Skip to main content
Aasquare Academy logo AasquareAcademy
← All posts

"Are We Cooked?" What CMU Faculty Say Kids Should Learn in the AI Era

Takeaways from a Carnegie Mellon faculty panel on AI, careers and learning, and how they align with project-based engineering at AA Square Academy.

aasquare.academy 7 min read
  • AI skills
  • Parents
  • Future of work
  • Careers

If you are a parent wondering whether artificial intelligence will render your child’s future skills, career prospects, or computer science education obsolete, you are in good company.

At Carnegie Mellon University (CMU)—one of the world’s leading centers for computer science, robotics, and artificial intelligence—the student-led Carnegie AI Safety Initiative (CASI) hosted a faculty panel titled “Are We Cooked? CMU Faculty on AI, Careers and Learning.”

You can watch the full panel discussion on YouTube.

The panel brought together distinguished academic leaders and industry pioneers:

  • Po-Shen Loh — Professor of Mathematics, entrepreneur, and former coach of the USA International Mathematical Olympiad (IMO) team.
  • Zico Kolter — Professor and Director of CMU’s Machine Learning Department, Co-founder of Gray Swan, and Chair of OpenAI’s Safety and Security Committee.
  • Daphne Ippolito — Assistant Professor in the Language Technologies Institute (LTI).
  • Aditi Raghunathan — Assistant Professor in the Computer Science Department.
  • Vince Conitzer — Professor of Computer Science, Director of FOCAL, and Head of the Core Safe AI Research Center at CMU.

Their conclusions were remarkably reassuring, grounded, and practical. Rather than suggesting that technical education is losing value, the professors outlined a clear evolution in how children must learn and which core human capabilities will truly matter.

This article synthesizes the panel’s key takeaways for parents and demonstrates how these principles align directly with hands-on, project-based engineering at AA Square Academy’s Innovators Tech Immersion (https://aasquare.academy).


1. The Fundamental Shift: From Code Generation to Logical Literacy

For decades, introductory coding education emphasized memorizing syntax, writing boilerplate functions, and manual code assembly. Today, AI coding agents can generate routine code in seconds. However, every CMU panelist agreed on a critical distinction: AI handles routine code generation, but humans must master logical comprehension, system design, and code inspection.

The Calculus Integral Analogy

Professor Zico Kolter offered a compelling analogy for why learning to code remains vital in the age of AI:

“Learning how to code is about learning how to think sequentially… It’s like learning integrals in calculus. You might not manually compute integrals in your future job, but learning calculus trains your mind to structure logical thoughts into explicit instructions.”

Why Debugging is the New Superpower

Professor Daphne Ippolito pointed out that as AI agents write more code, they frequently introduce subtle, hard-to-detect bugs. A student who cannot read and trace code logic manually will be unable to spot or fix errors when an AI model fails.

To thrive, students do not need to act as human syntax compilers. They need to become analytical system architects who can read, test, and debug automated outputs.


2. Four Core Pillars for Future-Proof Technical Education

The CMU faculty highlighted four central themes that define how students should prepare for an AI-driven world. Here is how those academic insights translate into real-world learning:

CMU Faculty InsightWhat the Professors EmphasizedHow It Looks at AA Square Academy
1. Read & Debug LogicAI generates routine code; human value lies in reading code, spotting subtle bugs, and understanding system behavior.Hands-On Physical Debugging: When a robot drifts off a track or misses an obstacle, students trace code logic, form hypotheses, and fix real physical behaviors.
2. Frame Problems Worth SolvingAs execution costs drop, curiosity and problem formulation become the distinctly human competitive edge.Real-World Build Challenges: Students begin with genuine human needs—designing, building, and testing real mechanical and software solutions rather than solving static worksheets.
3. Demystify the MachineTo avoid treating AI as magic, students should understand how models are built and trained from first principles.First-Principles AI Foundations: Students collect data, train, and test small machine learning models themselves, transforming AI from a black box into a usable tool.
4. Build Networks of TrustAI makes output volume easy to game. Evaluation is shifting toward trusted mentor recommendations and peer verification.Cohort & Mentor Verification: Students work alongside experienced engineering mentors in consistent peer cohorts who can truthfully verify their genuine growth and character.

3. Deep-Dive: Exploring the Four Pillars

Pillar 1: Reading, Inspecting, and Debugging Code

  • The Academic Insight: AI agents can draft code rapidly, but they lack real-world physical context.
  • The Action for Students: Learning to program remains essential because it builds sequential reasoning. At AA Square Academy, students progress from foundational logic to text-based code within physical systems (such as microcontrollers, sensors, and robotics). A physical robot provides immediate, objective feedback: if there is a subtle logical flaw in the code, the robot stops. This builds deep, resilient debugging habits.

Pillar 2: Finding Problems Worth Solving

  • The Academic Insight: Professor Po-Shen Loh shared how he conducts field research in communities ranging from West Africa to rural Georgia classrooms (where he teaches 6th graders as “Mr. Poe”) to discover what real people need. He noted: “If you can figure out what problem you’re actually trying to solve, the AI is a powerful tool to help you build it.”
  • The Action for Students: Instead of giving children step-by-step assembly guides, AA Square Academy’s Innovators Tech Immersion uses open-ended design briefs. Students learn to ask: Who is this for? What real constraint matters? How do we iterate when our first build fails?

Pillar 3: Demystifying How AI Works

  • The Academic Insight: Professor Kolter advised that every student interested in technology should write a small Large Language Model from scratch in Python to demystify how intelligent-looking outputs are captured algorithmically.
  • The Action for Students: When children only consume AI through chat prompts, AI feels like magic. In our AI curriculum module, students learn dataset collection, feature engineering, model training, and evaluation. By training and testing small models themselves, they develop healthy technical intuition and realistic expectations of technology.

Pillar 4: Cultivating Mentorship and Networks of Trust

  • The Academic Insight: Professor Loh spoke about the importance of a student’s “trust network” (what he playfully calls a “couch number”—the number of trusted peers globally who would welcome you into their homes). As AI inflates superficial metrics (like raw paper counts or generated code volume), admissions and industry evaluations are returning to trust-based systems: direct recommendations from mentors and peer cohorts who know your real work ethic.
  • The Action for Students: In the Innovators Tech Immersion, students work in consistent year-round cohorts guided by industry-experienced engineering mentors. Instructors observe how students handle frustration, collaborate with teammates, and solve complex problems under pressure—building authentic relationships that lead to meaningful recommendations.

4. Preserving Human Thought: Avoiding “AI Slop”

A passionate point raised by Professors Daphne Ippolito and Zico Kolter was the danger of generating unvetted “AI slop”—using AI to write essays, reports, or communications without personal thought.

“Writing is the process of codifying thought in your own mind. If you hand that off to an AI, you lose the thought process itself… Sending AI-generated text that you barely proofread is disrespectful of the reader’s time.” — Prof. Zico Kolter

At AA Square Academy, we embrace AI as an accelerator for research and testing, but we enforce strict standards on student expression:

  1. Oral Project Demos: Students present their engineering builds orally during class showcases, explaining why they chose specific mechanical or software trade-offs.
  2. Engineering Notebooks: Students document their own engineering logs, ensuring their critical thinking and personal voice remain at the core of their work.

5. Why Human Skills Matter More Than Ever: The Chess Lesson

When supercomputers like Deep Blue defeated world chess grandmasters, critics predicted that human interest in chess would plummet. Instead, as Professors Vince Conitzer and Aditi Raghunathan pointed out, chess is more popular today than at any point in human history.

Why? Because humans are fundamentally interested in human effort, human creativity, and human problem-solving. AI tools raise the baseline of what young minds can build, enabling them to tackle advanced concepts earlier—provided they build strong foundational habits.


6. How AA Square Academy’s Innovators Tech Immersion Fits Your Child’s Journey

AA Square Academy’s Innovators Tech Immersion (https://aasquare.academy) is specifically engineered for students in Grades 4 through 12 to build long-term, compounding technical capability in the AI era.

Core Program Highlights:

  • Year-Round Consistency: Real engineering skills compound over time. Weekly immersive sessions ensure continuous growth without summer learning loss.
  • Six Engineering Pillars: Seamlessly integrates Computer Science, Artificial Intelligence, Robotics, Software Engineering, Electrical Engineering, and Mechanical Engineering.
  • Projects Over Lectures: Over 80% of class time is spent building, coding, testing, and presenting physical hardware and software projects.
  • Prestige Competition Tracks: Structured preparation for national and international STEM competitions, including USACO (USA Computing Olympiad), ACSL (American Computer Science League), and MIT Zero Robotics.
  • Expert Mentorship: Learn directly from practicing engineers and computer scientists who provide genuine guidance, rigorous feedback, and verified recommendations.

Conclusion: Preparing Your Child to Lead in the AI Era

The overarching message from Carnegie Mellon’s faculty is clear: Do not attempt to compete with AI at what AI does best—speed and raw generation. Instead, empower your child to master what humans do best—curiosity, critical logic, system debugging, and meaningful human connection.

To explore how AA Square Academy can help your child navigate and lead in the AI era, visit https://aasquare.academy and discover our Innovators Tech Immersion programs today.


Source: Full panel recording on YouTube.

Disclaimer: This article summarizes public panel discussions hosted by the Carnegie AI Safety Initiative at Carnegie Mellon University for educational reflection. Carnegie Mellon University and its faculty are not affiliated with and do not endorse AA Square Academy.

Admission is selective. Effort is the only prerequisite.

We admit a small number of students each season — motivated beats experienced. Tell us about your student and we’ll set up an intro conversation.

Apply for admission