What Questions Should I Ask Before Taking an AI Job?

I have sat on the hiring side of hundreds of these talks. So I know which questions tell you the most about whether a job is actually good. Most candidates skip them. They ask about salary and title. What they should ask about is the stuff that decides whether they will grow, be treated well, and still have a job in two years.

Here are the questions I'd ask before accepting any AI role.

1. "Is this a real AI role or an AI-flavored version of an old role?"

Many "AI jobs" posted today are old roles in new packaging. The word "AI" gets added to the title to pull in candidates and to justify lower pay. Ask what share of the work really uses AI tools and methods. Then ask how much of it is the same work people did before. A real AI role will have a clear answer. A repackaged one will get vague.

2. "How does the company actually use AI, today?"

Some companies have AI built deep into the work. Others have one ChatGPT subscription and a press release. Ask for real examples of how AI shows up in the job you would be doing. How exact the answer is tells you a lot. You will either do real AI work, or you will be the one who has to invent the AI strategy from scratch. That can be great or awful, depending on what you want.

3. "Who would I learn from?"

In a field that moves this fast, the people around you set how fast you grow. Ask who you would work with and what they have done before. A role where you are the top AI person and no one can teach you may pay well. It can also stall your growth. Early in your career, being close to people who are ahead of you matters a huge amount.

4. "How does the company think about AI replacing parts of this role?"

This is the awkward question, and you have to ask it. Ask it straight: as AI tools get better, how do you see this role changing? A thoughtful employer will have an honest answer about how the role grows into higher-value work. One who has not thought about it, or who gets defensive, is telling you something. They have not planned for the thing that will hit your job security hardest.

5. "What does success look like in 6 months?"

This shows you whether the company has thought the role through or is just hiring on reflex. Clear success metrics mean clear expectations. They also give you a fair shot at hitting them. Vague answers mean you get judged against a standard no one ever wrote down. That rarely ends well.

6. "How is compensation structured, and how does it grow?"

At startups, a lot of the pay for an AI role comes as equity. Ask what the split is between cash and equity. Ask what that equity is really worth. Ask how pay grows with performance. AI hype has puffed up some equity packages that may never be worth what the cap table implies.

7. "What happened to the last person in this role?"

One of the most revealing questions you can ask in any interview. Did they get promoted? Leave for a better opportunity? Get let go? The answer tells you about the role's trajectory and the company's treatment of people. A pattern of churn is a red flag.

The questions I love getting asked

When a candidate asks me these kinds of questions at TTGC, it tells me they think like an owner, not an employee. They are working out whether this is a good use of their time and talent. That is the exact mindset that makes someone valuable. The ones who only ask about salary and vacation days are telling me something too.

The honest framing

A job is a two-way decision. The company is sizing you up, and you should size it up just as hard. That goes double in a field moving as fast as AI. A wrong choice can mean two years in a role that does not grow your skills while the field passes you by. Ask the hard questions. Good employers will respect you for it. The ones who do not have just told you something important.

Sources

Related reading: What to Ask Before Hiring an AI Development Team · How to Negotiate Salary for Your First AI Job