I was with a group of faculty members as they watched a live demonstration of a new built-in AI tool in the learning management system. With a single line of command, the vendor representative generated an entire course module in seconds: topic descriptions, learning goals, readings, PowerPoint slides, practice activities, quizzes, and exams. The quizzes and exams could even be graded automatically.
Everything many of us had spent years learning to design appeared instantly.
When the presenter asked whether the faculty would be interested in adopting the tool, the Zoom room fell silent.
Finally, one professor spoke:
“If we ask students not to use AI to cheat, maybe we shouldn’t either.”
A Walk Down Instructional Memory Lane
As I listened, I found myself revisiting nearly thirty years in instructional technology, a field rich with theories and models for evaluating the value of AI versus human teaching.
If I were to hashtag the ideas running through my mind, they would be:
- #CAI
- #Training_vs_Teaching
- #AI_vs_Human_for_Teaching
All leading to:
#GoodTeaching_with_AI_and_Human
Computer-Assisted Instruction (CAI)
From the invention of personal computers to the rise of online learning, technology has traditionally played an assistive role. In the 1950s, B. F. Skinner introduced his “teaching machine,” pioneering programmed instruction. Computer-Assisted Instruction (CAI) was designed to support human instruction, not replace it.
That was true, until the arrival of AI.
Within just a couple of years of ChatGPT’s emergence, studies have suggested that AI-powered tutoring systems can significantly accelerate learning in certain contexts. A 2025 comprehensive review of Intelligent Tutoring Systems (ITS), for example, found that AI can offer scalable and cost-effective personalization, something difficult for humans to replicate in large classes.
Yet what AI primarily delivers is instruction or training. And training is not the same as teaching.
Training vs. Teaching
The difference between training and teaching lies in their purpose, scope, methods, and outcomes.
Training is a systematic process aimed at developing specific skills for particular tasks. As Edwin B. Flippo (1984) defined it, training increases knowledge and skills for doing a particular job.
Teaching, by contrast, is broader. It cultivates understanding, critical thinking, and intellectual growth. According to Anita E. Woolfolk (2016), teaching guides learners to construct meaning through structured experiences.
Years ago, I attended a workshop on cognitive load theory. The presenter asked: What is the best way to demonstrate an Excel feature? Participants suggested help documents, graphics, short videos, and Q&A sheets. Hands went up in agreement.
When I suggested beginning with a real-life situation that required that Excel function, no hands were raised.
The silent reaction seemed to say: Are you out of your mind?
I later realized I was the only participant from a school setting. The others worked in corporate training. Their performance goals centered on efficiency and task completion. They were there to learn how to build user skills quickly and with minimal “unnecessary” cognitive load.
My goal was different: to help students use quantitative reasoning to navigate life.
On our way out, my friend joked, “Sharon, you’re so daring. If I gave your answer to my boss, I’d be fired.”
Today, I wonder whether the roles have been reversed. When it comes to designing and delivering training, AI outperforms humans in speed, scalability, and availability.
But raising a human being, intellectually and emotionally, is not the same as training task efficiency.
That is teaching.
Human vs. AI for Teaching
Research consistently shows that humans continue to outperform AI in areas such as:
- Emotional intelligence and empathy
- Mentorship and relational trust
- Managing behavioral dynamics
- Holistic understanding of student well-being
To teach well in the age of AI, we must understand what distinguishes us.
Human vs. AI in Teaching
Relatable vs. Irrelative
Humans share a biological journey of growth and limitation. AI is an invented, non-biological system.
Empathic vs. Phlegmatic
Humans feel and share emotions. AI simulates response without lived experience.
Active vs. Passive
Humans can act or adjust actively by reading the emotions, context, and the environment. AI operates as an answering machine to prompts.
Vulnerable vs. Powerful
Humans are exposed to risk and limitation, yet vulnerability enables connection. AI strives for comprehensiveness and error minimization.
Rigorous vs. Sycophantic
Humans can stand firm in truth and expectation. AI is often overly accommodating, optimized to please.
Humans possess embodied presence. We sense tone shifts, hesitation, and confidence. We go “off script” when the moment calls for it. We adjust not because of an algorithm, but because we care.
Even our weaknesses, our vulnerability to error, risk, and uncertainty, allow us to build authentic relationships.
That is not programmable.
Good Teaching with Humans and AI
Recently, a colleague responded to an update I shared about my son’s internship application. He had not passed the screening test. As a first-time taker, he mismanaged his time and became stuck on minor bugs.
Her email did not merely offer advice. It offered perspective, empathy, and encouragement. She shared stories. She normalized the struggle. She conveyed warmth and sincerity.
With her permission, I am sharing the message with notes that I’ve added to each sentence, from which I could build a list of what constitutes the meaning of teaching, like a human.
From the message alone, I could construct a list of what teaching, like a human, looks like:
These are deeply human acts.
AI can generate feedback.
But it cannot care.
Living with the Reality of AI
In an article called Innovating the Future: DePaul’s approach to artificial intelligence, DePaul political science professor Dick Farkas pointed out that there’s no uninventing AI. It is like nuclear weapons, you cannot just say, “let’s just push it aside.” It is a reality.
Within that reality lie both threat and opportunity.
It is a threat to those who teach as if they were content-delivery machines. If teaching becomes indistinguishable from algorithmic transmission, replacement becomes plausible.
But for those willing to let AI handle knowledge transmission and skill automation, AI can become an enabler, freeing educators to focus on what matters most:
- Human growth
- Human care
- Human development
The Algorithm of Good Teaching
The future of good teaching is not AI or human.
It is AI for efficiency, and humans for meaning.
AI for training, and humans for transformation.
AI for speed, and humans for depth.
Because in the end, education is not merely about producing correct answers.
It is about shaping lives.
And that remains, irreducibly, human.
