Is Artificial Intelligence a Good Thing for the Future of PHP Developers

In this article you will learn about:

1. What Is Artificial Intelligence

2. When Artificial Inteligence Should be Used or Not

3. Is There Something Better Than Artificial Intelligence

4. How Can a Developer Take Advantage of Artificial Intelligence to Have a Better Career

1. What Is Artificial Intelligence

In simpler terms, Artificial intelligence is a set of approaches to emulate the intelligence of humans using machines programmed using the software.

It consists basically of three phases:

1. Signal detection: collecting information using sensors

2. Machine learning: analyzing and recognizing patterns

3. Decision making:  processing of recognized patterns

For instance, let's say you want to process a picture of a person to detect who that person is.

1. Signal detection: use an image scanner to capture the picture of a person

2. Machine learning:

2.1 Training: Process the picture to get the shapes of the picture of the person

2.2 Pass the shape details to a neural network and tell that person is X, where X is the name of the person

2.3 Adjustment: Pass the shape details again of that person and check if the neural network can identify the person.

2.3 Adjust the neural network parameters if the neural network cannot detect the person in the pícture correctly.

2.4 Repeat the process going back to the step of Adjustment step until the neural network identifies the person correctly.

2.5 Repeat the process with another person going back to the Training phase.

3. Decision making: if the person was detected correctly, give the person access to certain system privileges.

This is just a rough description of how the process works. Some details are not so simple to implement in practice. You need to learn more about those details to implement a real artificial intelligence solution that works well.

2. When Artificial Inteligence Should be Used or Not

As you may have read in the article section above, artificial intelligence relies on successful pattern detection and recognition.

Like all human beings, it may be subject of mistakes. For instance, you may have experienced a situation that when you go to a supermarket you use a credit card and the payment machine refuses your credit card.

Often that refusal is caused by mistakes in the detection and recognition of patterns. In practice this means that the artificial intelligence system had a false alarm assuming that you are not the real owner of the credit card that you used.

This may happen more often on the Internet when you travel to another country the artificial intelligence system may not be aware that you went to another country and may assume that somebody has stollen your credit card.

Often the store attendant, uses another payment machine of a different payment company and that other machine is able to succesfully accept your credit card. The difference is this other payment company uses a different artificial intelligence system that is better prepared to detect if the credit card is really yours.

So in the end the artificial intelligence system only succeed with the help of the store attendant that knew that these payment systems may fail to accept payments due to false alarms. So she gave you another machine to make the payment.

So the answer about using artificial intelligence to develop software systems is it depends.

When you know the pattern of the solution to a problem, like detecting if the credit card is from the customer or is it stollen, you can use regular programming algorithms as you already do when you program your software systems.

When you don't know the pattern of the solution for a problem, like for instance, detecting if a person in a picture is a certain person, or the solution for that problem would be too complex to solve using algoritmic programming, you should use artificial intelligence.

So the answer is, if the problem is simple enough, use algorithmic programing. If you do not know the pattern of the input data or the pattern is too complex, use artificial intelligence.

3. Is There Something Better Than Artificial Intelligence

As you may read in the article section above, sometimes humans are better at solving problems, for instance, using a different payment machine when one device refuses to accept the payment.

The store attendant and the customer decided to use a different machine. Both are humans.

What they have that is better is more experience than the artificial intelligence systems used by the payment systems. Having a greater understanding is what I call Natural Intelligence.

So I suggest always relying on human beings with more experience, even knowing that they are not perfect and can also make mistakes.

Machines using artificial intelligence can also make mistakes because they learn wrong patterns from the past decisions of human beings.

So be aware that the final decision should be yours because you are the primary person responsible for your life's success.

If you do not know something, you can ask other human beings, Google, ChatGPT, or something else.

To know for sure what is the correct answer you need to test your solutions to verify if they work well. This last test can also be automated. Still, the final decision of what to do with the results of the tests should be yours.

4. How Can a Developer Take Advantage of Artificial Intelligence to Have a Better Career

As you may be aware by now, if you read the article above, you should understand that artificial intelligence is just a tool that you may use or not, depending on your problem.

A tool that helps to solve problems is a good thing because it may save a lot of time, and using manual methods would take a lot of time.

I recently tried the ChatGPT solution to check if it could help me to write this article about signing PDF documents. I did that to test how valuable is the ChatGPT tool. I used the tool just to suggest code to generate PDF documents.

I was going to show a picture but right now ChatGPT is overloaded. Non paying customers cannot use it now. Maybe someday I will pay if I have a better use case for the tool.

Anyway, the answer of ChatGPT was helpful, but I had to change it to make it simple. Still, I only used it to write a part of a section of the article. I would say it was less than 10% of the time I spent writing the essay. Still, it was a gain in productivity because I used a tool to automate part of the job.

So my advice to you is to learn about as many tools as possible, regardless of whether they use artificial intelligence or now. That will help you to prepare for your future as a developer, which will be more valuable if you know how to use modern tools like ChatGPT.

Certainly, ChatGPT will be improved to address the valid criticisms it is already getting. Getting criticisms is a good sign for an OpenAI company. Nobody criticizes something that is not relevant.

So keep an eye on the developments of ChatGPT and other artificial intelligence-based products so you can be one of the first to take advantage of those tools when they improve. They will undoubtedly be more reliable and valuable.

So, for now, I suggest you go to the OpenAI site for developers and learn about the uses of the different Open AI APIs. This way, you can be prepared when you have an application to develop that can take advantage of artificial intelligence to solve problems around unknown or complex input data patterns.

I did not use ChatGPT this time to write this article. In this case, that tool was not helpful for me. Still, I used another device that uses Artificial Intelligence called Grammarly to fix my English and make it more straightforward for you.

I even pay a subscription because it helps improve the article's quality in less time than it would take me if I asked an English teacher to fix my English. Still, some fix suggestions were not good because they changed the sense of the phrases. So some tips I had to ignore some.

So, in this case, Grammarly was an excellent tool to help me write the article. Still, it only allowed me to do less than 10% of writing the essay. The rest of the work was done by myself because I had had long experience with artificial intelligence since 1992 when I was still in college doing my engineering degree.

The conclusion is always that you should use the right tool to do the job. The best tool that you can use yourself with your natural intelligence and life experience. Use artificial intelligence tools and other tools to improve the quality of your work.