My son wants to study computer science.
I build software with AI every day.
You would think that would make me feel confident.
It does not.
Gabe is a rising senior. He is interested in computer science and computer engineering. He likes building things. He likes understanding how things work. He is good at it.
And I am sitting here at the kitchen table trying to figure out whether the thing he wants to study is going to look the same by the time he finishes studying it.
I am not some random parent worrying about AI from the outside.
I run CollegeHound. I use AI to build our product. I watch it write code, generate content, answer questions, summarize information, and solve problems that would have taken me hours to research.
Ed sees this every day too. He is my husband, Gabe's dad, and CollegeHound's CTO, and he uses AI tools constantly.
These tools are not a gimmick.
They are genuinely good.
And that is exactly why this feels complicated.
Because the two adults in Gabe's house are using AI every day to build a company. We see what it can do. We know how powerful it is.
And Gabe?
Gabe is pretty resistant to AI.
He does not want it to do the thinking for him. He is not excited by the idea of using AI as a shortcut. He is interested in computer science and computer engineering because he likes building things and understanding how they work.
That matters to me.
Because while Ed and I are using AI every day, Gabe is looking at the same thing from a very different angle.
He is not asking, "How can AI do this for me?"
He is asking, "If AI can do this, what am I actually supposed to learn?"
And honestly, that is a fair question.
Why This Feels Complicated
AI is changing computer science. It is not replacing the need to understand it.
The students who will thrive are the ones who learn to think, not just code. College, at its best, is still where that kind of thinking can happen. But families should go in with clear eyes, honest conversations, and a plan.
Not panic. Not blind faith. A plan.
The Strange Part in Our House
The strange part is that Gabe's skepticism may actually be one of the reasons I think he could be okay.
He is not dazzled by AI.
He is not trying to avoid learning.
He is not looking for the fastest shortcut to a tech salary.
He wants to understand how things work.
That is the part I keep coming back to.
Because if AI is going to make anything less valuable, it is probably surface-level skill. The ability to produce something that looks right. The ability to follow a tutorial. The ability to generate code without understanding what it does.
But deep understanding?
Systems thinking?
Problem-solving?
Knowing when something is wrong even though it looks right?
That feels more important, not less.
What I Know and What I Do Not
Here is what I know.
AI can write functional code. It can debug. It can generate a working app in minutes that would have taken a junior developer days. I have watched it do this in real time while building CollegeHound.
Ed uses these tools constantly. So do I.
They are not toys. They are not just fancy autocomplete. They are changing the way software gets built.
Here is what I also know.
When something goes wrong, when the code breaks in a way AI did not anticipate, when a design decision needs to account for security or scale or the messy reality of real users, you need a person who understands what is actually happening.
Not someone who can prompt an AI.
Someone who understands the system.
That distinction matters.
And I think about it every time Gabe and I talk about college.
Because I do not want him to spend four years learning something that will be outdated before he graduates.
But I also do not want him to skip the hard learning because a tool can produce an answer.
Those are not the same thing.
What Professors Are Worried About
Jay Caspian Kang has been writing in The New Yorker about what AI may do to higher education. His pieces have stayed with me because they are not just asking whether students will cheat.
They are asking a bigger question.
If AI can help students get through the work, how do we know they are actually learning?
That question matters a lot when your kid wants to study computer science.
Because the goal cannot just be finishing assignments. The goal has to be understanding.
Can the student reason through a problem?
Can they tell when an AI-generated answer is wrong?
Can they understand the tradeoffs behind a design choice?
Can they build something that works in the messy real world, not just in a demo?
That is where I keep landing.
AI may make some assignments easier. It may make some beginner coding tasks faster. It may change what students need to memorize.
But it does not remove the need for judgment.
And honestly, judgment may be the whole point now.
That is what I want Gabe to learn.
Not just how to get the answer.
How to know whether the answer makes sense.
The Irony of How I Use AI
There is another irony here.
I have spent nearly 30 years as a speech-language pathologist helping other people find ways to communicate. I have helped people use augmentative communication devices, visual supports, scripts, strategies, workarounds, and tools to get their thoughts out when spoken or written language did not come easily.
And now I am the one using a tool.
I have aphasia. My thoughts are there, but language does not always come out as cleanly as it feels in my head. AI helps me shape the language of my work. It helps me organize what I am trying to say, find the words, and turn the ideas in my head into something other people can read.
That does not mean AI is doing the thinking for me.
The thinking is mine. The experience is mine. The questions are mine. The judgment is mine.
But AI helps me access the language.
As an SLP, I would never tell a person that using a communication device made their thoughts less valid. I would never tell someone that a workaround meant they were cheating.
The whole point was access.
So I have to hold both truths at the same time.
AI can be an incredible accessibility tool. It can help people communicate, build, organize, and participate in ways that might otherwise be harder.
And AI can also make it harder to know whether a student is learning, thinking, or just submitting polished output.
That is why this conversation is complicated for me.
I use AI. I value it. I believe in tools that support access.
I also believe students still need to learn how to think.
Get Organized Before Senior Year
What I Tell My Son
Gabe and I have talked about this.
More than once.
Probably more than he wants to.
Here is what I tell him.
AI is not going to replace people who understand systems.
People who can think through a problem from first principles.
People who know what questions to ask before they start building.
People who can look at a piece of AI-generated code and know whether it is actually good or just looks good.
AI may replace some of the easier, entry-level execution work. It may change what junior developers do. It may make certain tasks faster, cheaper, or less valuable.
But that is not the same as replacing understanding.
The difference is understanding versus output.
And college, at its best, teaches understanding.
The question I cannot answer for Gabe is whether every college still does that consistently.
Will the programs he is looking at hold the line on rigor while everything around them gets easier?
Will his professors still assign the hard problems?
Will he still have to struggle enough to learn?
Will he be surrounded by students who want to understand, or students who want to get through?
I think some programs will get this right.
I think the best programs will adapt by going deeper, not shallower. They will teach students to use AI as a tool while still requiring them to understand the fundamentals underneath it.
The programs that do that will produce graduates who are genuinely valuable.
The programs that do not may produce graduates who are hard to distinguish from someone with a three-month bootcamp and a ChatGPT subscription.
That is a hard thing to evaluate from a campus tour.
But it may be one of the most important things to ask about.
The Parent Paradox
I am aware of the irony here.
I am building an AI tool for college planning while questioning what college will look like for my own kid.
Ed and I use AI every single day to run and build CollegeHound. We believe in it. We see what it makes possible.
And we are also parenting a son who is skeptical of AI and interested in studying the very field AI is changing.
So no, this is not theoretical in our house.
It is dinner-table conversation.
It is car conversation.
It is campus-visit conversation.
I am not anti-college.
I am not anti-AI.
And Gabe is not anti-technology.
We are just a family trying to make a good decision with incomplete information.
I think that is every parent reading this right now.
We all want certainty.
We want someone to tell us the right major, the right school, the right plan. We want to know that the investment will pay off. We want guarantees.
There are no guarantees.
There never were.
But right now, the uncertainty feels louder than usual.
AI is reshaping career paths across every field, not just tech. And the families who are going to navigate this well are not the ones who pretend nothing is changing.
They are the ones who talk about it honestly.
At the kitchen table.
In the car.
On the campus tour.
Without panic.
Without false confidence.
Just honestly.
The Part of College AI Cannot Predict
Here is the part that keeps pulling me back from the ledge.
College is not only job training.
At its best, college gives students access to people, ideas, classes, projects, and conversations they would not have found on their own.
A student walks into a class because it fits the schedule, or because a professor sounds interesting, or because a friend recommended it, and something shifts.
They read something they did not expect to care about.
They meet a professor who changes how they think.
They work on a project that opens a new door.
They realize they are good at something they had never considered.
You cannot optimize for that.
You can only put yourself in places where it might happen.
That is one of the reasons I am not ready to write off college, even in the age of AI.
A bootcamp can teach a skill.
A tool can generate an answer.
But a strong college experience can still change the way a student thinks.
What Families Should Actually Focus On
If your kid wants to study computer science, computer engineering, or anything tech-related, here is what I think matters right now.
Look at the program, not just the name.
Computer science at one school is not the same as computer science at another. Some programs emphasize theory and systems thinking. Others emphasize applied skills. Some are adapting their curriculum for an AI world. Others are pretending nothing has changed.
Ask questions.
Look at course requirements. Talk to current students if you can. Ask what students are building. Ask how AI is being handled in classes. Ask what students are still expected to do without AI.
Pay attention to what is still hard.
If a program has eliminated the challenging coursework because AI makes it inconvenient, that is a red flag.
The hard parts are where the learning happens.
A program that still expects students to wrestle with algorithms, data structures, architecture, systems design, ethics, security, and real project work is a program that is still teaching something AI cannot simply hand over.
Talk to your student about what they actually like.
There is a difference between liking technology and liking problem-solving.
Both are valid.
But the students who will thrive in a changing CS landscape are probably the ones who genuinely enjoy thinking through hard problems. Not just the ones who want a tech salary.
If your student is drawn to a different path entirely, that is worth exploring too.
Do not panic.
The sky is not falling.
Computer science is changing, but demand for people who deeply understand technology is not disappearing. The Bureau of Labor Statistics still projects strong growth for software developers over the next decade.
What is changing is the bar.
The floor is rising.
The people who will be fine are the ones who were always going to learn more than the minimum.
Have the conversation.
Your student is going to hear about AI whether you talk about it or not.
Better that they hear your honest thinking than some TikTok hot take.
Students are hearing messages that college is not worth it. They need adults who are willing to sit in the uncertainty with them and think it through together.
I do not have the answer figured out for Gabe.
I have a direction.
I have questions I want him to ask on campus visits.
I have a list of things I want to know about every program he applies to.
And I have the conviction that understanding how things work will always matter, even when the tools for building them change.
That is not certainty.
But it is enough to keep moving forward.
And honestly, that is all any of us have right now.
How CollegeHound Helps
If your family is working through these same questions, you do not have to do it alone.
Scout, our AI planning assistant, can help your student think through what they want to study, compare programs, and build a college list that actually fits.
It will not pretend to have all the answers.
Neither will I.
But it can help you ask better questions.
And right now, better questions are worth more than false confidence.
Get Organized Before Senior Year