The End of JS
256 segments
One of my favorite conference talks of
all time is the birth and death of
JavaScript. Now, you're probably
thinking, "Okay, the death of
JavaScript, you you have my attention.
What are we talking about here?" Well,
before you get too excited, it's
actually a talk from 2014 by Gary
Burnernhard while he was at Destroy All
Software. By the way, great name. In
today's day and age, I mean, I guess in
some sense, we've really lived up to
that name of destroying all software by
producing unceasing amounts of it. What
makes this talk so good is that he
attempts to predict the future 20 years
in advance. A lot of it was a lot of
good jokes. It's one of the best
conference talks ever created, but it
also has some pretty interesting ideas
in it. And one of the best parts is our
year right now 2026 is actually one of
the pivotal moments inside the
prediction. The beginning of the
conference talks does go through some of
the old lore of JavaScript, the 10day
creation that Brendan Ike was able to
somehow birth this language into the
world, the pain of actually using it,
you know, some of the classic fun ones.
So his entire prediction actually relies
on a technology called Azam.js. Now this
may be a technology before your time,
maybe you were before its time, but
effectively it is supposed to be the
assembly for the web. Now this should
sound familiar because this is actually
what Wom is based off of. Effectively
Azom is just JavaScript except for every
single kind of operation you need to
enforce its type. So right here you can
see that square this function square
takes in X. X equals the plus sign then
X. That is a prefix operator that
converts whatever X is into a number.
And then you return X * X and then you
reconvert that into a number. diagonal.
Same thing. X equals number of X. Y
equals number of Y. We're going to
square X, square Y, which again renumbs
X and renumbers Y. And then we're going
to squirt it out with a number at the
end. That's right. Squirt it out, baby.
You're probably thinking, okay, why
would we do this? This looks horribly
inefficient. Well, the idea is that the
JIT should be able to read all this
stuff in and go, okay, we are doing a
number operation. So square can simply
become a mole operation, right? That
means it can just just be a single
instruction like mole register whatever
it and itself and boom you got a number
back out or at least that's the idea.
Now I do actually remember this time I
was in the valley during this kind of
big era of AOM and I actually
specifically remember Epic Games and
Unreal Tournament. This is the Unreal
Engine 3 game engine which was used for
high-end games in the early to mid
teens. And this is a a short screencast
of it running inside of Firefox. Uh the
last demo, the Python demo is running
inside of Chrome, which does not support
AOM natively. That was actually running
as pure Java uh JavaScript. And uh here
you see a a video game engine running at
a perfectly playable frame rate. Not 60
frames per second, but playable. First
off, a couple things. Perfectly
playable, not 60. I mean, even 60 frames
per second. Like, are we even are we
even doing that these days?
Ridiculous. Can you believe what people
put up with in like 2013? This is what
big tech did to tech. Okay, this is what
they did. They took something that could
run better and then made it objectively
worse and been like, dude, this this is
the future. This is what you get to look
forward to. But it was exciting.
Honestly, it was exciting because this
meant there is a possibility of us not
having to write JavaScript. You could
write any language you want and you
could compile it down and actually
target it and run it in the browser. Now
granted, what came out of these things
were literally three megabyte bundles.
And you know, 15 years ago, the internet
was certainly not that great to be able
to have every single website delivering
you many, many megabytes of just
JavaScript. Let alone were the computers
able to run that. The, you know, the VMs
inside of V8 were not that good as they
are today. There were a lot of problems
but nonetheless it worked. It made
people very very excited. So this is
kind of like the crux of Gary's talk
which is okay there's this technology
which you don't have to write JavaScript
anymore but it's powered by JavaScript.
JavaScript will kill JavaScript via AOM.
Obvious side note there was no such
thing as WOM at this point. So he he
based it purely off of AOM. And just
kind of prove the point Gary does flex
on us a little bit. So here's uh the
running inside Chrome with Chrome
inside Firefox.
It's just compilers. Once you have a
compiler, all you need is the the the
libraries being used. Uh so here what we
have is the running against Wine
running against an X Windows shim all
compiled from C into AOM. That AOM is
running on Chrome. And now we have
another wine-like problem, which is that
was the Mac version of Chrome running
inside the Mac version of Firefox, but
the Mac windowing stuff is all closed
source. So we had to do what we did in
the '90s and reimplement it. And because
we like confusing names, the open-
source reimplementation of Coco is
called Cacao.
And then uh Cacao and Chrome are both
compiled from C into Azim. And all of
that is running inside of Firefox. now.
And so really he just proved you can run
anything. If it compiles in C, baby, you
can run it in the browser. And this is
kind of his fundamental idea that he has
for our future, which is that everything
can just simply be compiled into this
Azim and it can just run in the browser.
But he actually ends up making a
significantly bigger prediction cuz this
in itself shouldn't really kill
JavaScript. So this is where it actually
gets interesting cuz this is where he
makes his first big prediction that is
during our time period which is the
first major adoption of AOM.
>> So uh that takes us to 2025 when you
start to see thick extremely large
applications ported uh into AOM.
>> Now now is this prediction correct?
Well, I think one thing that probably
threw off Gary's old prediction was AI,
right? Because during 2020 to 2025 in
his chart, uh, there was just a massive
war. You know, kind of funny. We had CO
during that time. He predicted massive
war. He was just wrong. You know,
pandemic, not war. But either way, his
prediction did not include this idea of
AI and generating code. And so, he
thought that people would just get so
frustrated with JavaScript that
everybody would just want to use AOM.
And he's not completely wrong because
the successor to Azom Wom is now in 3.0
stage which does happen to have garbage
collection, better exception handling,
JavaScript string built-ins, custom text
format annotation, deterministic
profiling, like a bunch of stuff that is
actually making it pretty good to use.
Now, my last time I used WOM, I I could
easily say I did not like it that much,
but now, okay, this is looking a bit
more interesting. Is there a world in
which you could you could use WOM? Well,
it actually turns out that Wom is
actually making inroads into the old web
development. Cloudflare workers can use
Wom and that means you could develop
workers using Rust or C++ or Go or
Python. Not really sure about that last
one, but you could you could use any of
them because they all can compile down
to Wom. So, it's actually kind of
happening. There is this small world
where WOM has seen a small uptick in
adoption. There's been a couple notable
apps that have used WOM. Figma being one
of them. So though this prediction is
really I mean ultimately I think we can
all agree is wrong. It it's not like
crazy far off. But it's his next
prediction that it feels unbelievable.
So his final stage prediction is
actually called metal where instead of
having like your operating system and
these calls between ring zero and ring 3
and having to have a bunch of overhead
from having indirection of memory
instead you just put a JavaScript VM in
the kernel and everything just has VM
isolation and everything runs AOM. Every
program you compile it's AOM. Every
dream you ever had, Azam. That girl you
were talking to earlier, that's Azam.
Now, he does claim that it it's hard to
sell this idea in 2014. While probably
still kind of hard to sell this idea in
2026, but hey, there's hosting companies
out there that are fully bought into
this. It does seem like WOM is making
some forward step. Maybe Gary isn't that
wrong. Maybe, I know this sounds crazy,
maybe he could actually be right. And if
this doesn't happen, I'm just going to
have to blame AI. You know what I mean?
Like they they ruined it. We could have
had we could have had JavaScript in the
kernel if it wasn't for AI. Thanks AI
for ruining everybody's ability to
innovate. Anyways, I I wanted to go over
this talk because this is this was one
of the most legendary talks of all time.
I highly recommend going and watching
it. I'll link it in the description.
Hey, the name is Let me show you my ASOM
face, baby. A jen. I forgot the Ajet at
the end, so it kind of kind of ruins it.
That That might have just simply been
harassment again. Hey, do you want to
learn how to code? Do you want to become
a better back-end engineer? Well, you
got to check out boot.dev. Now, I
personally have made a couple courses
from them. I have live walkthroughs free
available on YouTube of the whole
course. Everything on boot.dev you can
go through for free. But if you want the
gamified experience, the tracking of
your learning and all that, then you got
to pay up the money. But hey, go check
them out. It's awesome. Many content
creators you know and you like make
courses there. boot.dev/prime for 25%
off.
Ask follow-up questions or revisit key timestamps.
The video discusses Gary Bernhard's 2014 conference talk, "The Birth and Death of JavaScript," which attempted to predict the future of web development 20 years in advance. Bernhard's core prediction revolved around Azam.js (a precursor to WebAssembly or Wasm) enabling any language to compile and run in the browser, thereby diminishing the direct use of JavaScript. The speaker reviews these predictions, noting that while the first major adoption prediction for 2025 wasn't entirely accurate, partly due to the rise of AI, Wasm is indeed gaining traction in areas like Cloudflare workers and applications like Figma. The talk also touches on Bernhard's more radical "metal" prediction of a JavaScript VM in the OS kernel. The speaker concludes that despite its bold nature, Bernhard's vision might not be entirely wrong, humorously blaming AI if it doesn't fully materialize.
Videos recently processed by our community