What is vibe coding—and why are educators suddenly building apps without “knowing code”?
In this AI Foundations video from Ed3, we explain vibe coding: a new way of building software by describing the intent, feel, and outcome of a tool—instead of writing code line by line.
With vibe coding, you tell an AI what you want and why it matters, and the AI translates that intent into working code. You’re not programming logic step-by-step—you’re shaping behavior through language. A well-structured prompt becomes the “blueprint” for the tool.
This video covers:
A common misconception is that vibe coding replaces programming. It doesn’t. It reframes it. The future isn’t code-free building—it’s intent-driven building, where AI generates structure but humans supply purpose, judgment, and responsibility.
This video is part of the AI Foundations series by Ed3, supporting educators worldwide in making informed, ethical, and human-centered decisions about AI in classrooms.
👉 Learn more about Ed3: https://www.ed3global.org
👉 Explore professional learning, courses, and events designed for educators navigating AI responsibly.
👉 Join our community of practice: https://community.ed3global.org
00:04 A classroom tool built with no code
00:32 What vibe coding means
00:46 Language → working code
01:07 Ed3’s GECCO prompt structure
02:04 Why this matters for educators
02:31 Examples of tools teachers can build
03:06 The risks of “effortless” creation
03:33 Three responsible practices
04:02 Vibe coding doesn’t replace programming
04:13 Humans supply purpose
1
00:00:04,866 --> 00:00:06,266
Check this out.
2
00:00:06,266 --> 00:00:08,633
It's a classroom mood tracker.
3
00:00:08,633 --> 00:00:10,200
And when you start a new lesson,
4
00:00:10,200 --> 00:00:10,866
you can log
5
00:00:10,866 --> 00:00:12,200
whatever mood the class is
6
00:00:12,200 --> 00:00:14,266
in at any point in time.
7
00:00:14,266 --> 00:00:15,666
You can even see a mood summary
8
00:00:15,666 --> 00:00:18,800
from each lesson and even analyze trends.
9
00:00:19,300 --> 00:00:20,733
But the best part is
10
00:00:20,733 --> 00:00:23,366
I didn't write a single line of code.
11
00:00:23,366 --> 00:00:26,033
I just described what I wanted and voila,
12
00:00:26,033 --> 00:00:27,166
it worked.
13
00:00:27,166 --> 00:00:28,366
That experience has a name
14
00:00:28,366 --> 00:00:30,133
now and people are calling it
15
00:00:30,133 --> 00:00:32,400
vibe coding.
16
00:00:32,400 --> 00:00:34,533
Vibe coding is a way of building software
17
00:00:34,533 --> 00:00:36,366
by describing the intent,
18
00:00:36,366 --> 00:00:38,100
feel, and outcome,
19
00:00:38,100 --> 00:00:39,300
instead of writing precise
20
00:00:39,300 --> 00:00:41,400
code instructions yourself.
21
00:00:41,400 --> 00:00:43,066
Rather than telling a computer
22
00:00:43,066 --> 00:00:45,466
how to do something step by step,
23
00:00:45,466 --> 00:00:46,966
you tell an AI what you want
24
00:00:46,966 --> 00:00:48,400
and why it matters.
25
00:00:48,400 --> 00:00:49,666
The AI translates
26
00:00:49,666 --> 00:00:52,166
that intent into working code.
27
00:00:52,166 --> 00:00:54,900
You're not programing logic line by line,
28
00:00:54,900 --> 00:00:57,900
you're shaping behavior through language.
29
00:00:57,966 --> 00:00:59,300
The logic in vibe coding
30
00:00:59,300 --> 00:01:01,400
is generated by your prompt.
31
00:01:01,400 --> 00:01:02,433
With a specific prompt
32
00:01:02,433 --> 00:01:03,366
that names goals,
33
00:01:03,366 --> 00:01:04,966
inputs, and constraints.
34
00:01:04,966 --> 00:01:07,566
You can build pretty incredible things.
35
00:01:07,566 --> 00:01:09,000
Here's how we structure our prompts
36
00:01:09,000 --> 00:01:11,433
at Ed3. We use GECCO's.
37
00:01:12,666 --> 00:01:13,333
First,
38
00:01:13,333 --> 00:01:14,466
we identify what
39
00:01:14,466 --> 00:01:16,200
we're hoping to accomplish.
40
00:01:16,200 --> 00:01:17,533
This is the goal.
41
00:01:17,533 --> 00:01:19,200
We want this to be something specific,
42
00:01:19,200 --> 00:01:21,433
but we also want to name our intentions
43
00:01:21,433 --> 00:01:22,700
for the idea.
44
00:01:22,700 --> 00:01:24,733
Do we want to increase human connection?
45
00:01:24,733 --> 00:01:25,900
or do we want to make something
46
00:01:25,900 --> 00:01:27,733
more efficient?
47
00:01:27,733 --> 00:01:28,500
Next,
48
00:01:28,500 --> 00:01:31,400
we add any external inputs like data,
49
00:01:31,400 --> 00:01:34,266
background information, or context.
50
00:01:34,266 --> 00:01:36,200
Constraints are the rules and boundaries
51
00:01:36,200 --> 00:01:37,800
we want to set, things
52
00:01:37,800 --> 00:01:40,700
we don't want the bot to do or create.
53
00:01:40,700 --> 00:01:41,966
Conditions specify
54
00:01:41,966 --> 00:01:44,300
how the AI should change its behavior
55
00:01:44,300 --> 00:01:46,200
given certain scenarios.
56
00:01:46,200 --> 00:01:47,300
For example,
57
00:01:47,300 --> 00:01:49,966
if it encounters a data security issue,
58
00:01:49,966 --> 00:01:50,933
should I keep going
59
00:01:50,933 --> 00:01:53,566
or stop to ask questions?
60
00:01:53,566 --> 00:01:56,400
And the O in gecko stands for output.
61
00:01:56,400 --> 00:01:57,300
This identifies
62
00:01:57,300 --> 00:01:59,100
what the format of the final product
63
00:01:59,100 --> 00:01:59,766
should be,
64
00:01:59,766 --> 00:02:02,366
like a web app or a report, or a booking.
65
00:02:03,300 --> 00:02:04,466
For educators,
66
00:02:04,466 --> 00:02:05,833
vibe coding changes
67
00:02:05,833 --> 00:02:07,566
who gets to be a builder.
68
00:02:07,566 --> 00:02:08,933
Teachers who have never taken
69
00:02:08,933 --> 00:02:10,400
a programing class can create
70
00:02:10,400 --> 00:02:12,133
simple classroom tools,
71
00:02:12,133 --> 00:02:13,666
prototype ideas for student
72
00:02:13,666 --> 00:02:16,100
reflection, feedback or collaboration,
73
00:02:16,100 --> 00:02:18,000
test solutions to real problems
74
00:02:18,000 --> 00:02:19,133
they see every day,
75
00:02:19,133 --> 00:02:19,666
and build
76
00:02:19,666 --> 00:02:21,233
with students and encourage them
77
00:02:21,233 --> 00:02:22,733
to prototype solutions,
78
00:02:22,733 --> 00:02:23,700
bringing practices
79
00:02:23,700 --> 00:02:25,566
like design thinking and project
80
00:02:25,566 --> 00:02:26,400
based learning
81
00:02:26,400 --> 00:02:29,400
to totally new, practical levels.
82
00:02:29,533 --> 00:02:31,066
Some things that a teacher might vibe
83
00:02:31,066 --> 00:02:33,500
code might be a daily checking app,
84
00:02:33,500 --> 00:02:35,066
a peer matching tool,
85
00:02:35,066 --> 00:02:36,733
a classroom mood tracker,
86
00:02:36,733 --> 00:02:37,966
or a reflection journal
87
00:02:37,966 --> 00:02:40,500
that adapts to student responses.
88
00:02:40,500 --> 00:02:42,633
The shift is powerful.
89
00:02:42,633 --> 00:02:45,066
You're moving from using tools
90
00:02:45,066 --> 00:02:46,966
to shaping tools.
91
00:02:46,966 --> 00:02:48,600
Another cool thing about vibe coding
92
00:02:48,600 --> 00:02:50,433
is that while educators and students
93
00:02:50,433 --> 00:02:51,500
are using the GECCO's
94
00:02:51,500 --> 00:02:53,366
prompt to build a tool,
95
00:02:53,366 --> 00:02:54,933
they're actually better understanding
96
00:02:54,933 --> 00:02:58,200
how AI works. By identifying the goals,
97
00:02:58,200 --> 00:02:59,900
constraints, and conditions,
98
00:02:59,900 --> 00:03:01,866
and then assessing the final outcome,
99
00:03:01,866 --> 00:03:03,400
they're learning about the limitations
100
00:03:03,400 --> 00:03:05,500
and potential of AI.
101
00:03:05,500 --> 00:03:06,666
But here's where we need
102
00:03:06,666 --> 00:03:08,333
to slow down a bit.
103
00:03:08,333 --> 00:03:10,933
When I handles the technical details,
104
00:03:10,933 --> 00:03:11,600
it's easy
105
00:03:11,600 --> 00:03:13,866
to stop asking important questions like
106
00:03:13,866 --> 00:03:15,833
why does this behave the way it does?
107
00:03:15,833 --> 00:03:18,066
What assumptions did the system make?
108
00:03:18,066 --> 00:03:19,666
And who is this designed
109
00:03:19,666 --> 00:03:21,266
for and who might it miss?
110
00:03:22,266 --> 00:03:22,900
If we treat
111
00:03:22,900 --> 00:03:25,400
vibe coding as effortless creation,
112
00:03:25,400 --> 00:03:26,766
we risk producing tools
113
00:03:26,766 --> 00:03:28,966
we don't fully understand, control,
114
00:03:28,966 --> 00:03:30,000
or solve the problem
115
00:03:30,000 --> 00:03:32,066
we actually have.
116
00:03:32,066 --> 00:03:33,266
There are three ways to use
117
00:03:33,266 --> 00:03:35,866
vibe coding responsibly in education.
118
00:03:35,866 --> 00:03:38,866
First, start with problems, not products.
119
00:03:39,400 --> 00:03:41,166
Name the need clearly before
120
00:03:41,166 --> 00:03:42,533
building anything.
121
00:03:42,533 --> 00:03:45,300
Second, interrogate the output.
122
00:03:45,300 --> 00:03:46,733
Test edge cases,
123
00:03:46,733 --> 00:03:49,333
ask what happens when things go wrong.
124
00:03:49,333 --> 00:03:50,333
And third,
125
00:03:50,333 --> 00:03:53,166
use vibe coding as a learning experience.
126
00:03:53,166 --> 00:03:54,600
Invite students to reflect
127
00:03:54,600 --> 00:03:56,700
on how language becomes logic
128
00:03:56,700 --> 00:03:59,533
and how ambiguity shapes outcomes.
129
00:03:59,533 --> 00:04:02,133
Vibe coding doesn't replace programming.
130
00:04:02,133 --> 00:04:03,933
It reframes it.
131
00:04:03,933 --> 00:04:06,333
The future isn't code free building,
132
00:04:06,333 --> 00:04:09,066
it's intent driven building.
133
00:04:09,066 --> 00:04:10,733
AI may generate the structure,
134
00:04:10,800 --> 00:04:13,800
but humans still supply the purpose.
135
00:04:13,800 --> 00:04:15,200
As educators,
136
00:04:15,200 --> 00:04:16,566
knowing what vibe coding is
137
00:04:16,566 --> 00:04:17,300
helps us separate
138
00:04:17,300 --> 00:04:19,466
the hype from the reality
139
00:04:19,466 --> 00:04:20,500
so we can make wise
140
00:04:20,500 --> 00:04:21,866
choices for our classrooms.
