We asked ChatGPT if it would manage the MITC exam. Here's its response:

Translate
You:
Would you be able to pass the MITC exam for testers, where you need to interact with software to solve tasks?

ChatGPT:
As an artificial intelligence, I don't have the ability to interact with software in the same way a human does. My functionality primarily relies on processing and generating text, as well as providing information and advice based on accumulated knowledge. However, if you have theoretical questions regarding software testing, different testing techniques, tools used in testing, or other related topics, I'd be glad to help you answer them.

It is clear that current AI cannot perform tests that are typically done by manual testers. However, it does have the ability to generate answers to questions or test code. The next step was to test ChatGPT on a MITC trial exam. Since ChatGPT doesn't interact with the exam itself, we had to provide it with full exam information.

1. Tasks like designing test data were transformed into questions:

None of the given sequences has 30 characters. 0 points.

2. Tasks like reviewing test run reports were also adapted into questions:

Validation worked, so the result was incorrect. 0 points.

 

The final exam result:

 

0% exam is failed. ChatGPT couldn't handle this task.

 

How would it fare in the actual exam?

Because we're using a real exam, we are unable to disclose the content of specific questions, but we can describe the level of difficulty faced by AI.

1. Theory questions:

In the MITC set, there might be zero to one theory question. We used AI to answer a question, but it was only partially correct.

According to the statistics, the AI-powered assistant had a 33% success rate in answering one question, which indicates a potential 1% correct answer rate throughout the entire exam.

2. Tasks involving running test cases:

The current generative AI cannot run tests and determine whether a test was successful or not.

AI failed. 0% success rate.

3. Tasks involving designing test data:

To determine test data during MITC exams, it is necessary to interact with the software. ChatGPT can't do this, so we decided to assist it with our intelligence, describing the guidelines on the field and the expected test value. AI succeeded in 1 of 3 cases, earning 1 point towards the final score.

4. Defect discovery and reporting:

AI can't search for defects in software, thus can't report them either. With our assistance, in describing the defect and requesting a defect report from ChatGPT, we noticed that to match its response to the required defect report, we'd practically have to provide all the data. 

AI is not capable of identifying defects in software, so it cannot report them either. With our help, we attempted to describe the defect and requested ChatGPT to provide a defect report. However, we realized that to get an accurate response, we would have to provide all the necessary data ourselves.

Translate:
You:
Create a defect description for incorrect validation in the phone number field - thera are some accepted characters that shouldn't be.

ChatGPT:
Title: Error in Phone Number Field Validation - Acceptance of Inadmissible Characters
Description: While entering data in the phone number field on the registration/order form (or any other place where this field is used), a validation issue was observed. Currently, the phone number field accepts characters that should not be allowed. Inadmissible characters, such as special characters, letters, or other non-standard characters, are currently tolerated, leading to the input of incorrect data.
Steps to Reproduce:
1. Go to the registration/order form.
2. Navigate to the phone number field.
3. Enter inadmissible characters such as special characters (!@#$%^&*), letters (abc), or other non-standard characters.
4.Notice that the form accepts these characters as valid data for the phone number field, despite their inadmissibility.
Expected Behavior:
The phone number field should only accept digits and possibly special characters used in international numbers, such as plus (+) or hyphen (-). Letters and other non-standard characters should be rejected as incorrect data.

We consider AI unable to handle this task.

5. Metrics tasks:

Some questions assess the criteria for ending tests and require basic testing knowledge and simple math. ChatGPT handled these flawlessly, earning one point.

 

The ChatGPT final score is 2 of 26 points, which is approximately 8%. Therefore, ChatGPT is unable to pass the MITC exam, as the passing threshold is 80%.

 

Therefore, MITC is unable to answer any MITC questions without human support to compile information from the task and the application. Even with assistance, it can't achieve a passing score. Randomly selecting answers in the MITC exam, including binary questions, would likely yield better results than attempting to provide the correct answer.

This simple experiment leads to the conclusion that MITC is a better evaluation method for testers than the ISTQB® exam, when it comes to assessing knowledge, work, and thinking skills.

Finally, let's ask ChatGPT which certification is considered more effective in verifying the knowledge and skills of testers?

Translate:
You:
Which exam for testers better verifies knowledge and skills - Modern IT Certification or ISTQB?

ChatGPT:
Assessing which exam better verifies knowledge and skills can be subjective and depends on many factors, such as:

1. Scope of material: ISTQB offers a structured certification program covering a wide range of topics, from basic to advanced aspects of software testing. On the other hand, Modern IT Certification (MITC) may focus on more specialized and advanced topics, but the scope of material might be narrower.
2. Difficulty level: ISTQB Foundation is generally considered accessible to beginners, while advanced levels may be more demanding. MITC could be more advanced and demanding, but it may depend on the specific exam and its level.
3. Recognition: ISTQB is widely recognized worldwide and often required by employers as a standard certification for software testers. MITC, being a relatively new certification, may be less known or recognized in some industries or regions.
4. Relevance: MITC may have an advantage in terms of relevance as it may be more aligned with modern technologies and current trends in software testing.

Ultimately, which exam better verifies knowledge and skills will depend on your individual career goals, level of experience, and preferences. It's worth researching, comparing exam materials, and reading reviews from others who have gone through the certification process to make an informed choice.