Is it still worth studying computer science in the era of artificial intelligence? How long until general artificial intelligence (AGI) takes our jobs?

I’ve been hearing this since the late 80s 😃 Specialized systems, CASE tools…no-code… AI. There will be changes, that’s for sure. Will programmers disappear? I highly doubt it, what will happen is a shift in focus. We’ll probably be responsible for coordinating multiple AI agents, learning, etc.

We’re living through a great technological leap in what AI can do and the opportunities are very promising. Image generation, natural language processing, voice recognition, self-driving cars, the news doesn’t stop.

In the early 90s, my professors believed so much in CASE tools that they strongly recommended us to become analysts, as programmers would disappear soon. I’ve been working for 30 years and I’m still waiting 😃. The study to become an analyst gave a boost to my career, since I ended up managing those who were only programmers. As an analyst, you’re closer to the client and the company’s business, which helps in the long run to better understand problems and appreciate solutions.

A lot of people are giving up on the area, but if you think about it, if AGI really comes out next year, there won’t be a 100% safe area.

What I believe:

  • Do a college course in something you want to learn and that motivates you (you study the subject by yourself, even without pressure).
  • Never stop learning
  • Learn to write well
  • Learn English
  • Learn as much math as possible

By studying and preparing, when there’s a change, the person will be ready to adapt… or adapt faster than those who didn’t do anything.

In computer science:

  • Learn and continue studying algorithms
  • Understand what a library does, how it works. Don’t just use it.
  • Learn multiple programming languages. Python, a hard language (C, Rust, Zig) and at least one functional (Haskell, Scala, Scheme, F#).
  • Program, just looking isn’t enough.

Another professor, already in university, explained to me that most of the jobs we had back then didn’t need college. They were mere typists or network administrators, support. A sign for those starting now is if the job you want doesn’t need college… it may be at risk, suspect. Don’t be content with doing without understanding, in repeating without thinking. These are the jobs that will be replaced first.

Don’t believe everything the hype says. Install the LM Studio, download a model, create a local server, make queries with multiple prompts, evaluate the result. See it for yourself. Study how it works, books and videos aren’t scarce.

In the current situation, it’s still very difficult to generate a 100% complete system using AI. The prompts already help a lot, especially in creating code snippets, but you still need to coordinate everything into a system. You have to understand and translate what the client needs to solve a problem. If AI needs a super-prompt written by the client with everything they want to do, our jobs will be guaranteed, since we’ll be writing such prompts. And mainly evaluating the generated code, correcting errors, assembling such code with others to get a system that works in the end.

Futurology is a very easy game to lose. Prepare for the unknown, it’s what we’ve been doing for generations. Even today I haven’t seen a profession with as many possibilities as computer science. In the worst case, you may be able to automate everything in your new profession, but don’t give up and don’t count on AI, nor give up on everything by the possibility that it will exist one day.