On AI, Leadership, and Learning: Ari Balogh’s Message to the Class of 2026
A Message from 2026 CDS Convocation Speaker Aristotle Balogh
Distinguished faculty, proud families and friends, and most importantly — graduates of the Boston University Faculty of Computing & Data Sciences — congratulations!
Graduates, today is your day.
Whether you’re receiving your bachelor’s, master’s, or Ph.D., you’ve accomplished something outstanding. You chose one of the hardest, most rigorous disciplines in modern academia. You didn't just learn to build models and code. You learned to think — to model complex systems, to reason about data, to build things that actually work. That is a rare and powerful skill set.
Along the way, you’ve wrangled messy data sets, mastered complex algorithms, and survived late nights staring at screens until the model finally converged or the code worked. On top of all that, you’ve navigated a global pandemic and the early stages of an AI revolution—all before most of you have even had your first full-time job.
That is genuinely remarkable.
I know that beneath the excitement, beneath the pride of wearing that cap and gown, there is also a current of anxiety. You are graduating into a moment unlike any that’s come before, into a world that is profoundly different from the one most of us entered, and changing at a rate that’s hard to fathom. You read the same headlines I do. You hear the predictions that generative AI is going to automate software development; that analyses prompts will replace data scientists; that the very skills you’ve spent years developing might be done by machines in seconds.
I'm not going to stand here and tell you those headlines are all completely wrong (though many are aimed at getting clicks). But I am going to tell you that they are profoundly incomplete, and that your future is not only secure, but so much more exciting than the past we are leaving behind.
Today, I want to share three themes, actually challenges, that may help you navigate this new world and may even make your careers more fulfilling and impactful than any generation before you.
Starting with the first theme: It has always been about learning. And it still is.
Let me paint you a picture of what data science and software development looked like not that long ago. You'd spend — no exaggeration — half, maybe 60% of your time on toil. Data munging and cleansing, boilerplate code, debugging cryptic error messages, writing documentation that, realistically, nobody was going to read.
So much of the job was drudgery.
Now here's the good news — and it’s genuinely great news: Most of that toil is disappearing.
AI tools can clean your data, write your analysis code, draft your documentation, do most of your debug, and more.
Data science and software development, at their core, are some of the purest forms of knowledge work. And what remains when you strip away all that toil … is the very best part! You are going to spend more of your time doing genuinely challenging, intellectually stimulating work than any generation of technologists before you, could ever have imagined.
Your job is to be the mind that:
- Asks the right questions.
- Understands the tradeoffs.
- Determines the choices between competing options.
- Designs models and systems that are robust, ethical, and aligned with real-world needs.
Now the direct consequence of this is that you have to truly love what you do, ‘cuz you’re going to be doing a lot of it!
The other part, and here’s where the challenge comes in, you must be deeply curious. When you use AI, and you will constantly, … you must use it inquisitively. Don’t just use it to do the work and move on. Ask yourself: Why did it choose this approach? What are the tradeoffs of this specific algorithm? Where are the blind spots in this data analysis?
Digging deeper, think about how you already use AI tools today:
You might ask an assistant, “Write me the program to clean this dataset.”
But a better question is:
“Show me three different ways to handle missing values here, and explain the tradeoffs.”
You could say, “Generate a model for this problem.”
Or maybe better is:
“Propose a baseline model, a more complex one, and a non-ML approach. Explain when each is appropriate.”
The difference is subtle, but profound.
- The first mindset treats AI like a shortcut.
- The second treats AI like a collaborator in your learning.
Those who will thrive in this new world are not the ones who use AI to quickly get through a task or avoid thinking. The ones who will thrive are the ones who use AI to think deeper, to explore further, to learn faster. You chose this field – presumably – because something about it grabbed you here & here [head & heart]. Something about the elegance of well-designed algorithms, the thrill of patterns emerging from noisy data, the satisfaction of building something that just works.
Hold on to that and feed it. Because in a world where the routine is automated, the love of learning is your most durable competitive advantage.
If you love the craft—if you’re curious, if you ask “why” and not just “how”—you’re going to do work and have impact that my generation could barely imagine.
This brings us to the second theme. Because the mechanical part of your job is getting easier, the human part of your job is about to become much, much harder—and vastly more important.
That social side of work is about to matter more than ever. And you're going to have to work harder at it.
Here's why. As AI handles more of the technical execution, the problems that come your way will be the ones that are too complex, too ambiguous, too important for any one person – from a single discipline and with a single perspective – to solve alone. These are problems that require teams. Diverse teams. Teams made up of people with different backgrounds, different perspectives, different roles, different ways of seeing the world.
And here's an uncomfortable truth: diverse teams are harder. They're harder because you will work with people across different disciplines. People who disagree with you. People who challenge your assumptions – not because they're being difficult, but because they genuinely see the problem from a different angle.
That's not a defect, that is a feature.
So the first part of this challenge is that it only works if you build the skills to make it work. And those skills are deeply, fundamentally human. I'm talking about 3 specific skills:
- First, Build trusted relationships – not just transactional ones. Build relationships where you show up reliably, where you give credit wherever it’s due, where you’re honest when you don’t know something, where you can talk about difficult things.
- Second, Give constructive feedback – not the watered-down, ambiguous kind, but clear, caring, specific feedback that actually helps someone grow, or that explains a possible flaw in their approach, but without tearing them down.
- Third, I'm talking about Dealing with conflict – this one is so hard. Don’t avoid conflict, and do not see it as a contest to win, rather, engage productively, stay in the conversation when it gets uncomfortable, and seek a productive outcome, because the best ideas are usually forged through conflict, not watered down compromise.
Developing these skills is hard, and actually implementing them daily is even harder.
These three skills will be the foundation of your most critical career asset: which is, the trusted network that you build and maintain – the group of people you can lean on, argue with, brainstorm with, and pressure-test ideas against … as you tackle truly hard problems. This trusted network consists of:
Peers you can call near midnight to sanity-check an idea or just be a sounding board.
- It includes Mentors who will tell you what you definitely do not want to hear.
- It also includes Teammates with whom you can argue and strive together – not for sport, but because you both care about getting to the best solution.
Now, can AI help you explore a solution space? Absolutely. It can generate options, surface research, model scenarios. But AI cannot provide the judgment that comes from a passionate, respectful debate between people who trust each other enough to disagree. AI cannot navigate the organizational politics of getting a controversial recommendation adopted. AI cannot sit across the table from a skeptical, senior stakeholder and build the credibility needed to drive a fundamental change.
Only you can.
So invest in those skills. Take them as seriously as you took your algorithms coursework. Build your network not just wide, but deep. And remember: in a world increasingly mediated by technology, the ability to connect authentically with other humans isn't just nice to have – it's your superpower.
The future belongs to those who can combine technical excellence with social intelligence and constantly keep growing both.
This brings us to the final and most important theme: Own your impact.
Many of us fell in love with tech for its own sake: we loved the elegance of a clever algorithm, or the thrill of a converging model. That passion is beautiful, and it matters. But you cannot stop there.
The work you do in data science and computing doesn’t live in a vacuum. It lives:
- In hospitals deciding who gets what treatment.
- In banks determining if someone gets that loan for their first house.
- In social platforms shaping what people see and believe.
- In governments setting policy.
Your models, your code, your systems – carry power, and with it, comes responsibility. The best of us in these fields are not just brilliant technologists, they:
- Deeply understand the business or institution they’re in and how it achieves its mission,
- they Actively engage with the customers or citizens they serve, and
- they Relentlessly focus on outcomes, not just outputs.
Rather than focus on how to do the task at hand, they ask questions like:
- “What problem are we really trying to solve?”
- “How will we know if this is actually working?”
- “Who benefits from this system – and who might be harmed and how?”
My most important challenge for you today is whatever organization you join – whether it's a startup, a Fortune 500 company, a hospital, a nonprofit, a government agency, a research lab – I want you to imagine that you are the CEO. Not in some arrogant or authoritarian way, but in a responsible way. If you were that CEO, what would you care about? What would keep you up at night? What's the highest-impact thing that data science or technology could contribute to drive the mission forward?
Your official title doesn’t matter, you can still ask:
“What is the highest-impact thing I can contribute to drive this mission forward?”
When you think that way, everything changes. You stop waiting to be told what to do, and start seeing where the opportunities are. You stop optimizing for technical elegance and start optimizing for outcomes. You become the person in the room who connects the dots between what's technically possible and what actually matters.
That will cause you to dig deeper, develop valuable opinions, and sometimes it means you’ll say:
- “We’re measuring the wrong thing.”
- Or, “We’re optimizing for efficiency, but ignoring fairness or bias.”
- Or, “This might be great for revenue, but terrible for the user experience.”
And remember, these systems encode values and assumptions—some explicit, many hidden.
Your job is to surface those assumptions, question them and the ethics behind them, and, when necessary, challenge and change them.
You have a unique vantage point:
- You understand what’s technically possible.
- You see where the data comes from and how messy it really is, how sensitive the analysis is to slight input perturbations.
- You know how fragile models can be when the world changes.
That makes your voice incredibly important, since you’ll be in rooms where no one else understands the technology, the data, and the tradeoffs like you do. The question to carry with you is not:
- “How do I create the coolest model or system?”
It is:
- “How do I build the model or system that most effectively advances this mission, for these people, in this context – while minimizing risk of harm or bias?”
That’s what it means to own your impact. And, by the way, that person is irreplaceable, so be that person.
To sum up, graduates, we are standing at the threshold of a new technological revolution, at least as impactful as the Internet was over the last 30 some years. It’s natural to feel a little apprehensive of the unknown. But I want you to leave here today not with any anxiety, but with excitement – because you are extraordinarily well-positioned.
Look around you. Look at the faculty who’ve guided you. Look at the families who’ve supported you. Look at the friends you’ve made here at BU. You are not facing this alone.
You have been trained at one of the finest institutions in the world. You have learned how to learn. So:
- First - Never stop learning. The toil is going away and what’s left is the good stuff, the deep, challenging, creative work. Stay curious, and use AI to learn faster, think deeper, and embrace the intellectual challenge of your work.
- Second - Invest deeply in the people. Build your trusted network. Get good at the hard, human work of collaboration, feedback, and constructive conflict. In a world full of algorithms, authentic human connection is more valuable than ever.
- And third – most importantly: Own your impact. Don't just be a technologist. Understand the mission. Take responsibility for outcomes. Be the person who connects what's possible to what matters. Step up as leaders, owning your impact and driving the mission forward.
Graduates, you are no longer students of the Faculty of Computing and Data Sciences. As of today, you are the architects of the future. And I, for one, cannot wait to see the world you are going to build.