19 May 2026 - Douglas Thain
(Remarks to the Computer Science and Engineering Class of 2026 at the University of Notre Dame)
Good afternoon graduates, distinguished faculty, family members, and guests.
Thank you for inviting me to speak on this very happy Ascension Sunday and graduation day.
Whenever I go on a long trip, and then return back home, it always feels like I just departed. Stepping back into the South Bend airport, I always think “Wasn’t I just getting on the plane here 5 minutes ago?”
I’m sure that it must feel the same way to you at graduation.
You arrived here in August of 2022. We were still wary of COVID, but things were mostly back to normal then. You probably moved into your dormitory on a sticky hot day. You met your roommate, and I’m sure that he or she had some strange habits, kept unusual hours, and had poor taste in music. (Or possibly, it was you who had some strange habits, unusual hours, and questionable taste in music.)
That first year, you probably struggled a bit – finding the right building on the first day of class, approaching people that might become friends, showing up to a new club. Maybe you did even some homework now and then. Perhaps you encountered some mysterious segmentation faults with Prof. Bualuan.
But over these four years, I hope that you found your people, that you learned some new skills, tried some new foods, travelled somewhere new, and finally learned how to merge a conflicted branch in git from Professor Bui. You probably got some grades that you didn’t like. (Maybe from me.) But you rose to each of those daily challenges, both small and large.
And now here we are, everyone a little more confident, a little more relaxed…. (very well dressed!) … about to start a new chapter in life. What’s next?
Well, it is no secret that we live in interesting times. There are a number of challenges facing our society: political polarization, economic inequality, global conflicts, … artificial intelligence.
Ok, do we need to stop and talk about AI?
I think we do, if only because your parents are going to ask about it at dinner.
AI can refer to a lot of computing technologies – in fact, we have quite a number of faculty who are experts in image recognition, natural language processing, machine learning, and so forth.
But let me narrow it down to Large Language Models (LLM) which seem to affect so much of what we do today. This is the specific technology that drives chatbots, and code generators, awful customer support agents, and pretty much anything that produces human language. ChatGPT in particular took off just as you began college in late 2022, to be followed by Gemini, Claude, CoPilot, and many others.
When these tools first appeared, many people had the misconception that we would no longer be plagued by, you know… thinking. The machine would be able to do all of our thinking and talking for us. But that’s not quite right.
LLMs seem to be quite excellent at regurgitating and remixing things that are already known. After all, they are trained by the vast quantity of human-authored text found on the Internet. So if you want sample code for a personal website, if you want a summary of Jane Austin, if you want a generic business plan, then LLMs will be happy to do it for you. “Absolutely! I’ll get on that right away.”
Now, this is, in fairness, pretty helpful. If you know nothing about a particular topic , and need a little structure to get started, the LLM will provide it for you. I particularly like using LLMs to get a quick starter example when learning a new technology.
The bad news, of course, is that (about two years ago) employers thought that LLMs would replace all knowledge workers: lawyers, authors, tax attorneys, and programmers.
But, what our profession has experienced is that LLMs create chaos when applied to a task that is beyond trivial or out of the ordinary. A little example program works at first, but if you ask the LLM to add capabilities, it eventually gets turned into a pile of spaghetti that no one can understand or debug. LLMs make test cases “pass” by simply removing them. An LLM can make a plausible circuit board design, but has no knowledge of current and voltage requirements, and so it catches fire due to a circuit overload.
So what do we need to solve these problems? Should we throw more AI at it?
We need Engineers with a capital E!
The computing work of the future will require much more intellectual framing than it did in the past. Engineers are needed to articulate the purpose of a system, to design an overall architecture, to specify the requirements, to design acceptance tests, to ensure the security and safety of the users, and to audit the details of how the system works.
In short, this is what Engineers have always done when building a bridge, designing a powerplant, or creating an electric car. Each of these activities was once carried out by unlicensed hackers, who cut corners and shaved costs until these creations fell down, caught fire, or killed their occupants. Today, computing systems fill roles that are equally important as bridges, powerplants, and cars. We need Engineers to design and to take responsibility. I think that this will bring computing more into the mainstream of professional engineering. That’s a good thing for society. And it’s a good thing for all of you sitting here.
Now, a caution – in the early years of your career, don’t be overly quick to delegate or avoid work that may, at first glance, appear tiresome. Be willing to practice and make mistakes, especially on the small things.
A great chef learns his or her craft by cooking many dishes over and over again, burning some, oversalting others, until excellence becomes second nature. A great musician practices scales and exercises over and over again, making mistake after mistake, until excellence becomes second nature.
In the same way, the craft of computing must be practiced. You should write lots of programs, build lots of systems, crash lots of computers, and debug many segmentation faults. Only by getting your hands dirty will you develop strong skills, good taste, and the solid intuition that will allow you to become an Engineer with a capital E.
Henry Ford once said that most people don’t recognize opportunity knocking… because it shows up wearing overalls, and looks like work.
Sorry, I got a little off track on AI. Where was I?
We live in interesting and challenging times: political polarization, economic inequality, global conflicts, artificial intelligence…
…and the prospects of a 16 team college football playoff.
I do not bring these challenges up to discourage you. But you should know that a certain amount of chaos is in fact the normal state of the world in every age. We live in a broken, imperfect world, fractured by original sin, and compounded by our own sins.
So, every generation has challenges it must face, and yours is not different.
Our predecessors in the previous century had to navigate the influenza epidemic, the great depression, two world wars, the social upheaval of the 1960s, and many others through the years. In every generation, people just like you confronted these challenges with courage, persistence, and hope.
And so your generation will have to address its own challenges. And I have every certainty that you will bring the same virtues to bear: courage, persistence, hope.
What’s more, I think that the Catholic education that you have received here at Notre Dame makes you uniquely prepared for these challenges.
As you enter the working world, I am reminded of the old story of the three bricklayers:
The great fire of London in 1666, which levelled the third largest city in the world. There was a great effort to rebuild the entire city in a few short years. The great architect Christopher Wren led the reconstruction of 50 some churches, including the great Cathedral of St. Paul.
The story goes like this: The architect visited the construction site of St. Paul’s one day, and observed three bricklayers, side by side, at work on a scaffold, building the same section of wall.
The architect asked the first bricklayer: “What are you doing”? The first bricklayer said, “I am bricklayer, stacking up these bricks one by one. This will take me all day long.” He looked grim and exhausted, and did not look up.
The architect asked the second bricklayer: “What are you doing”? The second bricklayer was standing and looking ahead, saying “I am a builder, I am making this wall that stretches from here to there.” He looked more satisfied.
The architect asked the third bricklayer: “What are you doing”? The third bricklayer was doing exactly the same work, but had a gleam in his eye, looked up to the heavens, and said “I am building a cathedral, a great monument to Almighty God.”
Wherever your life and work may take you – be like the third bricklayer – lift up your eyes and look to the heavens – no matter how great or humble, let your work serve the glory of God.
Congratulations on all you have accomplished at Notre Dame. We wish you every joy and success in what comes next. Don’t forget to call home at least once a week. And come back to visit us in South Bend now and again. Thank you very much.
« Prev: Evolving Programming Models for Massively Scalable Scientific Applications