How much do you agree with this statement: "AI in Testing will benefit all testing, QA and quality engineering professionals"?

🙋 Questions
,
1

We would love to know your thoughts - do you agree that “AI in Testing will benefit all testing, QA and quality engineering professionals”?

Do you have any strong opinions? What will those benefits be? If you disagree why?

  • Strongly disagree
  • Somewhat disagree
  • Somewhat agree
  • Strongly agree
0 voters
2 Likes
2

So I’ve said strongly disagree. “OMG Cal suuuuch a strong viewpoint” I hear you cry.

My reasoning is that I think some organisations will implement AI badly and this will have a negative impact on testing overall that in the future we’ll have to unpick. Where it’s implemented well, many good experienced testers will also lose work or be pushed into work that they don’t enjoy. So i didn’t think this was a benefit to them.

2 Likes
3

Somewhat Disagree!

Why? I have reasons to explain:

  • AI has created, “creating bullshit” easier.
  • Testers will have to fight this “bullshit”.
  • “ALL” is an absolute and as a tester I know that there is something good for someone with any new change and also something bad for someone with any new change.
  • Only people who know good testing will be able to leverage AI to something meaningful and good. Good testing skills are still rare.
  • AI comes with it’s limitations too. Most people just ignore it. Accepting duality is complex and needs a lot of cognitive dissonance at times.

However, I am still optimistic that it will increase the value of good testers.

2 Likes
4

Somewhat Disagree. In theory it’s another tool that testers can use, however it’s likely to lead to dumbing down of the role when managers assume that any old person will be able to test (and deadlines will be squeezed even more than they are already) - not to mention the unrealistic expectations that vendors will set…

5

I’ve put somewhat agree. Because I’ve been using it on a specific project for about 8 months now. Everything from helping me with testing architecture to monitoring and alerting in production. Analysis of AC, creating POM files, converting scripts, performance tests anything in the whole SDLC.

Was it perfect? No… As I mentioned before, I believe you become an “QA-AI Reviewer” like a Dev Team who does 80% code review and 20% adding his own stuff. This is exactly how it is.

I see people doing “disagree” but have you actually used it for several months? Not just “once” ? It helps me greatly with the boring stuff. It requires a mature environment to use it, because:

If you put shit in your AI-Agent, you can only get shit as an outcome
Compare it to writing a shit analysis, you’ll only get a shit product. It’s not the same damn thing, this does not make it a “product problem” but an “analysis problem”
Hence it’s not always an “AI problem” but an “information or prompting” problem.

People need to learn how to prompt and put in valuable information. I literally cannot see how one would disagree with the statement (based on my experience !!! ) even the smallest thing is “help” - making a list of things, yea you have to review it but you also have to write it yourself otherwise and come up with “everything”. Sometimes there are things in there that I wouldn’t have thought about so it helped :slight_smile:

Don’t get me wrong here, sometimes the output is indeed not what I expected, I have only shared my good experiences here… but yea there are bad experiences (and I had to learn how to prompt and train my AI-Agents) also but overall if I look back I would say it helped me more then I have been frustrated. Q:AI in Testing will benefit all testing, QA and quality engineering professionals
So: Yes

And that’s why I choose “Somewhat agree” :slight_smile: <3

But this is an implementation problem and not an AI problem.
With this reasoning, if I implement test automation badly, “all test automation will not benefit testing” and that’s not the question.

With this I agree, the rest I do not. Testers fighting bullshit is either having a badly trained AI-Agent or prompt engineering, yes there will be times where the answers are shit but in most cases, it’s just fine.

AI is indeed limited to what it learns and knows, that’s why you have to keep training and feeding it new information. And still it will go badly, it’s a never ending process but as I said it’s 80% reviewing and 20% writing yourself.

6

Oh, as the original question was “will it help quality engineering professionals”, I didn’t read it as will it benefit testing overall. My comments were framed around that, with me feeling that a bad implementation would not help all testing professionals (I also agree that bad automation implementation likewise means that automation doesn’t help every tester).

If the question is “will it help the wider discipline of testing more generally?” then I’d answer somewhat agree; because as you’ve said it does take away and speed up some basic things. Although there’s a big risk of anti patterns and groupthink biases that could cause a wider community problem unless considered early.

(Yes I have been using it more than just once)