Caleb I just tried your project. Thanks for sharing it.Your experience developing this project is in line with this NYT article I just read this morning
"Coding after Coders" https://www.nytimes.com/2026/03/12/magazine/ai-coding-programming-jobs-claude-chatgpt.htmlI think this is a KEY article that anyone interested in AI. (not just AI and CS) might want to read)The last paragraph really speaks to why we all need to be paying attention and why we need AI literacy and also how AI will make it necessary for us to take another look at the standards we prioritize and the learning progressions we have developed in a world BEFORE AI.
""How things will shake out for professional coders themselves isn’t yet clear. But their mix of exhilaration and anxiety may be a preview for workers in other fields. Anywhere a job involves language and information, this new combination of skills — part rhetoric, part systems thinking, part skepticism about a bot’s output — may become the fabric of white-collar work. Skills that seemed the most technical and forbidding can turn out to be the ones most easily automated. Social and imaginative ones come to the fore. We will produce fewer first drafts and do more judging, while perhaps feeling uneasy about how well we can still judge. Abstraction may be coming for us all."Because I knew it was a long article to share, I also asked AI to summarize it below-- although I don't think the summary came close to identifying the many thoughts I had while reading the article"from ChatGPT:Here is a clear, structured summary of the New York Times Magazine article “Coding After Coders: The End of Computer Programming as We Know It” by Clive Thompson.
Summary: “Coding After Coders” (NYT Magazine, March 2026)
1. The Core Argument
The article argues that AI is fundamentally transforming software development, shifting programming from writing code manually to directing AI systems that generate code.
Instead of typing thousands of lines of code themselves, developers increasingly:
Describe what they want in natural language
Let AI tools generate the code
Review, test, and refine the results
This shift is sometimes called “AI-assisted programming” or “vibe coding.”
The article suggests we are moving from “coding” to “orchestrating AI that codes.”
2. New AI Tools Are Shockingly Capable
The piece highlights rapid improvements in AI coding systems such as:
ChatGPT
Claude
GitHub Copilot
Codex
These tools can now:
Generate entire programs
Debug complex systems
Read large codebases
Write tests and documentation
Suggest architectural changes
Developers interviewed for the article say AI can produce huge amounts of working code quickly, sometimes completing projects that once took weeks in hours.
3. Programming Is Becoming a Different Job
The article stresses that the role of programmers is evolving rather than disappearing.
Future developers will spend more time:
Less time on
Writing syntax
Routine coding
Boilerplate programming
More time on
Designing systems
Defining problems
Evaluating AI output
Debugging and validating code
Thinking about architecture and user needs
In other words, programmers increasingly act as editors, architects, and supervisors of AI systems.
4. Productivity Is Exploding
AI coding tools can dramatically increase productivity.
Developers report:
Building prototypes in hours
Completing “side projects” that once cost hundreds of thousands of dollars
Producing far more code with smaller teams
This may lead to a huge expansion in software creation, because barriers to building software are dropping.
5. But There Are Risks
The article highlights several concerns:
Reliability
AI can still:
hallucinate
produce insecure code
introduce subtle bugs
Developers must verify everything.
Skill atrophy
Some worry that younger engineers may rely on AI without learning fundamentals.
Job structure changes
Entry-level programming jobs may decline because AI performs many basic coding tasks.
6. A Historical Perspective
The article frames AI coding as another step in the long evolution of programming tools:
Era How software was created 1950s Punch cards and machine code 1970s–90s High-level programming languages 2000s Frameworks and libraries 2010s Open-source reuse 2020s AI generating code Each step raised the abstraction level, letting humans focus on higher-level thinking.
AI may simply be the next layer.
7. The Big Question: Who Builds Software Now?
One major implication discussed in the article:
Many more people may become software creators.
If AI handles the coding, then:
designers
scientists
entrepreneurs
teachers
students
may be able to build software directly by describing what they want.
Programming becomes closer to problem-solving and system design, not just syntax.
8. The Paradox: More AI Could Mean More Programmers
Some developers argue that AI could actually increase demand for software developers.
Reason:
When software becomes cheaper to create, people build more of it.This phenomenon is sometimes described as a Jevons-style productivity paradox: greater efficiency increases total demand.
One-Sentence Takeaway
AI isn’t ending programming—it’s transforming programmers into designers and supervisors of AI systems that write code.
Wow...Claude helped me make my "Learning by Teaching" DumBot, now called https://www.nonothin.com/, open to teachers and students with no account, login, tracking, or cost!!Use at will! Feedback desired.A very rough draft of the teacher guide is below.NoNothin.com - Teacher & Adult Guide
ROUGH DRAFT By Caleb Clark | March 13, 2026
“If you can’t explain it to a six-year-old, you don’t understand it yourself.”
— Attributed to Albert Einstein
About
NoNothinBot (nonothin.com) is a free “Learning by Teaching” web app. Students teach it—instead of being tested. It asks questions, draws simple pictures of what it’s being taught, and generates a report when the lesson is done. Then it forgets everything. There is no login, no tracking, no cost.
Just go to nonothin.com and start teaching it something, or just type “Hi.”
QuickStart Lesson Plan
Go to nonothin.com. It is built to be used anonymously and be easily accessible, so there’s no login, no accounts needed, and no tracking or cost.
Try it yourself. Just start teaching it something—anything. Or just type “Hi” and it will help you learn to use it! It will ask questions, try to draw what is being explained to it, and then create a temporary anonymous report for copying before forgetting everything.
Challenge your student/child. Ask them to teach NoNothinBot a topic you want to see how well they know.
Copy the report. Make sure they share the report NoNothinBot creates when they are done. See Assessment below.
Wipe memory. Have them click “New Session” and NoNothinBot forgets everything, including the report—which you cannot get back—and starts fresh.
Reflect together on the report.
Repeat with new topics. Offer as an alternative assessment.
Example Starter Prompts
Confused or impatient students can just type “Hi” and NoNothinBot will help them get started and clarify that they are the teacher.
Younger students can just type statements of what they want to teach, like “2 + 2 = 4.”
Older students can use full sentences like:
“I am going to teach you about the water cycle on Earth.”
“I am going to teach you what a sentence is.”
“I am going to teach you what the U.S. Constitution is.”
“I am going to teach you how to divide 360 by 8.”
Intended Ages and Focus:
Age 13 to end of high school: Assess knowledge, deepen learning, identify gaps, concept checks, explanation practice, discussion starters.
Post high school: Class preparation, tutorial-style learning, and metacognitive practice.
Out of school: Useful for self-directed learners who want to test the depth of their understanding beyond surface-level familiarity—teaching a concept in plain language quickly reveals what you actually know vs. what you’ve memorized as jargon. And it’s fun!
Troubleshooting & Facilitation Moves
If it asks too many questions: Say, “Move on” or “Ask fewer questions so I can finish the core idea.”
If it seems wrong: Treat it as a prompt for discussion; ask students to identify the issue and correct it.
If behavior is odd: Ask, “Explain what you are doing and why,” then redirect with a clearer request.
Report issues to: [log in to unmask]
Learning Theory
NoNothinBot acts like it is learning from the student during conversations (which it forgets after each session), thereby revealing student knowledge and deepening learning. It challenges students to teach it to see what the student knows, and thereby shows the student and their teachers/parents what they know through text and very simple line drawings that mirror what it’s being taught.
NoNothinBot builds on these established educational ideas:
Learning by teaching, an approach in which students consolidate knowledge by explaining, answering follow-up questions, and repairing gaps in their reasoning.
The protégé effect—greater effort and deeper learning when a student expects to teach.
The Socratic tradition, where progress is driven by probing questions and clarifications that help learners articulate assumptions, define terms, and connect steps in an argument.
Visual learning through simple drawings. During each session, NoNothinBot draws simple SVG line drawings of what it’s being taught. These aren’t decorative—they’re a feedback mechanism. If the bot’s picture doesn’t match what the student meant, that gap becomes a teaching moment. This approach draws on dual coding theory: learners who process information through both words and images build stronger, more retrievable memories than through words alone (Clark & Paivio, 1991; Mayer, 2009).
The Feynman Technique (see below).
The Feynman Technique
Richard Feynman (1918–1988) was a Nobel Prize–winning physicist and one of the most celebrated science teachers of the 20th century. He was famous for his ability to explain deeply complex ideas in plain, vivid language—and for insisting that if you couldn’t explain something simply, you didn’t truly understand it. His approach to learning has been distilled into four steps:
Choose a Concept and Study It: Write the topic on a blank page and write down everything you know about it.
Explain It to a Child: Explain the concept in simple, non-technical language as if teaching it to a 12-year-old.
Identify Gaps and Go Back to Source Material: Review your explanation to find areas where you struggled or relied on jargon. Return to your notes or textbook to fill those gaps.
Simplify and Review: Organize, refine, and simplify your notes into a clear narrative to ensure complete understanding.
How NoNothinBot Uses Drawings
NoNothinBot draws simple SVG pictures throughout each session to show what it thinks it’s learning. These drawings serve several purposes:
Instant visual feedback. When a student teaches the bot about the water cycle, it might draw arrows going from a cloud down to the ground. If the student hasn’t explained evaporation yet, the picture will be incomplete—and the student can see exactly what’s missing.
A mirror, not a test. The drawings reflect the student’s teaching, not the bot’s knowledge. If the drawing is wrong, that’s a signal the explanation needs work—a low-stakes way to notice gaps.
Final session drawing. At the end of a session, the bot creates a slightly larger summary drawing that appears in the report. This gives teachers and parents a quick visual snapshot of what was taught.
The drawings are intentionally simple—basic shapes, lines, and labels. This keeps the focus on the student’s explanation, not on the art.
Assessment
NoNothinBot can be a good option for people who hate tests.
NoNothinBot changes how assessment feels for some learners. Instead of feeling directly tested, students are asked to help a confused beginner, which for some can lower anxiety and make it easier for them to explain what they know, notice gaps, and try again. The bot’s drawings add another layer—students can see whether the bot “got it” visually, which often prompts them to clarify or correct their teaching without being told to.
Students can even joke about the bot for being “dumb” as they try to teach it. In the process, they also get to feel their own knowledge gaps—hopefully with more of a curious sensation of wonder and realization about their own understanding.
Related Traditions, Sayings, and Quotes
Famous quotes have referred to the idea of learning by teaching, such as:
Seneca: “While we teach, we learn” (Docendo discimus).
Savitribai Phule (India, 1831–1897): India’s first female teacher and a pioneering social reformer. Phule’s educational philosophy emphasized that students should immediately share what they learned with their families—a practice of “learning-by-sharing” that reinforced understanding through teaching. (From historical accounts of her educational work in Pune; see Mani & Sardar, 1988.)
The Upanishads (Ancient India): The concept of Shravana-Manana-Nididhyasana—learning happens in three stages: hearing (Shravana), reflecting/internalizing (Manana), and meditating/mastering (Nididhyasana). Mastery requires the deep reflection needed to explain it to others.
Japanese Proverb: “Manabu ni oshieru” (To learn is to teach)—the acts of learning and teaching are two sides of the same coin; explaining a concept to someone else is what actually solidifies your own understanding.
Joseph Joubert: “To teach is to learn twice.”
Lightweight Rubric (Adapt as Needed)
Ways to Describe NoNothinBot to Students
Sometimes you need a quick phrase to explain the idea. Here are some that work:
For test haters
The bot that needs to be taught
The bot that forgot
The Anti-AI—it knows nothing
Test the bot, not the student
References
Primary Sources for the Feynman Technique
The technique is a synthesis of Richard Feynman’s learning philosophy as described in James Gleick’s biography Genius: The Life and Science of Richard Feynman and Feynman’s own book Surely You’re Joking, Mr. Feynman! Modern interpretations are often credited to Farnam Street (fs.blog).
Gleick, James. Genius: The Life and Science of Richard Feynman. (Biography describing Feynman’s “Notebook of Things I Don’t Know About”—the origin of the technique.)
Feynman, Richard. Surely You’re Joking, Mr. Feynman! (Feynman’s own account of his teaching and simplification mentality.)
Farnam Street. “The Feynman Technique: The Best Way to Learn Anything.” https://fs.blog/feynman-technique/
Practitioner Resources
Li, James. “What Is the Feynman Technique and Why Does It Work So Well?” The Stuyvesant Spectator, Issue 6, Volume 113. https://stuyspec.com/article/what-is-the-feynman-technique-and-why-does-it-work-so-well
Edutopia (George Lucas Educational Foundation). “The Benefits of Having High School Students Learn by Teaching.” https://www.edutopia.org/article/having-students-learn-teaching/
Todoist. “Beginner’s Guide to the Feynman Technique” (YouTube video). https://youtu.be/XdM0k6EL7Dw
Encyclopaedia Britannica. “Socratic method.” https://www.britannica.com/topic/Socratic-method
InnerDrive. “What Is the Protégé Effect, and How Does It Work?” By the InnerDrive team, edited by Bradley Busch. https://www.innerdrive.co.uk/blog/what-is-the-protege-effect/
Wikipedia. “Learning by teaching,” “Socratic method,” “Pedagogical agent.”
Academic & Peer-Reviewed Sources
Chase, C. C., Chin, D. B., Oppezzo, M. A., & Schwartz, D. L. (2009). Teachable agents and the protégé effect: Increasing the effort towards learning. Journal of Science Education and Technology, 18(4), 334–352. https://link.springer.com/article/10.1007/s10956-009-9180-4
Kobayashi, K. (2019). Interactivity: A potential determinant of learning by preparing to teach and teaching. Frontiers in Psychology, 9, 2755. https://pmc.ncbi.nlm.nih.gov/articles/PMC6336728/
Kobayashi, K. (2019). Learning by preparing-to-teach and teaching: A meta-analysis. Japanese Psychological Research, 61(3), 192–203. https://doi.org/10.1111/jpr.12221
Yale Poorvu Center for Teaching and Learning. “Formative and summative assessments.” https://poorvucenter.yale.edu/teaching/teaching-resource-library/formative-summative-assessments
Hudesman, J., et al. “Using formative assessment and metacognition to improve student achievement.” ERIC. https://files.eric.ed.gov/fulltext/EJ1067283.pdf
Visual Learning & Dual Coding
Clark, J. M. & Paivio, A. (1991). Dual coding theory and education. Educational Psychology Review, 3(3), 149–170.
Mayer, R. E. (2009). Multimedia Learning (2nd ed.). Cambridge University Press. DOI: 10.1017/CBO9780511811678
Wikipedia. “Dual-coding theory.” https://en.wikipedia.org/wiki/Dual-coding_theory (An accessible overview of Allan Paivio’s theory that the mind processes verbal and visual information through separate channels, and that combining both strengthens memory and comprehension.)
--Best,Caleb Clark-----------------------------------------------------------------------
Search the SCHOOL-IT Archive
Manage your Subscription to SCHOOL-IT
--Lucie deLaBruere (she/her)
www.CreateMakeLearn.orgGoogle Voice (802) 404-1640
[log in to unmask]
--------------------------------------
Nothing is really work unless you would rather be doing something else.
- James M. Barrie
---------------------------------------Google Certified Educator / Google for Education Certified Trainer, Raspberry Pi Certified Educator
-----------------------------------------------------------------------
Search the SCHOOL-IT Archive
Manage your Subscription to SCHOOL-IT
