I recently received a proposal to dedicate my time to a job.

In the proposal, shortly after beginning the conversations, a challenge was included, creating a code with microservices, using hexagonal architecture, as they commented.

It occurred to me that since I have reviewed many job offers and I have not seen such a challenge in any of them, doing the task in the unusual way in which the unexpected challenge was proposed, using code generated by artificial intelligence, would be my challenge and would test its effectiveness.

What I did was copy the description of the problem they posed, which is in a Github repository, and give it to Artificial Intelligence, so that it could solve it, because that’s what it’s about, that it works for us, not us for it.

The code was generated, reviewed, so that it at least compiled and sent to those responsible for the challenge. They didn’t like the result, because they said they wanted to see how I thought, to reproduce or understand it, I suppose. They told me to review it and do some code by hand.

It is really understandable that being used to writing code by hand, code that is not written by hand, but that solves the problem, is not accepted by definition. I guess they wanted to test what is indicated on my resume and generate code, I guess, like AI does.

What occurred to me was to add functionality, write a microservice that they had not requested, but since it was a demo and a personal experiment that I was doing, it did not matter much, so I sent it to them.

They gave the answer with an AI, I didn’t respond.

And this is the result of the experiment I did, for such an unusual challenge proposed by those responsible.

Artificial intelligence is capable of generating code for problems posed to it; it generates functional and complete code that conforms to the definition of the problem.

The result can be expanded or modified with subsequent queries and generates new functional code, incorporating the modification that we indicate.

The problem must be specified very well, so that the result is adjusted to what is wanted.