AI: More Testers, More Money

jason arbon
5 min readJan 22, 2024

--

There is an AI divide coming — and quickly. AI will create new haves and have-nots in the software testing world. The introduction of AI to software testing will change the jobs of testers, vendors, and tools. But, the change won’t be in the ways most folks are fearing or dreaming. The narrative that AI will simply replace the work of human testers soon— is wrong. The wishful thinking that AI won’t be smart enough to do much of the testing work, is also wrong. There will be more testing jobs, and they will soon pay more. Some will lose their job in this transition, but it is up to them. Only the open-minded and optimistic testers will survive this transition — and they will be heavily rewarded with job security and compensation.

AI’s Impact on Testing Jobs

AI won’t reduce the number of testing jobs in the near term — there will be more demand for testers. Many testers fear AI will eliminate their job — this isn’t true for most. “AI-assisted” testers, those who apply AI to their job, will be more productive. Those who do it quickly will see the most job security, progress, and rewards.

Perhaps the biggest question in the $10B+ world of testing is how to measure value. Engineering teams know they need testing, but it is difficult to quantify, and the reality is that most testing efforts are marginally useful today, without AI. When testers are assisted by AI — the speed, scope, and value of testers will grow to the point of obvious value. Most software testing today barely covers basic functionality and some edge cases, at great expense, and slows the rest of the engineering team down.

AI-Assisted testing however will give these human testers ‘super powers’ and their contribution will be overwhelmingly obvious. As a friend in the testing world described it last week — soon, testers will wear AI-‘Iron Man’ suits. The business will want to add more AI testers because they finally see an obvious positive return. These AI-assisted testers will be fast enough to be useful in the current sprint, instill enough confidence to ship, catch enough issues that impact the business, and expose what is still left to be tested. There will be increased demand for AI-assisted testers because they will add visible value to the business — just like a developer that adds features and fixes bugs.

AI, also means that developers are creating more software. The speed-up for AI-assisted engineers is very real, and AI also means more people can create software. All that new software needs to be tested by more AI-enhanced testers.

The number of testing jobs will increase in the near term — but not for all people. Those testers who don’t embrace AI will soon be lost to the sands of time. Some vocal testers will cry into the winds of change, as they are swallowed up by technological progress. Interestingly, their jobs won’t be lost, they will just be let go for a new crop of Testers who are excited about testing. and eager to leverage AI. Those jobs won’t disappear, but they will be populated by the testers embracing AI.

Testers will be paid more thanks to AI. Effective wrangling of AI to help with testing tasks isn’t easy, and requires better testers. AI will replace the basic testing — perhaps even do most of the testing. AI can generate hundreds, thousands, even millions of permutations of tests — and execute them, all without humans. Engineering teams will need AI-assisted testers to decide how to spread the scarce compute/time resources across all possible coverage. A key role of AI-based testers is to customize the testing bots to align with the team and business goals. These AI-assisted testers will be needed long after the developers that are automated by AI.

“A key role of AI-based testers is to customize the testing bots to align with the team and business goals.”

No more hiding in a cubicle or in JIRA — these AI-assisted testers will actively communicate with the other humans on the team. Top-notch communication skills focused on efficiency and clarity will be in high demand. Non-AI Testing today means Testers don’t get to exercise these skills because basic, inadequate coverage is easy to understand and execute. AI-assisted testing means engineering teams will demand more experienced testers and more of them.

“Prediction: Testers who left the field to be Product managers or Developers will soon flock back to the relatively high-paying and secure senior testing jobs”.

The AI-Divide

During this transition to AI-assisted testing, there will be many negative voices. “Only humans can Test”. “AI isn’t reliable enough”. “Test Automation didn’t take my job...” “I’m still needed, because...”. “AI can’t think.” These folks won’t disappear overnight, but they will quiet down. These are folks proactively self-selecting out of the testing population.

A few will realize their folly ($/Market are powerful forces), start applying AI quietly, and delete their anti-AI social media posts. More will be politely “let go”, and be increasingly unable to find another testing role.

Sadly, some testers will succumb to the very biases they preach to avoid in their testing work. They won’t admit a personal bias for continuity. Many will succumb to Survivor Bias. Their teams and circles are an echo chamber of emotional resistance in a technical field. Ironically we’ve even seen many self-declared expert testers already decide AI is too early, too dangerous, or ‘not useful’ before they even ‘tested’ the new AI-assistive technologies. Follow these negative humans, and memes, at your own risk!

Hint: Late adopters will continue to be late adopters, and that isn’t good for job security in a fast-moving world of AI.

Some folks are slower to adopt new technologies or are more risk-averse simply because of their life situation. As a testing community, we should try to help these folks by making sure they are aware of what is happening in our field. We should build AI-assisted testing in a way that is easily adopted and applied to testing. We should continually demonstrate and share the value of AI-assisted testing. We need to make sure folks aren’t simply left to the whims of vocal legacy testers and folks with a personal agenda and bias to maintain the status quo.

Summary

AI is a litmus test for testers. AI is quickly doing more testing but it also increases the need for experienced human testers. AI will also expose the negativity and bias in legacy testers, and they will soon disappear. Not even testers are happy with the state of software testing — but AI-assisted testers will quickly change this reality.

AI is a litmus test for testers.

In the long term — the AI will learn from all the AI-assisted testers. Both AI and AI-assisted Testers will continually up-level in a virtuous loop. This virtuous loop will make software better and improve the lives of the software testers who proactively embrace it.

— Jason Arbon

--

--

jason arbon

blending humans and machines. co-founder @testdotai eater of #tunamelts