Six Simple Projects Using Cursor
Key Points
- The speaker showcases six community‑built projects using Cursor (a weather app, a video‑search tool, a non‑coder’s Trello clone, a child‑created chatbot, a polished macOS voice‑to‑video app, and a Python AI demo).
- All of the highlighted examples are relatively simple, prompting the question of whether Cursor can handle larger, more complex applications or extensive codebases.
- Three hypotheses are offered for the prevalence of simple demos: the 20 k‑token context window may limit how much code the model can reason over at once.
- Additional possible constraints include the current UX of Cursor that favors quick, single‑file generation rather than long‑term project scaffolding and a lack of community‑shared best practices for scaling up.
- The speaker suggests that examining these patterns across the ecosystem will help determine Cursor’s strengths, its current limits, and the kinds of tools or improvements needed for more ambitious development tasks.
Full Transcript
# Six Simple Projects Using Cursor **Source:** [https://www.youtube.com/watch?v=I4WzI6eO3RI](https://www.youtube.com/watch?v=I4WzI6eO3RI) **Duration:** 00:05:23 ## Summary - The speaker showcases six community‑built projects using Cursor (a weather app, a video‑search tool, a non‑coder’s Trello clone, a child‑created chatbot, a polished macOS voice‑to‑video app, and a Python AI demo). - All of the highlighted examples are relatively simple, prompting the question of whether Cursor can handle larger, more complex applications or extensive codebases. - Three hypotheses are offered for the prevalence of simple demos: the 20 k‑token context window may limit how much code the model can reason over at once. - Additional possible constraints include the current UX of Cursor that favors quick, single‑file generation rather than long‑term project scaffolding and a lack of community‑shared best practices for scaling up. - The speaker suggests that examining these patterns across the ecosystem will help determine Cursor’s strengths, its current limits, and the kinds of tools or improvements needed for more ambitious development tasks. ## Sections - [00:00:00](https://www.youtube.com/watch?v=I4WzI6eO3RI&t=0s) **Diverse Projects Using Cursor AI** - The speaker reviews six community-built examples—from a simple JavaScript weather app to a video transcript search tool and a non‑coder’s LLM coding journey—to illustrate Cursor’s strengths, limitations, and potential applications. ## Full Transcript
what are people actually building with
the AI Cod assist cursor I found six
different examples I'm going to go
through them briefly here and then I
want to actually step back and ask what
we can learn from looking across the
ecosystem about what cursor is good at
what it's not good at what we haven't
seen yet and I think that will help us
answer some of the larger questions that
come to mind like is cursor good for
complex apps is cursor good for large
code bases things like that so example
number one uh this is from one little
coder I'm going to put all of these in
the YouTube comments uh weather app can
we write a weather app in JavaScript and
get it runnable very simple uh and it's
exactly what it says in the tin can you
basically pull publicly available
weather data turn it into a nice
graphical user
interface uh all about AI the creator
has a YouTube Tech search app it looks
simple but I do appreciate the
functionality here it basically allows
you to pull a video and do a text search
through the video uh because it pulls
all of the transcript down and you can
actually search through and you can get
a time stamp for particular things
you're looking for so if you're going
through and you're trying to understand
like where does this person talk about X
or Y you can just search for the word
and find it really
fast Riley Brown uh is building a
following just talking about his journey
as someone who doesn't code but is
coding using llm so that's his whole
thing and he uh talked about cursor in
the context of building a Trello app
CopyCat um looks pretty simple to me uh
but I will definitely link it in you can
look at
it uh a simple chatbot uh the
interesting thing here uh the Creator
Ricky had uh is teaching his kid to code
and the kid built the chatbot herself
and she's eight uh which as someone with
a kid very impressive my kid has not yet
built a chatbot
um and then Amar who's a designer at 11
Labs took the weekend this past weekend
and built a Mac app fully functional
that's designed to solve for importing
voice into video and that's like he's a
Creator like it makes sense for him to
do that and I was impressed I mean maybe
we shouldn't be surprised he's head of
design it looks gorgeous like it's a
really well-designed app it looks the
most polished of any of these that I've
SC um and then the last one I want to
call out is uh chantastic is writing
about or doing a video on cursor and
python so if you're wondering how does
cursor help you build in Python for AI
use cases there's a video for that too
so those are six different use cases
again I'll link them all
below I want to step back though when
you see the pattern here I think you
won't be surprised when I call out that
most of these are really really simple
use cases and and I have three different
hypotheses for why that might be true
one is the context window the context
window right now is 20,000 tokens on
cursor is it because the context window
is short uh well not short right that's
still dozens of pages but is it short
enough that it's an issue for larger
code bases
maybe is it that it is simple because
people are rushing to production want to
show what they've built there's this
sort of Creator urge to get out there
and show you can push to production and
people aren't really thinking through
more complex use cases yet
maybe is it because the lack of design
thinking leads people to solve really
simple pinpoint problems that are useful
to them but maybe not more fundamental
and may not require a larger build and
that one kind of cuts both ways because
a lot of the advice for MVPs is to build
it simple build for a paino that you
know really well so in a sense cursor
enabling people to act on that general
advice for sort of how you get into
software how you get into building
things that are
useful I'm not sure which it is uh
cursor has ways of handling context
windows that sort of stretch around uh
the 20,000 limit sometimes they don't
really talk about that uh very much I
wasn't able to find out a lot about it
other than that they apparently have
some proprietary methods of doing
so I think the design thinking is
something I'm taking away I think it's
going to be increasingly important if
software costs come down to be really
thoughtful about what you want to build
how you want it to look and why to stand
out and so in that sense I think a Mars
app uh really shows the way that you can
you can build something with an llm even
if you don't know how to code and you
can make it look really nice you can
make it look really polished and that's
going to become increasingly important
in order
to justify the value of your software in
a world where anyone and everyone is
coding up web apps at the drop of a hat
which is the other big take way here
like if an 8-year-old is coding a chat
app software is essentially free for
anyone to produce so all that being said
I'm curious what your take is where what
patterns do you see in this list of use
cases I found are there use cases that I
miss that people are building in how do
you see cursor playing into the larger
llm ecosystem