AI is currently killing my business and threatening my old CS career, but I have things to retreat to: a network, friends, hobbies, nature, art. I find answers in the countless books and essays I have read, instead of flying blind. Liberal arts matter.
I did not see the use of those things when I was laser-focused on my career, but they added so much richness to my life outside of work, and now they're the only thing between me and despair.
That aside, do not neglect networking, negotiation and other soft skills. They do much more for your career than being a really solid coder. Software engineering is engineering. It requires herding a bunch of human beings to achieve unclear, changing goals. Code is just an artefact of that endeavour.
The other thing is get experience or look at companies that aren't technology companies, look at sciences, manufacturing, etc. and services they need. Maybe consider a minor in some other field that also interests you and focus on companies in that field.
But the advancement was so exponential and fast that by the time I finished college, the profession that I was studying almost became obsolete due to AI. No company was hiring juniors anymore, and if they did, they were requiring 4+ years of experience.
My advice is get familiar with coding agents as soon as possible and start learning sales and business fundamentals, and make your own product. If it fails make a second one, and just keep trying until something works out.
Or turn towards jobs such as carpentry, ceramics, electricity etc. Those jobs are experiencing so much scarcity in workforce that it is practicaly imposible to not get clients.
I am in the same situation as you are in terms of how the improvements to LLMs aligned with my education. On top of this sounding like a very depressing career, do you actually know people who are reliably using this strategy? New projects are done for - unless you strike the gold in the top 1% and break through, anything you make will be drowned out by thousands of projects that do the same thing. Almost all easily solvable problems have been solved, and the rest often require a non-junior amount of knowledge. I don't want to discourage you or say that it's impossible, but getting anywhere in this industry seems to be an ever-increasing uphill slope for those who haven't made it already. And existing success is no guarantee of future success anymore. I'm out.
For example, certain high-quality software; certain task scheduling, etc.
Maybe this is true if you own a business and can pocket the productivity difference. For many workers, however, higher productivity is rewarded with more work. Most people aren't paid per unit of work, but by time, so they expect you to fill your 8 hours with work, and all the newfound efficiency goes to someone else.
If you can show you are more mature than your peers, you will get more opportunities.
Signals of maturity:
1. Be a sound person, don’t make promises you cannot keep.
2. Learn how to negotiate and say no properly.
3. Learn how to grit your teeth and finish something you have zero interest in but have to do anyway.
4. Learn how to present. Learn how to lie about something without being accused of lying.
5. Figure out your own ways and do not believe anything you hear online blindly.
People skills never become obsolete and are useful outside of IT.
Business runs on relationships.
Network.
Good luck.
There's reason to have hope for the future, but I am very confident that we do not know what the world will be like in 5 years, much less our industry in particular. There are people who insist that AI isn't a big deal, and that your career will look like theirs with some minor changes on the surface. I encourage you to listen to scientists instead, which are screaming from the rooftops that artificial minds change everything. Everything.
This is a pretty powerful essay, famous long before it ended up being spookily accurate, nailing the arrival of the singularity to within a ~year (30 years from 1993); it might help you understand why I say that there are no reliable predictions to be had: https://frc.ri.cmu.edu/~hpm/book98/com.ch1/vinge.singularity...
Best of luck, and welcome to the field. I'd enjoy the Greek sun a bit more, but I also wouldn't quit school; no matter what the world ends up looking like, if learning to think critically about hard problems doesn't end up mattering, nothing will. And to say the least, computers aren't going to go away!