i was just thinking today about how lucky of a boat that the class of '26 (my year) has been in coming into freshman year (especially those who knew they wanted to do computer science). the reason i say this is because of generative ai for all the skills it has been able to help with outside of the classroom (as well as some inside of class). if you think back to how people perceived getting a degree in computer science a few years ago compared to now, there's a huge difference. and maybe this is coming from a slightly more privileged background since i had learned bits of cs here and there during high school, but i still think this applies.
it was not an easy feat getting a cs degree a couple of years ago, and not saying it's easy now, but i would say it's a lot easier. with the lack of online resources (and previously ai), i think it's fair to say it was a lot more challenging. it also used to be a lot less competitive because people were just starting to pick up on how good of a career path it is (to make money at least) and so the job market was a lot better than it is now. this big shift is because of oversaturation and ai.
i think around the time that i entered freshman year it had started getting really saturated. within 3 months of starting college, the first wave of ai (when chatgpt was announced) had begun. the biggest advantage was having the opportunity to use ai when building projects and i think that is why my year of students have had it the luckiest. being given such a powerful tool when our brains are the most malleable has been both really useful and also dangerous. i would not be nearly as technically capable as i am now if i hadn't started using ai immediately. college doesn't prepare you well for the real world, but the things i've learned from ai definitely have. however, we can see a lot of the dangerous effects of it more recently.
this might be kind of mean to say, but students entering computer science since ai (my year included) have been getting dumber every year because they are being too overly reliant on ai. one of my friends (adi) recently attended code for good and mentioned how any team member he talked to always redirected him to using ai. i think that's a big shift alone. people don't value their own input anymore and have basically become "npcs". situations of teamwork like those won't serve anyone any good from copying and pasting something from ai. answering other people's questions help both parties out and ai should be a mid-last resort. another example is this beginner introduction to python course (engr 102) at texas a&m that every engineer has to take. and because everyone is using ai on all of their assignments, the level at which students actually know how to code is rapidly declining (people can't even write a for loop correctly). these are just a couple things, but i'm sure there's a lot more like this.
the current cs advantage is that sure we have ai, but it needs to be used in an advantageous way where you are still learning/motivated and you can stand out from the oversaturated market. unless something drastically changes, i personally think that my year has had it the best and it's only downhill from here.
- aayush