You Don’t Need SQL Anymore? That’s How Data Disasters Happen

I recently saw a post on LinkedIn claiming that analysts no longer need to learn SQL because AI can write SQL queries for them.

I strongly disagree with this point of view.

Yes, AI can write SQL, but that doesn’t mean analysts should skip learning it.

SQL is a foundational skill. It isn’t just a syntax skill.

It’s how analysts think about data, how various data tables relate to each other, how metrics are defined, how joins affect results, and how filters subtly change meaning. If you don’t understand SQL, then you don’t truly understand the numbers you’re presenting.

You, the analyst, understand your business, your stakeholder and the context; AI doesn’t.

It doesn’t know which table is the source of truth.
It doesn’t know which join causes duplication.
It doesn’t know that using a LEFT JOIN instead of an INNER JOIN just inflated revenue by 18%.

It will still confidently give you a query, and if you don’t know SQL, you’ll confidently run it and share the results with your stakeholders.

That’s how most data disasters happen, and that’s how trust in data erodes.

You should already know by now that AI makes mistakes, it hallucinates, so you can’t trust it 100%. At least not yet.

However, if you don’t know SQL, then you can’t spot where AI is making mistakes. So, skipping the fundamentals, you are heading towards a disaster.

I use AI all the time because AI is incredibly powerful. In this context, it is powerful for those analysts who already understand SQL.

AI can speed up writing queries.
It can help you explore unfamiliar schemas.
It can reduce time spent on syntax and formatting.
It can even spot inefficiencies or suggest optimizations.

But it only works safely when the analyst can:

• Read a query and understand every join

• Question assumptions baked into the logic

• Validate results before sharing them

We all use calculators. But we know that a calculator doesn’t replace math skills, it amplifies them. If you don’t understand math, you won’t know which buttons to press or when the answer is wrong.

The same applies to SQL.

So no, analysts shouldn’t stop learning SQL.

They should learn SQL and then use AI to move faster, not think less.

That combination of human judgment plus AI acceleration is where real leverage comes from.

I hope that this makes you pause before skipping the fundamentals.

SQL is not an optional skill for analysts; it’s foundational. AI doesn’t replace that foundation; it builds on top of it.

Analysts who truly understand SQL will build long, resilient careers.
Analysts who rely on AI to generate SQL without understanding it are taking a shortcut that eventually catches up with them.

If this perspective resonated with you, feel free to share it with someone who might be navigating this decision right now.

If you have any questions or a different perspective, feel free to leave a comment.

—
Anil Batra
Your Digital Analytics Coach and Guide
Connect or follow me on LinkedIn: https://www.linkedin.com/in/anilbatra