Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Welcome back to the deep dive. Today. We're really getting
into something fascinating anything. It's this AI powered app builder,
and honestly, it's making some pretty bold claims, redefining how
products get built, all without touching code. So yeah, our
mission here is to unpack all that. What are the claims?
What can it actually do? We've looked at what the
(00:20):
creators are saying and importantly dug through a ton of
user comments on product hunt. Real experiences will cover what
it is, how it works, who it might be good for,
and yeah, what people are saying.
Speaker 2 (00:30):
And it's quite something to think about, isn't it. The
ambition We're not just talking faster apps. This is about
like closing that huge gap between just an idea and
a real working product, super fast, super easy. If they
pull it off, well, it could genuinely democratize software creation.
Open it up.
Speaker 1 (00:46):
Okay, so let's unpack this. Where did anything even come from?
Well it actually grew out of an older AI app
builder called Create. Apparently Create got something like four hundred
and fifty thousand users over a couple of years, pretty popular.
But then just a few months back, the creators drove
AMN and Marcus Lowe. They said they had this breakthrough
and that led them to basically rebuild the whole thing
(01:09):
from scratch, centered around what they call a radically new
AI coding agent. Right, so this new version anything, it
has this core promise the world's best agent for making
products without coding. The big idea taking you from prompt
to production. Just imagine typing what you need and out
pops a real application.
Speaker 2 (01:27):
That prompt a production idea. I mean, that's the real
shift here. It's a huge leap, not just from you know,
writing code yourself, but even from the older no code tools.
It really puts the AI agent front and center handling
all that complex stuff underneath. Think about it. The cost
to just test an idea drops massively, which could mean well,
an explosion of niche apps maybe.
Speaker 1 (01:46):
So what can it actually do? Let's talk features. It
ships both mobile apps like for your phone and regular
web apps. But the big differentiator, it seems, is that
everything is built in and they really and the AI
agent at sounds pretty advanced. Sheriff Kurraganti an early user.
He said it can write huge chunks of code, make
(02:07):
updates on the fly, refactor intelligently fix its own bugs.
And even search the web when it needs help. That
last one is wild. He basically said, it gets rid
of that whole headache of stitching together ten different.
Speaker 2 (02:18):
Tools A boys the duct tape exactly.
Speaker 1 (02:21):
And then there's design. That's often a weakness with generated stuff, right,
but they're aiming for designs that don't look aima designs
that feel human made, and users seem to agree. Jesse
Bray called it the sexiest app building app. He said
how it feels is often just as important as how
it works, which is true. Another user, yb just said
the designs were a different league. Hi, praise. Okay, so
(02:42):
you build your app, it looks good. What about getting
it out there? They talk about one click deployment, the
app store, just hosting it on the web. Tajus Bocta
called that feature insane cool and we're already seeing people
like Nieherrika Hemmer, Johnny getting ready to submit their first
step built with anything to Apple.
Speaker 2 (02:58):
That's wow. That brings up a key point though, the
technical depth is it just a black box? Alan van
Toy actually asked about source code access and the answer
was a clear yes, Oh yes, you can edit or
download the source code full access. You could even pair
program with the agent, and the back end isn't just
simple stuff. It handles things like LM processing, using those
big language models to understand your prompts, and setting up
(03:20):
APIs so you can build proper full stack apps with
complex logic behind the scenes. This mix of no code
simplicity but with code access that can empower really wide
range of people. Couldn't it from founders to devs.
Speaker 1 (03:33):
Just wanted to go faster, right, And they even let
you choose the AI model powering it. You can switch
between like anthropics Sonnet four or open ais GPT five.
Apparently they're seeing better performance with Sonnet right now, but
having the choice is interesting. Oh, features sound impressive, but
the real proof is you know what happens when people
actually use it, and the stories are pretty remarkable. Take
(03:55):
Lindy Ryder, she's a project manager okay, not a coder,
first time building an app. She built something in a
few hours, something she estimated would have taken her dev
team a few weeks, called it fun to use, and
highly recommended it. That's a huge difference. And it's not
just non coders having these moments, developers too. Mihi Crasan
a developer literally said, building with Anything is so fast
(04:16):
that I am genuinely scared for my job as a developer. Wow.
He gave an example building a business card app and
Anything automatically added support for dot PCF files, the standard
contact card format. He didn't even ask for it. The
creators say, actually a lot of dems are using it
purely for that speed advantage.
Speaker 2 (04:31):
And what jumps out from all these stories is how
much less hassle it seems to be. Evan Marshall put
it perfectly, least headache I've ever had working with any
no code product. That says a lot. Andrew Jamison talked
about solving the apikey hell Er duct taping tools problem
that resonates. He was also blown away by the AI.
The frontier agent cranking out one hundred k plus lines
(04:51):
autonomously called it absolute sorcery. So it's not just speed,
it's removing the frustrating roadblocks absolutely.
Speaker 1 (04:58):
And some examples are just staggering. Even Axel built two
fully functional, production ready web apps in a few days.
He estimated the savings at over twenty thousand dollars compared
to traditional development. Called anything in a whole nother league.
Speaker 2 (05:12):
Then there's usman usef built an app for a specific need,
ready to use, perfectly serves the exact purpose in just
two hours, two hours with no breakdowns or bugs. He said,
it does sound a bit like magic.
Speaker 1 (05:23):
It really does stretch the imagination. And there are more.
Toby built this massive app for the arts and entertainment industry, scheduling, budgeting,
equipment tracking, time sheets, even a subscription model, fifteen databases,
eight interconnected pages, cool tools, a dashboard built the core
in minutes. He even named his AI agent Max called
them besties. Shows a kind of connection forming there. Oh
(05:46):
and Alex Avery idea to functional app in two minutes,
just two minutes.
Speaker 2 (05:50):
So pulling this together, who is this actually for? Well,
the main target the ICP, it seems to be people
building real software products, things they want to launch, maybe
make money from. Often folks who are semi or non technical,
maybe deve adjacent like product managers, designers who understand software
but don't code day to day.
Speaker 1 (06:08):
Yeah, that sounds right, but it seems the appeal is
broader too. You mentioned product managers, designers, marketers and bigger
companies using it for quick prototypes. Maybe internal tools. Students
are using it for games, personal projects. The reach seems
pretty wide already, and their ambition it's huge. They want
Anything to be just the best way to create software,
(06:28):
period full stop. Zakari Sealifu called it an impressive leap
for no code.
Speaker 2 (06:34):
But you know, with something that promises to do everything,
you have to ask about the tradeoffs, right, like is
there a risk of vendor lock in? What if you
need something super custom the AI just can't do. Yes,
you get the source code, which is great, but relying
entirely on AI generated code for something really complex at scale,
maybe that brings its own challenges. Perhaps the developers job
shifts more towards like verifying AI code quality, integrating it
(06:57):
rather than writing it all from scratch.
Speaker 1 (06:59):
That's a really good point. It's not just about creation
but maintenance and scaling too, And the Anything team seems
aware they need to keep evolving. They're listening to feedback. YB,
for instance, love the designs, but worried about the credits
needed just to test properly, especially in a crowded market.
Suggested demos or clearer roadmaps makes sense. Other users ask
(07:21):
for practical things like drag and drop image uploads. Soup
Paul Signy mentioned that and more dev focused features too,
like cloning GitHub repos or running NPM commands. That suggests
developers want to integrate it even deeper into their workflows exactly.
Speaker 2 (07:35):
And that raises the question of how that user feedback
shapes what's next. Can they add those more complex developer
features without losing the core ease of use for everyone else?
It's a balancing act. Eli who won an honorable mention
in one of their hackazons, he noted things like faster performance,
get of integration that were probably coming, but he also
stressed that even now it works great out of the box,
(07:56):
so the foundation is solid.
Speaker 1 (07:58):
Okay, So let's wrap up this. Anything is clearly pushing
hard at the boundaries of what AI can do in development.
It's making building products much faster, much more accessible, really
blurring the lines between no code and code, and that
leaves us with a big question for you, the listener.
We're seeing people build complex apps, ship them to stores,
(08:18):
save tens of thousands of dollars already. How is this
kind of acceleration going to change things for individuals? For businesses?
What kind of magic maybe could you build if that
friction of coding was just gone.
Speaker 2 (08:31):
It's fascinating, isn't it. This idea of vibe coding, describing
what you want and having it appear. It might not
just make app building easier, It could fundamentally change who
gets to innovate and how fast they can do it.
It really points towards a future where maybe creating software
is less about knowing the right syntax and much more
about just having a clear vision the future, driven purely
by ideas.