that I used to be.

3. Have someone else do it - Many, many people do this today. They have, or find, an analyst who knows how to talk to computers, and have that human translate their needs. This, however, requires budget and/or headcount, and is not a luxury everyone has when they need it. (Note that analysts do MUCH more than just translate English in a query - they work with you to understand your needs, ask better questions, provide insight into the data and more. Query translation is only a small part of their role!).

Having an AI tool that can create queries for you removes options #1 and #2 as the computer can do it… and that’s something they’re really good at. They won’t forget how subqueries work.  They won’t fat finger a project name in JQL. They won’t be missing a critical piece of info on how the language works. Those AI tools will take over a small part of what #3 can do - create queries. (I find there’s a mis-conception that analysts just write queries all day - they do a lot more!).

AI Advantages

That AI tool also has some advantages a human doesn’t.  It’s likely closer to the data, and related data, than any human could be. It likely doesn’t only understand YOUR data and YOUR question, but the similar questions of dozens, or hundreds, of others. It likely has access to millions of additional (anonymized) data points that it can use to make better queries.  A human doesn’t stand a chance at having all of that, making these tools incredibly powerful in some ways.

These tools eliminate two of the options we have (don’t get answers and learn it yourself) and also and offer some advantages that humans can’t possibly replicate (more data, faster responses, etc). This means that “just ask the computer” will make many, many, people much, much faster at accessing data more quickly.  This will (hopefully) lead to better, faster, decision making.

But wait, there’s more

While I feel these tools will help by eliminating the need to learn a query language, I still feel there is value in learning one. 

For me it’s similar to learning math even though we have calculators, or learning a bit about how a car works when we learn to drive - that underlying information teaches us concepts about how it works, allowing us to better use it and understand the answer that the magic box gives us.

I see several reasons learning queries is valuable:

1. Skill augmentation - For those of us who know query languages, these tools will help us augment our skills. This is similar to how a writer may use a tool like ChatGPT to draft a document, then edit it to their liking. AI can give you a starting point for a complex query that you can then expand on, or write more simple queries faster, saving you time. Eliminating the need to write simple queries frees up time to focus on more complex ones, allowing us to further improve our skills (I read somewhere that tools like this will turn a good engineer into a great one, and a great one into a fantastic one).

2. You’ll know what the machine is doing - Understanding how something works is incredibly important to understanding why it’s doing what it’s doing, and being able to explain it.  Knowing what a “Select” statement is in SQL, or how the “IN” operator works in JQL will mean you can read the machine’s output. At the very least this gives you more confidence in what it has churned out, but will also help you defend its results if anyone asks. I can easily imagine scenarios where someone is challenged on the data they present and only being able to offer “well, the computer told me” as a defense… (this would not be good!).

3. You’ll get smarter - Similar to augmenting your own skills, you’ll be able to expand how you think. The tool will have a different approach than the one you use in some cases, exposing you to new use cases or features you haven’t used yet. I find I tend to only have narrow use-cases for tools, resulting in me missing concepts or features outside that focus. Using AI tools will help break me out of that narrow window by showing me other ways.

Parting Thoughts

AI tools are here to stay.  They’ll continue to augment and improve our skills, and will continue to become more and more powerful as we understand them more. They are, and will continue to be incredibly useful in helping humans do more faster.  That said, if we’ve learned anything from comics books with that great power comes a great responsibility - not only to use it responsibly, but to understand how it does what it does.