<strong>Even with Copilot, you should learn to program</strong>

A few days ago, Google Trends announced a major interface change with new features. Among them, and perhaps one of the most useful, was the ability to obtain real-time data on searches related to a keyword. It's a very useful tool for understanding what's on people's minds. In fact, at the beginning of the pandemic, a post went viral showing a perfect correlation between Google searches for "loss of smell" and subsequent positive COVID-19 cases. When we want to learn or find something, we go to Google. We type in anything. Until the use of natural language processing systems like ChatGPT becomes more widespread and commonplace, we'll continue to rely on Google. That's why tools like Google Trends are so useful: they tell us, by geographic area, what people are concerned about.

The keyword “programming” shows an interesting evolution . Clearly, in recent months, it has become a topic of interest for society. It reached peak levels, but hasn't declined. Surely now, with generative artificial intelligence , it will experience a new boom. It's easier than ever to learn to program. We at Brain and Code have already incorporated it into all the workshops and bootcamps we offer our clients. We'll explain how, but first, let's understand what learning to program really entails.

Programming languages ​​are much simpler than natural languages. However, learning to program seems very complex to us. Human-machine dialogue, communication with a computer, has a strong logical component. This is precisely the first thing we must understand. The syntax we use to tell a computer to execute a set of instructions must be ordered, assembled, and combined to produce a result. But this isn't just stacking bricks; we need to know how to do it.

Computers, modern computing machines, unlike other machines, have a general purpose. That is, they aren't programmed to heat food or sweep our floors. They perform many different functions. They have a powerful processor and a language in which to represent those functions. Everything they do is controlled by a program, which includes what we mentioned earlier about the sequence of instructions in a language, which we can call "programming languages." When we at Brain and Code say that we want to offer an education that makes computing accessible to everyone, what we really mean is learning to master these machines.

This language for communicating with the machine is much simpler than English or Chinese. It has a simple structure. For example, the C programming language (one of the oldest and most widely used) has 46 syntax rules. In contrast, English grammar, as David Crystal states in *Making Sense of Grammar* , appears to contain around 3,500 rules. Furthermore, a programming language is precise. That is, it does not allow for ambiguities in its interpretation. A program is designed with a specific purpose in mind. However, a sentence in a human language can lead to different interpretations, which complicates its learning.

As communication languages, both English and programming languages ​​(of which there are many, just as there are different languages) have more favorable contexts than others. For example, we wouldn't use Java (another language we teach at Brain and Code) to arrange to meet our friends at 8:00 PM to watch a Real Madrid match. But if we need to calculate all the prime numbers between 1 and 1,000, a programming language is more appropriate, since it doesn't allow for ambiguity in the calculation. We went to Perplexity and asked it to program this exact program in Python . It returned the following:

Any idea that requires an algorithmic language (calculations in physics, formulations in chemistry, route optimization, rules in language, etc.) is ideally suited to be expressed in a programming language. Therefore, we believe that in the future, it should be part of mandatory school curricula. We cannot afford to exclude this knowledge from generations who will have to interact with machines in their daily lives.

The initial reality, however, offers little cause for optimism. Less than 1% of the population knows a computer programming language, and those who do almost certainly already have a job and a good salary. When we want to learn English, we speak the language, spend a summer in Ireland, or ask our company to send us somewhere we can practice. How are we going to learn this language if we don't have places to practice it?

Furthermore, with the advent of generative AI, we have new assistants. When we want to write software code in any programming language, we can now rely on GitHub Copilot. In fact, the name of this tool reflects our vision of the future human-machine coexistence: copilots (technologies) accompanying pilots (humans).

New languages ​​and computer literacy are here to stay; they will be necessary to be competitive in the future and also, why not, to enjoy the wonderful logical challenge they present.

Brain and Code ©
March 25, 2023