TryShadowing
Shadow YouTube. Nói tiếng Anh.
Trang chủ
Khám phá
Chép chính tả
NEW
Thư viện
vi
0
ngày
Đăng nhập
Common Mistakes With Vibe… — Y Combinator luyện shadowing | TryShadowing
TryShadowing
Shadow YouTube. Nói tiếng Anh.
Trang chủ
Khám phá
Chép chính tả
NEW
Thư viện
vi
0
ngày
Đăng nhập
Trang chủ
Khám phá
Y Combinator
Common Mistakes With Vibe Coded Websites
Common Mistakes With Vibe Coded Websites
Y Combinator
·
37:26 · 6 thg 3, 2026
Bắt đầu học
0:00
0:00
Ghi âm
×1
1x
VI
EN
JA
KO
ZH
FR
PT
TH
IT
DE
IPA
Chấm điểm phát âm chưa hỗ trợ trên trình duyệt này — bạn vẫn ghi âm & nghe lại được.
It's
easier
than
ever
to
design
with
AI
tools,
Đang dịch…
Bật Ghi âm để được thu giọng và chấm điểm
Thông minh
Karaoke
Câu
1
/927
0:00
It's easier than ever to design with AI tools,
0:03
but that doesn't mean you should always accept all changes.
0:06
So today I'll be joined by Raphael Shad to look at the good,
0:10
the bad, and the ugly of AI assisted design
0:14
and sort through how to use these superpowers to stand out.
0:17
Welcome to another design review.
0:26
All right, today I am thrilled to be joined by Raphael Shad,
0:30
one of the newest visiting partners here at YC.
0:33
And Raphael was the co-founder of Chron, a modern calendar tool,
0:37
which you sold to Notion a
0:39
while back and are also one of the top designers
0:43
that I know and have worked with.
0:44
So Raphael, it's an honor to have you here
0:46
and it's an honor to work together now at YC.
0:48
Yeah, super fun to be now back in in in YC
0:53
and helping founders next generation founders build iconic companies.
0:56
Well, you're the perfect person to have on here
0:59
because topic that we're going to be talking about today is AI design trends.
1:03
On the negative end there's the AI design slop
1:06
and on the positive end there's a lot of AI design trends
1:09
that that now are easier to build for anybody even
1:12
if they're not a designer.
1:14
So what are some of the common things that you've seen recently,
1:17
a lot of the AI top AI design trends
1:19
that are standing out
1:20
that that you'd want to call out for the audience before we jump into
1:23
some sites.
1:23
>> Yeah, sure.
1:24
This all kind of started
1:25
when I had like a late night thought
1:27
and tweeted that I see a lot of dumb hover effects on landing pages
1:31
of startups these days.
1:33
Presumably wipe coded and
1:34
so I was kind of curious to peel the the the layer back there.
1:38
It's like how did these like what I thought were dumb effects,
1:41
how did they make it into LLMs and why are they're everywhere now?
1:44
A couple other trends that we then identified was kind of like purple gradients.
1:48
All of a sudden all startup websites had purple gradients everywhere
1:52
or these sections that kind of like fade
1:54
as you go in
1:55
as you scroll and they fade in fade out.
1:57
Yeah, and it's not so much that those inherently are bad.
2:00
It's not purple gradients are bad.
2:01
Nobody should ever use them.
2:03
It's that now they're just
2:05
so ubiquitous everywhere that they kind of lose all meaning
2:09
and specialness and originality.
2:11
Totally.
2:11
And one of the key things was
2:13
when there was a good website kind of establishing a trend,
2:16
it took a while in the old world for others to kind of like
2:19
copy these trends.
2:21
But now with LLMs, if there's a good website with a purple gradient,
2:24
it makes it into the LLM
2:25
because the LLM gets trained on like the good examples
2:28
that get linked to a lot.
2:29
And then all of a sudden like the next week all the startup websites
2:32
look the same.
2:33
Yes.
2:34
And so this is both good
2:36
and bad if this power is used in the right ways.
2:40
So we've got a bunch of YC companies
2:42
that have submitted their websites
2:43
that were all built with AI coding tools.
2:46
So this should give us good grounds to figure out ways
2:49
that they have been used really really well
2:51
and things that we would do differently
2:53
if we were designing it ourselves.
2:55
So let's let's dive into some of this.
2:56
>> Let's have a look.
2:57
All right.
2:58
Okay, first up we've got new new.ai.
3:01
Solve your game testing with AI agents.
3:03
Save countless hours of manual game QA by using AI agents performing end-to-end tests
3:08
at scale.
3:09
Purple gradients.
3:10
>> The first thing that jumps out is very purple site which you know,
3:14
we were just talking about.
3:15
That's one of the most common things that you see.
3:17
So it's interesting like pre-LLM day,
3:20
I think I would have looked at this and gone like wow,
3:22
this looks really nice.
3:23
And now I look at it
3:24
and I'm like they used AI to design this.
3:26
Maybe they didn't but like that's what it screams to me.
3:30
Okay, so one thing is interesting as we as you start scrolling here,
3:33
we've got this line that's following us up and down.
3:37
And it's interesting because I actually find this this line to be a little
3:41
distracting.
3:41
I'm paying more attention to the line than I am the the things
3:44
that I think that it wants me to pay attention to.
3:47
It's it's almost like something
3:49
that it's kind of like probably hard to implement for a human
3:52
and you would have never thought to actually code this up before AI
3:58
and just because we now can just
4:00
because LLM's are kind of good at these type of like SVG you know
4:03
build ups or transforms doesn't mean
4:05
that it's actually a good design
4:06
and helps you convert potential you know visitors into into customers basically.
4:11
Yeah.
4:12
Yeah, I don't know anybody that's like we need a line
4:14
that goes through here.
4:15
Let's start coding that from scratch.
4:16
Like you would only implement something like this
4:18
because it was so easy to do.
4:20
Um but I don't think it actually adds value.
4:22
If anything I think it distracts from the rest of the page.
4:26
Um I can't recall anything from these sections based on this.
4:32
Um okay, so now we're down to everything in one place.
4:36
You I find this hard to read.
4:38
It's uh it's it's very light.
4:40
Um maybe I'm just old.
4:42
Um Yeah, contrast could be improved there a little bit.
4:45
So these cards are interesting, right?
4:47
Where >> Um this is actually pretty good.
4:49
Uh this is uh probably something
4:50
that would have been maybe too expensive to do
4:52
or only like the very best
4:54
or like website designers would have kind of like got gotten to the lengths
4:57
of of doing something like
4:58
that.
4:58
These hover animations for the cards.
5:00
But AI is so simple to do
5:02
and this is actually like a pretty tasteful um way of using leveraging
5:06
so like the powers of LLM's to to build more compelling websites.
5:10
>> Yeah, what's cool about it is it feels like it helps establish the
5:15
brand and it's unique
5:16
and it's creative and it's fun.
5:19
Um and it it's actually trying to you know multiplayers like two controllers with
5:24
lightning connecting them.
5:25
Um and it actually helps reinforce the points
5:28
that I think that they're trying to make here.
5:30
Um and this is a great example of you know where the line
5:33
that was following you down the page didn't add much value.
5:36
This is this would have been really hard to do
5:38
and now um you kind of get it for free.
5:41
Yeah.
5:42
Um you don't have to go from scratch building a lot of this stuff.
5:46
One of the things I want to try out is, um, up here.
5:48
Um, so we here have the main navigation.
5:52
One of the things that jumps out here is that as you hover,
5:55
you generally kind of want to make things pop.
5:57
But here, it actually kind of disappears.
6:00
So instead of inviting to click, to talk to them, um,
6:03
it kind of like goes into the background.
6:04
Same here, it kind of fades out
6:06
as you activate one of these menu points.
6:08
Um, this is something
6:09
that I think a designer would almost never choose to do.
6:12
And now, uh, in the default of browsers, it also doesn't work that way.
6:16
And now LLNL just somehow decided that that's a good, uh,
6:19
hover kind of like effect, um, but it makes no sense.
6:23
Um, think about it that way.
6:25
Like the browser already has like a built-in hover effect for free,
6:28
even without CSS.
6:29
It turns the cursor into a hand, right?
6:31
So that already indicates like I can click here.
6:33
So just leaving it white in this case,
6:35
or just kind of like the the bright purple,
6:38
and then having the hand already tells me like I can click on this
6:40
thing.
6:41
So if you want to add a CSS effect, then, you know,
6:44
make it maybe like a little pop.
6:46
Maybe make it like one shade lighter, um, or more active,
6:49
or add like a little, you know, a little, uh, glow around it.
6:52
But don't make it go away.
6:54
Yeah.
6:55
Do you have any rules of thumb when, you know,
6:56
you're designing something and you're trying to come up with hover effects?
6:59
Like, what are your goals
7:01
when you're trying to come up with the right hover effect?
7:03
>> Yeah, make it sort of like clickable.
7:04
Um, Steve Jobs famously said, you know,
7:06
he wanted to make things lickable with the Aqua interface back in the day.
7:09
I think, you know, make things, you know, clickable,
7:11
kind of like inviting to to click.
7:13
Stick to sort of like pretty standard.
7:15
Again, kind of I think the hand indicates that something is clickable,
7:18
and so I wouldn't go much beyond that.
7:20
Another thing that I see hover being used, um,
7:23
is a lot to kind of like reveal additional information.
7:25
So not just kind of like a hover effect to make the UI more
7:29
feel alive.
7:30
Like hover is great to kind of make UIs feel more alive, right?
7:33
Um, but I see it increasingly used to kind of like, uh,
7:36
disclose additional and critical information or functionality,
7:39
and I don't think that's the best use for hover.
7:42
This whole trend uh,
7:43
that maybe like kind of like Mac OS kind of started to over the
7:46
years get more and more into kind of like very clean interfaces.
7:49
Get the functionality and the tools and the buttons all out of the way.
7:52
So, there's all this space for content.
7:54
I don't want to have my computer to be just like all content.
7:57
I actually want it to be, you know, a bicycle for the mind,
7:59
a tool.
8:00
And so, revealing additional functionality only on hover
8:05
and you have to go hunt for what the tool can actually do.
8:08
I think it's kind of like an anti-pattern.
8:10
Yeah.
8:10
Not to mention that, you know,
8:11
there isn't a concept of hover on mobile as well.
8:14
So.
8:14
>> Yeah.
8:14
And long press, which is kind of the equivalent of of hover,
8:18
never really catch caught on on mobile.
8:21
Probably because it's so hard to discover.
8:23
You don't want to sit there long pressing all the time to try to
8:26
figure out if if there is any additional functionality behind the desktop interface,
8:30
a tool where you have to hover around everywhere to discover where the stuff
8:32
actually lives.
8:34
You know, the disclosure arrows
8:35
or like hiding something
8:37
or getting to a dot dot dot kind of menu to get to extra
8:39
functionality.
8:40
If that's all hidden behind hover, then, you know,
8:42
you're just not as efficiently communicating the functionality of your software.
8:46
>> Yeah.
8:46
Overall, I think this is a very nicely designed site.
8:50
I think it's easier than ever using AI design tools
8:53
and LLMs to get a professional-looking site.
8:56
There's almost no excuse for not having it now.
8:58
And especially as we're reviewing YC applications,
9:01
if you look at a demo
9:02
and that demo does not have a base level of quality design,
9:07
it seems like the person just didn't even try.
9:10
Because it's so easy to do it now.
9:12
So, Nunu, really cool site
9:14
and some some really interesting applications of using AI here.
9:18
So, well done.
9:19
All right, next up we got Rosebud AI.
9:22
Create games with AI.
9:23
And and once again, like you start to see the trend.
9:25
Here we've got the purple gradient again.
9:27
We've got a very similar kind of pinkish purple accent color up here.
9:32
And that's why like in isolation with one of of like it looks cool.
9:36
It looks modern everything, but you know,
9:38
one after the other they all start to kind of look the same
9:41
and it takes away a lot of the core brand
9:44
and originality and things like
9:45
that that I think most founders want for their site.
9:47
So, let's dig in here.
9:48
This is the first time we're looking at these live.
9:51
>> Yeah.
9:51
Um so, Rosebud.ai,
9:53
create games of AI
9:55
and there's a prompt to to to prompt to game kind of interaction,
9:59
I would assume.
10:01
Um That's cool.
10:02
There's something you can instantly play.
10:04
Yeah.
10:06
Let's try it.
10:06
>> Woah, it's like a 3D game in the browser.
10:10
That's cool.
10:10
>> There's a person walking around.
10:13
I can use the arrows.
10:15
Let's try it.
10:16
There we go.
10:16
Oh.
10:17
WASD, we have a old school gamer here.
10:23
>> >> Oh, you can jump.
10:25
All right, let's escape out of this.
10:28
Um but cool, definitely kind of like engaging.
10:32
Um I wonder whether this is actually their like product code running on the
10:36
website here or this is like live coded Yeah. sandbox environment.
10:41
Um Yeah, I would assume
10:41
that this is the type of thing you can create with their product,
10:44
but um that's an interesting point is that they don't really tell us.
10:47
Mhm.
10:47
You know, it doesn't say like play a game made with our product right
10:51
now.
10:52
Um it's interesting, is this following us this top bar?
10:54
>> these are just anchors jumping vertically around.
10:57
>> Yeah, so what's interesting is I didn't realize this at the time,
11:00
but when we went to that one we were just playing,
11:02
that was under modify games.
11:05
And so, what I wonder is
11:06
if this is a game
11:08
that is already been created
11:09
that we can modify in real time
11:11
and that's what they're trying to show off rather than the creation part.
11:14
Yeah.
11:14
So, maybe kind of like a thing to take away here is
11:17
that using this non kind of like non-standard navigation is tripping us off a
11:23
little bit on how to use this website.
11:26
Another thing that kind of jumped out here at the top is like I
11:28
don't think anyone would think of combining a red logo with purple
11:33
as an accent color necessarily.
11:36
Uh it's not necessarily complementary.
11:39
Uh and uh I guess whenever I see the use of emojis,
11:42
even though they're not the system emojis, I feel like it's a little lazy.
11:46
And so I I feel like LLMs kind of take the easy path
11:49
because they don't have any IP really themselves.
11:52
They're just like big models kind of cobbling things together.
11:55
They use like these standard icons everywhere
11:57
and it's just immediately kind of like a tell.
11:59
Yeah, I think we could just use more context up here, too.
12:02
And and you know, the the name of the company is is here.
12:06
It looks like it's almost part of the headline.
12:09
Rather than I would expect it to be up here by the logo.
12:11
Yeah, and then the value prop kind of being more clear.
12:14
You know, create games with AI, something something.
12:17
It's kind of like what's important for the H1 is typically kind of you
12:21
know,
12:21
what is it?
12:23
Who is it for?
12:24
And to what end?
12:26
Why should that person that it's for care?
12:28
And if those three things plus call to action are not kind of above
12:33
the fold,
12:33
then it's a harder time to convert convert visitors.
12:37
Yeah.
12:38
Um okay, we've got some I guess examples of games.
12:41
Yep, so games created with Rosebud AI.
12:44
Um okay, this is interesting.
12:46
So, there's another effect here.
12:47
You can see the light following my cursor um around the edges.
12:51
And it Yeah, this is the kind of thing where, you know, you're like,
12:54
"Hey, it costs nothing to do this."
12:56
Then maybe you're like, "Yeah, sure.
12:58
Let's throw it in.
12:58
It's fun and it's playful."
13:00
Um but if you had to spend a week coding this and you know,
13:03
getting it to work right, you'd be like, "Ah, that's not worth it."
13:05
Okay, so we've got a lot of a lot of game examples.
13:07
It looks like maybe this is just the rest of the site here.
13:09
Okay, and then we hit a footer.
13:11
There's a create your own game button.
13:13
>> Uh yeah, okay.
13:14
That says they disabled the button. >> of the some of the um projects
13:17
you can remix and
13:18
then it kind of puts up the conversion wall.
13:21
Okay.
13:22
Um that would go beyond sort of like landing page.
13:24
Mhm.
13:25
Very cool.
13:26
Awesome.
13:27
Well done, Rosebud.
13:28
Okay, so next up we've got Get Crocs.
13:31
And one of the first things
13:32
that I noticed here was kind of the fade in.
13:35
It was this automatic fade in.
13:37
Um Ooh, and another thing that I noticed is my scroll is hijacked here.
13:41
Guys, maybe a little hard to um to see on screen.
13:45
Uh but there's only some scroll jacking logic going on.
13:49
Um and there's also like a state where you see both the detached
13:53
and the fixed uh header menu all like both at the same time.
13:57
Like right right on here.
13:58
Yeah.
13:59
Um so typically kind of like they should just smoothly I like the sticky
14:02
header.
14:03
Um but it's just like smoothly transition one to the other.
14:05
And here you see something like that.
14:07
You know, they're clearly two distinct menus
14:09
that just happen to kind of be the same.
14:12
I'm sorry.
14:13
I'm so distracted by the button that's chasing me around on the screen here.
14:20
Um Yeah, this I almost feel like it's hard to get the button to
14:23
click on it cuz it's constantly moving.
14:25
>> Does it make you want to talk to the founders more or less?
14:28
It makes me distracted, I would say.
14:30
I'm not paying attention to what they do.
14:32
I'm just I'm like, "Why is this button following me around?"
14:36
Um it definitely got my attention, I would say.
14:39
Um one of the other things that I noticed first, too,
14:41
is almost these like meteors that are, you know, shooting down from the angles.
14:46
You can kind of see them over there.
14:48
Um that's another kind of thing where like it doesn't feel like it actually
14:52
adds a lot of value to the product.
14:54
And again, if you had to build that from scratch,
14:56
like it would never be worth your time.
14:58
Um but you it's so easy to do it that people are like, "Yeah,
15:01
let's throw that in.
15:01
It looks cool."
15:03
Um but I actually find it a little distracting.
15:05
Um I don't think it adds value to the the product here.
15:08
And the hero screenshot here is um I guess uh poster image for a
15:12
video.
15:14
Um Uh and it goes into what looks like a completely different video.
15:17
>> Yeah, that was confusing.
15:18
>> And the poster image is It's blurry here.
15:21
Um I don't know whether this is like an artifact of like using using
15:24
AI or not.
15:25
Um, but just as a general design principle,
15:28
I think all all your assets should be you know high-res, high-quality,
15:32
and this just looks like something very blur blurry.
15:34
Yeah.
15:35
And and again, back on the button, too,
15:36
cuz I still see even when we're down here,
15:38
it's like still chasing us around.
15:40
You know, this is another example of the type of thing
15:42
that you would only do
15:44
because it's easy.
15:45
But just because something is easy doesn't mean it's worth doing.
15:49
There's a lot of things that are now at your fingertips with AI that,
15:52
you know, anything imaginable is possible.
15:55
Just because something is possible doesn't mean you should say yes to it.
15:58
Here we're kind of like hitting a little bit on on just a lack
16:00
of visual consistency.
16:03
Like this looks like an entirely different visual language than anything we've sort of
16:07
like seen seen above.
16:09
Um, Where do you think that comes from?
16:12
I wonder.
16:13
Um, maybe different sections of the website being made by to like different parts
16:20
of the tool.
16:21
Um, maybe it's also just kind of like a function of like this here
16:24
looking very different um,
16:26
than so like the website content.
16:27
Yeah, sometimes just like a clear image with a little bit of text,
16:32
maybe a headline, and then some subtext, um, goes a long way.
16:35
Yeah.
16:37
And I got to say we're like halfway down the page right now,
16:39
and um, if you ask me what they did,
16:41
I would not be able to tell you.
16:43
Let's go back up to the top and remind ourselves.
16:46
AI creative strategist to make ads that work.
16:50
Ads that Okay, so it's something to make ads, it seems like.
16:53
It's interesting they do kind of have like a visual um,
16:58
like style to it with a lot of these uh,
17:00
it's almost like a vertical motion blur
17:02
or something on a color gradient
17:04
that we see in some of these images.
17:06
Um, like these right here, and we saw the same thing in these here,
17:10
and these here, um, and these here.
17:13
Um, but it's it's constrained to some of these elements, and it feels like,
17:17
you know, it's maybe a little overused here cuz it's hard to read these
17:20
and this feels like more like a rainbow rather than calling the attention to
17:23
it.
17:24
But up here does not have that style at all,
17:26
which is maybe you know,
17:27
the point that you're making earlier
17:29
and I could see something where the background here is more reflective of this
17:32
type of style and
17:33
then that might carry it through all the way through the rest of the
17:36
page.
17:36
>> Yeah.
17:37
Let me quickly reload the page and try something.
17:40
So we see all these sections kind of like fading in.
17:42
Um and here again,
17:44
it's just like fading in and one of the things that I noticed,
17:48
I just don't see the advantage of having these things not just right there
17:54
as I scroll because scrolling already like is the motion.
17:56
Why does it have to like also then kind of like fade in?
17:59
Um and one of the things
18:01
that you get is these weird um for example,
18:03
here, right?
18:04
We say, "We're here to answer all your questions."
18:06
And then there's one question
18:08
and we we we break to here earlier
18:10
and now it's like,
18:10
"Wow, that's a lame FAQ if there's only like one Q and one A."
18:16
Um but then it's like, "Aha, it just hasn't come in yet."
18:20
Um and so not a fan of that pattern.
18:23
Yeah.
18:24
What bothers me with the fade-ins a lot of time
18:26
and I'm not sure
18:26
if this does it
18:27
or not is when they're on a timer.
18:29
Sometimes I'm scrolling really quickly
18:31
and I'll get all the way past the section
18:34
as it's starting to fade in up at the top
18:36
and I miss the content altogether.
18:38
So I'm I'm curious to try
18:39
that here um and see how they did this.
18:42
So like if I'm just kind of zooming down.
18:45
Um so these seem to fade in um at least
18:49
when it's at the bottom of the page.
18:51
But it still feels like going through molasses because it's all scroll jacked.
18:55
Like hijacking the sort of like actual native browser scrolling um to do some
19:00
fancy thing with JavaScript to actually have the hooks to do all these animations.
19:05
>> Yeah.
19:05
All right.
19:05
>> Super interesting.
19:07
Very creative.
19:08
Uh so well done, Get Crocs.
19:11
Okay, next we got Sphinx, your last compliance hire.
19:15
All right, we got some animation thing going on here.
19:18
Um, I think one of the trends
19:19
that we've seen so far is just like way more animation these days than
19:24
there were,
19:24
you know, 3 years ago probably
19:26
because these AI design tools make it
19:28
so easy to do this.
19:29
>> Yeah.
19:30
Let's quickly talk about hierarchy here.
19:32
Um, so we start with um, a logotype,
19:35
kind of a stylized version of the name of the company, Sphinx, right?
19:38
That's cool, top left, pretty standard.
19:40
But then we have the H1, is also good, and the subtext,
19:45
but also we have like this kind of like weird in between,
19:49
like fourth style, right?
19:52
Um, not something that I typically ever saw designers
19:55
or engineers kind of come up with their own landing pages.
19:59
It's only something kind of like these many,
20:00
many styles mixed is something
20:02
that I increasingly see kind of LLMs do
20:05
because like somehow the LLM thought it's clever to have like a an extra
20:08
label here and an extra style.
20:10
Meet Sphinx, but we already have the company name, we already have, you know,
20:14
three different styles basically.
20:16
If you count count this here is like four styles.
20:18
This just adds a fifth, um, complicating the hierarchy, not really adding.
20:23
It adds a bunch of vertical space,
20:24
like we could just like move the H1 a little bit up and then,
20:27
you know, the value didn't trust it.
20:28
Um, logos would already be like a little higher.
20:31
Um, so this is something
20:32
that I increasingly see kind of like LLMs dream up
20:34
that just isn't really tight information hierarchy
20:37
and graphic design.
20:39
Yeah, I I think I going through these I I keep coming back to
20:42
the concept of like it's easy to do anything.
20:45
You really have to be um,
20:48
opinionated as the human that's in the loop in designing these around what you
20:53
think is the right thing,
20:54
not just kind of saying yes, accept all changes.
20:56
>> Yeah, you're now the editor.
20:58
>> Yes, yes.
20:59
And and another thing
21:00
that I'm noticing happening here is this section on the right,
21:03
it just keeps swapping around.
21:06
Um, if if this animation here for a second, this button over here,
21:10
the big star kind of button, uh, yeah,
21:13
like moves around and and now it's a circle button.
21:16
I think that's a button.
21:18
Yeah, I guess that's clickable.
21:20
Um, and then it's going to come back over to the left side
21:22
and it's going to be a big round rectangle kind of button.
21:25
And so we have a lot of different button styles.
21:27
I'm not sure why things are moving like this.
21:29
Um, and it's very confusing. >> not a button.
21:32
Like, yes, it shows the hand, but that's a head fake.
21:35
>> Yeah, I guess it just moves to the next thing. >> it doesn't really, well,
21:38
I guess it's kind of like a next button,
21:40
but then also on auto advance. >> very confusing to me.
21:43
I'm not sure what it's showing.
21:44
And and I can't tell what's happening in this animation either.
21:47
It seems to be some kind of workflow.
21:49
Um, actually you can just click anywhere and it it it advances.
21:53
So it's not even like this button
21:54
or the whole thing is a click target.
21:56
Yeah.
21:57
Um, this is again one of those things where it's hard for me to
21:59
imagine a human just being like,
22:01
let's have three different style buttons that every time you click it,
22:05
they all reorient and shift and, you know,
22:07
it's it feels like the type of thing
22:09
that an AI design tool would have suggested
22:12
and you would have said,
22:12
"Sure, that seems cool."
22:13
>> Yeah, it feels like the visual manifestation of LM hallucinations.
22:19
>> >> Um, let's go look at the menu here.
22:21
Um, all right.
22:25
Pretty standard stuff.
22:26
I never like when like the menus kind of like go go away like
22:29
in between.
22:30
Like there's a gap and then something goes away.
22:32
What's happening on the right side of a couple of those?
22:35
There were two of those
22:35
and there's there's kind of a grid on the right side,
22:37
yeah, with an icon.
22:39
Is that interactive or anything or is that Nope,
22:42
it's just something the LM thought was cute.
22:45
How about if you hover over on?
22:46
Yeah, okay, so it's just giving a different icon.
22:48
But icon.
22:49
And the icons are actually kind of cute.
22:51
Um, they seem somewhat original and on brand here.
22:56
Um, I would just never thought to to do this little little uh,
23:00
like if if you want to make a point
23:02
that on board is like four circles in the square arranged in
23:06
that way,
23:07
like if that was important, which it doesn't seem like it,
23:10
like you'll probably just put the icon here.
23:11
Kind of back to the like why reveal stuff
23:14
as you hover that is essential,
23:15
but it's probably non-essential, so you can probably just like leave it out.
23:21
What happens when we go further down?
23:24
Valued and trusted by Okay, we got a bunch of lines going everywhere.
23:31
All right, so we've got that line following you down the page pattern again.
23:34
>> And now we're scroll checked.
23:35
We're locked into this uh position here of the website in order to build
23:40
up this animation.
23:42
Um and I wonder what it wants to tell me.
23:46
Like why is it important to capture me here on this point to like
23:50
build out this animation versus just like showing it already built in the in
23:53
the built-out state.
23:55
>> And I find the animation is getting all of my attention rather than
23:58
what it says all over here on the left side.
24:01
So, I'm not even noticing any of this,
24:03
and this is not On the right side,
24:05
it's not giving me enough visual information to communicate something valuable about what they
24:10
do or how the product works.
24:11
So, I just kind of miss everything over here on the side.
24:13
The animation is too distracting.
24:15
Yep.
24:20
Outcomes.
24:20
>> It's also with these with these landing pages they kind of like build
24:23
up and then keep you in a in a section for a little
24:26
while.
24:26
It's really hard to know
24:27
when you're like how far into the >> book you you have read.
24:31
>> Yeah.
24:31
Right?
24:32
Like if you have a physical book, you're kind of like, "Oh, halfway through."
24:35
Here it's kind of like, I don't know.
24:37
Maybe it feels like we're halfway through, but maybe, you know,
24:39
it's going to keep me for another 15 seconds on some some like really
24:43
rich build-out um of a of a landing page.
24:46
So, the scroll indicator um uh is is a really important tool, I think,
24:51
kind of like to get uh a grasp on sort of like the you
24:55
know,
24:55
the landing page.
24:56
And here it's it's um, it's completely failing me.
24:59
Kind of like, where am I in the website?
25:02
All right.
25:02
Overall, like, visually I I think it looks nice.
25:05
Um, it looks modern.
25:07
There's a theme that they're, you know, you've got the orange accent color.
25:10
You've got the black as kind of your primary brand color.
25:14
Um, you've got the line that's kind of following you down.
25:16
And And my hunch is that that is uh,
25:19
sort of reproducing what the product looks like.
25:22
Cuz we keep seeing all these workflows with orange lines in the product itself.
25:25
And I think that's what they're trying to communicate here.
25:28
Um, and so, you know, that makes sense to me.
25:30
Their headline up at the top, I think uh,
25:33
does a great job of explaining what it is and who it's for.
25:36
Um, so I think they really nailed that.
25:37
Awesome, Sphinx.
25:38
Great work.
25:39
Next up we got Build Zero.
25:41
Let's take a look here.
25:42
Uh, build custom internal apps with AI.
25:45
Um, so right off the bat, you got some purple gradients.
25:48
Yeah.
25:48
There There There we go again.
25:50
Um, and uh, let's see here.
25:52
So, the menu nicely morphs.
25:54
Ooh, and then my pet peeve.
25:56
Um, dumb hover effects.
25:59
Um, it just doesn't add anything.
26:00
This vertical movement is just distracting and not cool.
26:05
Uh, it it pushes also.
26:06
There's another thing where, um, because of the effect,
26:09
it kind of like horizontally shifts the whole menu,
26:12
which kind of just seems like a bug.
26:14
Um, and then here's another like effect.
26:17
Um, kind of like, the arrow has to move somehow vertically.
26:21
Why?
26:22
Um, the arrow at least should go to the right also.
26:25
Yeah.
26:25
Yeah.
26:25
Yeah, it goes like backwards.
26:27
Um, this hover effect here I don't mind.
26:29
You know, it's like it's a black
26:30
and white button and
26:31
then it turns into the primary brand color.
26:33
In this case, kind of purple.
26:34
Like, that's not a bad hover effect here.
26:37
This here pretty bad.
26:38
Um, then what else have we got?
26:41
Uh, we've got a little like interactive element here.
26:44
Can click through it.
26:45
Ooh.
26:46
Okay, there seems to be some like bug with the selection here.
26:49
Yeah.
26:49
Um, if this was hand-coded, like,
26:52
no way someone would have not noticed it.
26:54
I almost wonder like if you just one-shot landing pages,
26:57
um do you actually go
26:59
and really like use it with the same attention detail
27:03
as if you if you had like built it um from scratch.
27:06
And then because you don't,
27:08
um you just ship uh things
27:10
that are that you wouldn't have um
27:13
because you don't even notice the the little bugs
27:15
that the LM made.
27:16
>> Um go back to the engineers tab here.
27:19
Now, if this is their product,
27:21
one of the other things
27:21
that stands out to me is this kind of dashboard um that's got,
27:25
you know, it's got like the red, green, blue,
27:29
purple kind of callouts up here.
27:31
That's one of the hallmark classic things
27:33
that uh a lot of uh AI design tools will will actually create.
27:37
Um Yeah, every fake dashboard looks basically like like something like that.
27:42
They've got the icons that are a darker version of the lighter background color.
27:46
Um it's it's usually like the Google colors, you know, like red, yellow, green,
27:51
blue. >> primary, yeah.
27:54
>> Yep.
27:54
So that that that we tend to see a lot.
27:57
Um ideas to working apps in minutes,
28:00
transform efficiency. >> these like kind of like bento bento boxes um is is
28:05
not a bad pattern,
28:07
um but it's also very non-original.
28:10
Um and so this LMs oftentimes kind of come up with these.
28:13
Um just like icon at the top uh
28:15
and then a bit of text
28:17
and you know,
28:18
uh in this case kind of like a 3x2.
28:21
Um that that's yeah, quite common.
28:25
>> Yeah.
28:25
Uh we've got some quotes.
28:27
We got the purple gradient.
28:30
We've got some frequently asked questions.
28:33
And then more gradients.
28:34
Yeah.
28:35
You know, we're I don't know,
28:36
a few into reviewing these now
28:39
and and I think we just keep seeing a lot of these common patterns
28:42
um which I think just highlights how ubiquitous they've become
28:45
and how um it it almost like takes away from the brand
28:50
and originality that I think most of these companies want to create um
28:54
when it it has elements
28:55
that every other site
28:56
and every other product has,
28:57
too.
28:58
Yeah, look.
28:58
I think the really cool thing about AI tools is
29:01
that it kind of frees you from having to kind of like,
29:04
you know, fiddle with like the technical details.
29:08
So, you can really work on the hard hard kind of questions of your
29:11
offering,
29:11
of your product, of your company, such as like, what are we making?
29:15
For whom?
29:15
Why is this valuable to them, right?
29:17
And that's the cool aspect.
29:19
But, if you then kind of like just use it um to create uh
29:22
what we sometimes call like AI AI slop,
29:25
I think this idea of having sort of like, you know,
29:27
all this power that you should use responsibly >> Yeah. um gets a little
29:31
bit wasted.
29:32
And so, that's the fine line to walk, I feel,
29:34
for for builders in 2026.
29:36
Yeah, there's there's nothing inherently bad about this, you know?
29:39
It's just that we've been through a few and we've seen this before,
29:42
you know?
29:42
That that's kind of what it feels like is I've seen this before.
29:45
And for any anybody else in the audience, like,
29:48
you lose a little bit of credibility, I think, you know,
29:50
when your customers are are viewing your product and they're saying like,
29:54
"Uh this looks like a bunch of other things I've seen.
29:56
They probably just used AI to build this and design it.
29:58
If that's the case, like, how good is the actual product, you know?"
30:02
Um or are they vibe coding that, too?
30:04
>> Use AI to build this by any means, um but just kind of like,
30:07
you know, really then evaluate the output
30:10
and uh use AI to make it even better
30:12
and more original and spend the time
30:14
that you saved one shotting it
30:16
or whatever,
30:17
um to then really think about your messaging and um and, you know,
30:21
designing a wonderful product that you can then showcase, etc.
30:23
Yep, that's great advice.
30:25
Awesome.
30:25
Build zero, thank you.
30:26
>> Thank you.
30:27
All right, next up we got Zarna, AI Associates for private capital markets.
30:32
Uh we're scroll jacked again.
30:35
Oh, yeah.
30:38
It's 10x everything, all the things.
30:40
Unlocked IRR.
30:42
This This site it like I'm in molasses.
30:45
It just feels so clunky to use.
30:47
Not to mention there just isn't a lot of like information.
30:52
I feel like I have like even this is deal volume 10x like Yeah,
30:57
this all everything 10x just feels made up
31:00
and it feels like there's just
31:01
so much space.
31:02
It feels like I have to scroll through this molasses forever to get to
31:05
anything that helps me understand what this is. >> of the things I personally
31:08
like if we go to the top right now the sort of background here
31:13
uses 100% of the vertical space
31:16
which kind of like almost like this visually kind of locks me in
31:20
and then there's like nothing more
31:23
and I hate when
31:24
when people kind of like put like scroll down to see more
31:26
because like of course we're on a website.
31:28
Of course I would know how to scroll down.
31:30
So one of the things
31:30
that I personally kind of like to interrupt the visual a little bit is
31:35
kind of like to start a landing page like maybe like like this,
31:38
you know, now I have kind of like something here at the top
31:41
but I know there's like more good stuff down here.
31:44
The the good stuff better be good
31:45
because close better deals faster doesn't you know tell me much
31:48
but maybe there's like you know like 10 pixels of like an awesome hero
31:52
image kind of like peaking up
31:53
or something like that.
31:55
That would be cool
31:56
and then I immediately without any indicators like no one wants to kind of
31:59
like scroll down and discover it,
32:00
right?
32:01
Yeah.
32:01
Well, we've stopped at the perfect point here
32:03
because you cannot see anything in the top navigation bar.
32:06
Yep.
32:07
It's you better know what's in the background
32:10
and if there's a dynamic background such
32:11
as a video playing
32:12
which might have all sorts of bright colors.
32:14
All of a sudden your bright menu isn't readable anymore.
32:18
>> Okay, so this is interesting.
32:19
>> So here we picked it up.
32:20
So it's smart but it's not smart when there's video behind it.
32:23
>> Yeah, you're right.
32:24
That is the the nuance.
32:30
Huh.
32:30
Okay, these I can't click on.
32:32
I would have expected that I could click on any of these.
32:35
Maybe the chevron.
32:38
Oh, I don't know what just happened.
32:43
Okay, now I can click on those.
32:45
Yeah, some of it is clickable and some of it is not, it appears.
32:48
And also it's like jumping me up and down on the screen.
32:54
Um it seems to just be moving on its own timer
32:56
and not letting me do anything,
32:57
which is confusing.
32:58
Um and again, this is one of those things where it feels like you
33:01
wouldn't intentionally design it to work this way.
33:04
Instead, you just It created a component
33:07
and you add And it doesn't happen
33:08
and it does sometimes work.
33:10
>> Only on those nested ones.
33:11
If you try the other ones, it doesn't do anything.
33:14
Yeah, and this better label here is um super super hidden.
33:18
There's sometimes a chevron, sometimes not.
33:20
Um yeah, this feels maybe like a little bit sloppy.
33:26
>> Mhm.
33:26
>> >> Yeah, one agentic layer generating an investment memo.
33:32
Um yeah, even the little bits of motion here is the kind of thing
33:35
that like AI will just give you
33:37
that for free and you're like,
33:38
yeah, okay, let's go with that, but you would never create yourself.
33:42
Um yeah, I can't even tell what I'm supposed to do with any of
33:45
this.
33:45
It hovers and expands a little bit, but like nothing is clickable.
33:51
Uh we got some fade-ins.
33:54
This just feels like it's lacking a lot of like useful information presented in
33:58
the right ways.
34:01
Um and at least not purple, though.
34:03
Got a lot of blue.
34:04
Yeah.
34:05
Yeah.
34:05
Yeah, fit and finish as well as kind of like meaty content.
34:09
>> Yeah.
34:09
Um like clear messaging is maybe lacking a little bit.
34:13
Um the visual style, if I didn't have to scroll through it,
34:16
um is actually kind of fresh.
34:19
>> Yeah.
34:19
Yeah, like it it reminds me of uh lovable
34:21
or or things like
34:22
that that kind of have
34:23
that like grainy zoomed-in kind of picture in the background. >> not sure what
34:26
these boxes um here represent uh in the background.
34:31
Yeah, some weird artifact.
34:33
All right, Zarna.
34:34
Thank you.
34:35
Okay, well, thank you everybody for submitting your sites.
34:37
And Raphael, we've seen a bunch of them now.
34:39
What are some of your takeaways here?
34:41
>> Yeah, I think definitely lots of purple gradients.
34:44
And in general, kind of like a lot of kind of like the same
34:47
the same patterns.
34:48
You know, we see animations kind of like for animation's sake just
34:50
because it's it's possible now.
34:52
I would have killed to have these kind of like AI superpowers
34:56
when I built my first website.
34:57
So, I I think it's wonderful that we have them.
34:59
But you are still kind of like responsible to not outsource your thinking to
35:03
LLMs.
35:04
And actually just use them as tools to get your brilliant ideas, designs, novel,
35:11
maybe kind of like interactions out on the web to dazzle your customers.
35:16
And I think that's what a landing page should do in the end.
35:18
It's kind of like a customer acquisition channel, right?
35:20
This is not your product.
35:21
This is not a website
35:23
as in kind of like a cool thing on the web.
35:26
There's space for that as well, but these are like startup landing pages.
35:31
And they're basically acquisition channels.
35:33
So, that's how I would really use these tools.
35:36
I totally agree with all of those.
35:37
Couple others I would add.
35:39
You know, one I think it's really important that you QA everything.
35:42
It's very easy to create something, but we went through and we found,
35:45
you know, some quirks or bugs or things like that here or there,
35:48
things that were confusing.
35:49
Ultimately, you know, you the founder,
35:51
you the human need to be the one
35:53
that goes through and make sure
35:54
that you know,
35:55
the interactions and everything else work right.
35:58
The other one is um you know,
36:01
it's important that your site and your brand,
36:05
they come through as the thing
36:07
that you want to represent your company
36:08
and that originality and something
36:11
that feels different.
36:12
And it's very easy using the same tools
36:14
that everybody else is using to end up in the same place.
36:17
You have to be intentional about ending up in a different place.
36:21
And so, rather than start from what does the AI spit out
36:26
and then let's tweak
36:27
that.
36:27
I would start from, you know, what color palette do we want to use?
36:31
And what is our brand?
36:32
And and how can we feed
36:33
that into the system
36:35
and use that to make sure we end up at an end goal
36:37
that represents us,
36:39
rather than just going with, you know, whatever the AI spits out there.
36:43
We've seen a lot of people do that, and I think, you know,
36:45
like you said, one of the most powerful tools ever to create incredible designs
36:50
and do things that would have been impossible
36:52
or taken forever that you can do just like this.
36:55
And so, to use
36:56
that for ways that
36:58
that actually help you move the lever upward
37:01
and make the sites more engaging
37:03
and convert better is one of the most powerful things
37:05
that you can do.
37:06
Yeah.
37:07
Well, thank you so much for joining, Raphael.
37:09
It's been amazing to get your perspective on all of these,
37:12
and thank you to everybody that submitted websites.
37:14
Thank you.
37:15
And we will see you on the next design review.
Thích
Chia sẻ
Y Combinator
Xem tất cả →
C1
Kinh doanh
Karaoke
22:36
OpenClaw Creator: Why 80% Of Apps Will Disappear
Y Combinator
5
C1
Kinh doanh
Karaoke
7:51
The New Way To Build A Startup
Y Combinator
C1
Kinh doanh
Karaoke
13:07
Inside The Startup Reinventing The $6 Trillion Chemical Manufacturing Industry
Y Combinator
C1
Kinh doanh
Karaoke
40:57
Demis Hassabis: Agents, AGI & The Next Big Scientific Breakthrough
Y Combinator
C1
Kinh doanh
Karaoke
10:27
How To Build A Company With AI From The Ground Up
Y Combinator
C1
Kinh doanh
Karaoke
53:20
The Future Of Brain-Computer Interfaces
Y Combinator
C1
Kinh doanh
Karaoke
50:10
Inside Claude Code With Its Creator Boris Cherny
Y Combinator
C1
Kinh doanh
Karaoke
21:49
How to Make Claude Code Your AI Engineering Team
Y Combinator
1
Video gợi ý
C1
Kinh doanh
Karaoke
5:49
You Need to Be Bored. Here's Why.
Harvard Business Review
7
C1
Kinh doanh
Karaoke
4:28
Lean Into Imposter Syndrome, Don't Give In to It
Harvard Business Review
C1
Kinh doanh
6:30
Identity Crisis: Why Defining Yourself by Your Career Is a Problem
Harvard Business Review
C1
Kinh doanh
Karaoke
22:36
OpenClaw Creator: Why 80% Of Apps Will Disappear
Y Combinator
5
C1
Kinh doanh
Karaoke
6:36
Fighting Workaholism: You Are Not a Success Machine
Harvard Business Review
C1
Kinh doanh
Karaoke
2:47
Can Work Make You Happy? Should It?
Harvard Business Review
C1
Kinh doanh
Karaoke
7:51
The New Way To Build A Startup
Y Combinator
C1
Kinh doanh
Karaoke
13:07
Inside The Startup Reinventing The $6 Trillion Chemical Manufacturing Industry
Y Combinator