Computerra: ChatGPT Making Us Dumber?

2025-06-27 14:24

Generative AI models like ChatGPT can produce dozens of test cases in minutes. However, without careful review by a human expert, the output can be useless — or even harmful.

According to Svyatoslav Veliev, Testing Technical Lead at Selecty, the efficiency of AI depends entirely on the intelligence of the person operating it:

«Yes, without it the pace may be slower, but core skills don’t disappear. AI simply shifts the focus of tasks and you no longer need to spend time on routine calculations, so you can concentrate on more complex challenges. The real question is how you make use of that freed-up capacity».

He points out that AI today is still primarily a suggestion generator, incapable of replacing human oversight. Every project has its own nuances and hidden pitfalls that neural networks are not yet capable of detecting. A single mistake in testing can lead to bugs in production, while an error in text has the potential to damage a company’s reputation.