WEBVTT

00:00:00.200 --> 00:00:04.170
So another year's almost over
and you know what that means,

00:00:04.180 --> 00:00:08.000
time for me to look back at it, and yeah,
obviously that video

00:00:08.039 --> 00:00:11.800
will be about AI. Not just about AI,

00:00:11.840 --> 00:00:15.440
but a lot of AI because, you know,
it's the thing that

00:00:15.540 --> 00:00:19.480
influences everything
and has an impact on everything,

00:00:19.500 --> 00:00:23.470
And believe me, I'm not a fan of having,
uh,

00:00:23.520 --> 00:00:27.110
AI everywhere. But it is what it is.
But yeah, let's take a step back.

00:00:27.140 --> 00:00:31.120
You, you might remember this year
was meant to be the year of

00:00:31.240 --> 00:00:35.140
AI agents. It was the term you heard
and read

00:00:35.220 --> 00:00:37.960
everywhere,
especially at the beginning of the year.

00:00:38.000 --> 00:00:41.470
You still, of course, hear
and read it everywhere,

00:00:41.540 --> 00:00:45.100
year started, everywhere on any-
every website you could read,

00:00:45.110 --> 00:00:49.040
"2025 is the year of AI agents."
And of course, an

00:00:49.140 --> 00:00:52.840
AI agent was basically anything that used

00:00:52.920 --> 00:00:56.780
AI for anything. You wrote a script
that uses

00:00:56.880 --> 00:01:00.540
AI to summarize a PDF? Congratulations,
you built an AI

00:01:00.640 --> 00:01:04.140
agent. Now, of course AI has made

00:01:04.580 --> 00:01:08.270
significant advances to some degree
when it comes to

00:01:08.320 --> 00:01:11.620
agentic workflows. Now, what
is an agentic workflow?

00:01:11.660 --> 00:01:15.170
Again, depending on how you define it,
it's anything that uses AI,

00:01:15.180 --> 00:01:19.060
but I would say it
is any kind of workflow where AI

00:01:19.100 --> 00:01:22.720
makes decisions. And keep in mind,
we're still talking about

00:01:22.800 --> 00:01:26.720
token generators,
but of course those tokens can describe

00:01:26.780 --> 00:01:30.510
tools the AI wants to use. And

00:01:30.510 --> 00:01:34.410
one big category of agents
that you definitely

00:01:34.460 --> 00:01:37.470
know is of course all these coding
assistants.

00:01:37.500 --> 00:01:41.420
I mean, they did not start this year,
you know that as well as I

00:01:41.440 --> 00:01:44.740
do. But in this year,
we saw lots of development on

00:01:44.780 --> 00:01:48.500
Cursor, we saw the release of Claude Code,
and all these

00:01:48.540 --> 00:01:52.120
tools are of course about giving us
developers

00:01:52.500 --> 00:01:56.000
assistants that help us write code,
that can

00:01:56.100 --> 00:01:59.880
generate code based on our instructions
or, if you're not

00:01:59.920 --> 00:02:03.520
that much into writing code,
they can of course generate

00:02:03.740 --> 00:02:07.680
code based on pretty vague instructions
and you can enter the

00:02:07.760 --> 00:02:11.140
territory of live coding. And live coding

00:02:11.640 --> 00:02:15.600
definitely was another big theme
and thing this

00:02:15.760 --> 00:02:19.320
year.
The term was coined by Andrej Karpathy

00:02:19.440 --> 00:02:23.180
in 2025 and as he described it, the idea

00:02:23.280 --> 00:02:27.040
really was that you don't really read the
code, that you

00:02:27.180 --> 00:02:31.170
go with the vibes,
you just basically see if it works

00:02:31.180 --> 00:02:35.120
or it doesn't and you just tell the AI
what to change or make it

00:02:35.200 --> 00:02:38.840
work and... Well, that
is how you can build software.

00:02:38.880 --> 00:02:42.609
Now, to be fair and very clear here,
he made it very

00:02:42.700 --> 00:02:46.280
clear that he still also writes code
himself.

00:02:46.300 --> 00:02:50.120
But of course,
you could use live coding for pieces of

00:02:50.180 --> 00:02:53.960
software that are maybe just a one-time
thing you need,

00:02:54.020 --> 00:02:57.870
maybe just a utility script you need,
or maybe if you're exploring some

00:02:57.920 --> 00:03:01.460
idea, building a prototype,
you could definitely

00:03:01.470 --> 00:03:05.320
try live coding for that. Now,
I have tried my fair

00:03:05.340 --> 00:03:09.180
share of live coding and I will say,
results have been rather mixed

00:03:09.200 --> 00:03:12.980
and it's definitely not my thing.
Wasn't my thing, still

00:03:13.160 --> 00:03:16.560
isn't. I will say though
that of course it can be quite

00:03:16.640 --> 00:03:20.580
helpful to let AI build a feature where
you maybe don't really know

00:03:20.640 --> 00:03:24.120
how to tackle it because you need to use
some browser

00:03:24.180 --> 00:03:27.980
APIs, for example,
which you don't know yet,

00:03:28.040 --> 00:03:31.980
help of the AI, which
is not the same as just letting the AI

00:03:32.000 --> 00:03:35.120
build something and then use it
and move on if it seems to work.

00:03:35.160 --> 00:03:38.960
That's a different thing. But yeah,
live coding was a

00:03:39.040 --> 00:03:41.680
huge thing and still is a huge thing.

00:03:41.720 --> 00:03:45.600
We have platforms like Lovable, v0,
and many others

00:03:45.700 --> 00:03:48.840
which are, in the end,
live coding platforms. Yeah.

00:03:49.700 --> 00:03:52.580
It's here, it will probably stay here.

00:03:52.640 --> 00:03:56.340
We saw lots of people experiment with it,
we definitely

00:03:56.400 --> 00:04:00.300
have live coders out there
that do live coding

00:04:00.320 --> 00:04:04.120
for a living, more or less I guess.
But yeah, let

00:04:04.160 --> 00:04:08.120
me be very clear here. Live coding
is not a

00:04:08.160 --> 00:04:11.800
good idea. Being an expert,
knowing how to write code, and

00:04:11.940 --> 00:04:15.870
using that expertise and
that knowledge in combination with AI is

00:04:15.940 --> 00:04:18.490
so much better than live coding. That's...

00:04:18.620 --> 00:04:22.060
I can only recommend taking this path
because your knowledge

00:04:22.320 --> 00:04:26.300
plus AI can get you really good results.

00:04:26.420 --> 00:04:29.760
Just relying on AI, probably not so much.

00:04:29.800 --> 00:04:33.330
Of course,
coding assistants like Cursor or Claude

00:04:33.440 --> 00:04:37.080
Code are not the only agentic systems we

00:04:37.140 --> 00:04:40.720
saw develop this year. OpenAI,
for example, also

00:04:40.780 --> 00:04:44.660
integrated an agent into ChatGPT
that can do stuff

00:04:44.740 --> 00:04:47.960
for you, that can click around on websites

00:04:48.040 --> 00:04:51.900
and do work for you. Now,

00:04:51.980 --> 00:04:55.860
maybe it's just my bubble,
but it doesn't really seem like a feature

00:04:55.920 --> 00:04:58.420
that everybody's using all the time.

00:04:58.440 --> 00:05:02.020
But yeah, it's here,
and combined with deep research

00:05:02.240 --> 00:05:06.110
agents,
which I think we also had last year

00:05:06.140 --> 00:05:09.900
which definitely became more of a thing
this year, you can

00:05:09.980 --> 00:05:13.810
offload some tasks,
especially some research tasks, to

00:05:13.860 --> 00:05:17.710
AI. Though as everything and always, it's

00:05:17.760 --> 00:05:21.520
typically not a super good idea to just
take those

00:05:21.580 --> 00:05:25.320
AI results and use them as they are,
but instead

00:05:25.340 --> 00:05:29.320
it's much better to use them as a starting
point and build

00:05:29.400 --> 00:05:33.320
up on them with your own knowledge,
maybe your own additional research,

00:05:33.360 --> 00:05:37.120
and then AI, in my experience,
can save you some

00:05:37.180 --> 00:05:40.660
time in certain situations.
And I definitely, for

00:05:40.720 --> 00:05:44.048
example, like using-...
Gemini Deep Research

00:05:44.568 --> 00:05:47.228
for doing research for a video like this.

00:05:47.328 --> 00:05:51.027
I still bring in my own expertise.
There are many points I

00:05:51.087 --> 00:05:54.808
cover in this episode which
were not brought up by

00:05:54.928 --> 00:05:58.527
Gemini, and also not by ChatGPT, but yeah,
it's a nice

00:05:58.647 --> 00:06:02.428
additional help. From that angle,
we could call this

00:06:02.488 --> 00:06:05.727
year the year of agents,
but to be very honest,

00:06:05.928 --> 00:06:09.888
not in the sense all these newspapers
meant

00:06:09.948 --> 00:06:13.558
at the beginning of the year.
It doesn't really seem like all those

00:06:13.587 --> 00:06:17.178
companies out there
are using agents for everything.

00:06:17.227 --> 00:06:20.288
Quite the opposite. There recently was an

00:06:20.368 --> 00:06:24.347
article about, uh,
Copilot by Microsoft not

00:06:24.368 --> 00:06:27.847
really delivering on its promise
or not being as

00:06:27.907 --> 00:06:31.788
popular as Microsoft would hope.
It really seems

00:06:31.808 --> 00:06:35.568
to be the case that companies want to use
AI, they

00:06:35.688 --> 00:06:39.308
all feel like they have to, of course,
and therefore, they are really

00:06:39.407 --> 00:06:43.328
searching for use cases for AI
and it's hard to find them.

00:06:43.388 --> 00:06:47.107
Coding and, um, AI-assisted coding
is definitely one of

00:06:47.168 --> 00:06:51.008
them,
but many others might not be there yet and

00:06:51.068 --> 00:06:55.048
we don't have all these agents
that do all the work humans used to

00:06:55.087 --> 00:06:59.068
do. And by the way,
when we talk about coding, I am well

00:06:59.128 --> 00:07:02.828
aware that in many companies out there,
there are executives that

00:07:02.868 --> 00:07:06.828
basically tell their teams,
"You either get double the work done

00:07:06.868 --> 00:07:10.738
more,
or we'll have to let some people go

00:07:10.768 --> 00:07:14.518
use AI." And that, of course,
is a horrible situation

00:07:14.548 --> 00:07:17.828
because whilst AI can be a very good

00:07:17.888 --> 00:07:21.868
assistant for coding, and I've said
that more than once, I say

00:07:21.888 --> 00:07:25.727
it in almost a- all my videos,
whilst it can be a good assistant,

00:07:25.788 --> 00:07:29.678
it's not a good replacement. And yeah,
it can make you more

00:07:29.688 --> 00:07:33.238
productive, but twice
or three times the output?

00:07:33.328 --> 00:07:37.238
I'm not so sure about that. But yeah,
it is

00:07:37.268 --> 00:07:41.107
what it is. That is something
which will probably get better as

00:07:41.168 --> 00:07:45.128
companies realize that even
when it comes to coding, AI is not the

00:07:45.227 --> 00:07:48.488
magic bullet,
but just another helpful tool.

00:07:48.568 --> 00:07:52.188
Or, of course,
all these AI models get much, much

00:07:52.207 --> 00:07:55.808
better, but until that point, it's a

00:07:55.847 --> 00:07:59.268
really, really bad situations. But yeah,
agents

00:07:59.368 --> 00:08:02.467
everywhere? I'm not so sure I don't see

00:08:02.487 --> 00:08:06.448
them. Now all that, of course,
does not mean that

00:08:06.508 --> 00:08:10.308
AI is not getting better and that there
is no chance

00:08:10.368 --> 00:08:13.707
of more agents doing useful stuff in the

00:08:13.748 --> 00:08:17.248
future. But when we look at this year, we

00:08:17.308 --> 00:08:21.087
definitely saw the release of many,
many AI models.

00:08:21.097 --> 00:08:24.918
(laughs) It, it feels like a new model
was released basically every

00:08:25.048 --> 00:08:28.707
week, right? Especially now in November,
we got like Gemini

00:08:29.147 --> 00:08:31.987
3, we got new models by OpenAI, by xAI.

00:08:32.008 --> 00:08:35.908
We get so many new models,
and not just in November.

00:08:35.948 --> 00:08:38.008
It's really model inflation out there.

00:08:38.048 --> 00:08:41.668
If you're a web developer like I am,
it feels like the

00:08:41.768 --> 00:08:45.688
JavaScript framework wars we had in 2019

00:08:45.768 --> 00:08:49.688
around that time. Uh,
but even worse because we didn't have as

00:08:49.708 --> 00:08:51.668
many frameworks as we have models now.

00:08:51.728 --> 00:08:55.238
Obviously, that will only continue,
but the question is

00:08:55.848 --> 00:08:58.788
will those models get better and better?

00:08:58.808 --> 00:09:02.647
To some extent,
to me it feels like they're not really

00:09:03.348 --> 00:09:06.568
I know they are continuously getting
better in all these

00:09:06.668 --> 00:09:10.568
benchmarks,
but we've seen more than once this year

00:09:10.668 --> 00:09:14.647
new model is released,
does very well in a benchmark, and then

00:09:14.728 --> 00:09:17.978
kind of falls flat in the popular

00:09:18.088 --> 00:09:21.858
opinion.
Isn't really living up to its promise and

00:09:21.887 --> 00:09:25.228
people are not really, uh,
happy about the model.

00:09:25.288 --> 00:09:29.248
We've seen that more than once.
One prominent example would be LLaMA

00:09:29.508 --> 00:09:33.448
4, which was released by Meta, uh, in,
in the first half of the

00:09:33.508 --> 00:09:37.348
year.
LLaMA 4 did amazing in all these

00:09:37.387 --> 00:09:40.568
course,
especially interesting because it's an

00:09:40.588 --> 00:09:44.428
But people didn't really like it.
It really felt like

00:09:44.488 --> 00:09:48.308
it was fine-tuned for these benchmarks as
it often feels these

00:09:48.328 --> 00:09:51.988
days with model releases and, well,
I'm not sure if

00:09:52.147 --> 00:09:55.858
anybody's still using LLaMA 4. Uh,
the result, of course

00:09:55.867 --> 00:09:59.647
(laughs) , of this, uh, failure,
if you wanna call it like this,

00:09:59.748 --> 00:10:03.228
was that Mark Zuckerberg started hiring

00:10:03.867 --> 00:10:07.748
all the talents in the AI space by paying

00:10:07.808 --> 00:10:11.488
astronomic amounts of money for the top
talent

00:10:11.828 --> 00:10:15.568
and we now have all
that talent war going on between OpenAI

00:10:15.647 --> 00:10:19.048
and Meta and Google
and bunch of other companies.

00:10:19.068 --> 00:10:22.008
But yeah,
regarding the raw performance of the

00:10:23.048 --> 00:10:26.627
we have yet to see
if we'll see a steep increase

00:10:26.768 --> 00:10:30.407
in the next year, 2026, or
if the curve kind of

00:10:30.468 --> 00:10:34.448
flattens out.
Benchmarks alone certainly do not tell

00:10:34.528 --> 00:10:38.367
the full story. Now, of course,
there's a lot of money being

00:10:38.407 --> 00:10:42.248
thrown on the development of all these
models and therefore, the year actually

00:10:42.367 --> 00:10:46.248
started with a shock when DeepSeek R1
was released

00:10:46.308 --> 00:10:50.048
because that was some Chinese underground
model

00:10:50.088 --> 00:10:53.748
which nobody expected, which at the time

00:10:53.768 --> 00:10:57.708
seemed to be trained very cheaply
and seemed to be able to

00:10:57.748 --> 00:11:00.477
compete with top-tier models back then.

00:11:00.488 --> 00:11:04.168
And of course now, at the end of the year,
we can say that all these

00:11:04.208 --> 00:11:07.968
Chinese models,
not just the DeepSeek models

00:11:08.048 --> 00:11:11.928
related to video or image generation,
play an important role

00:11:12.028 --> 00:11:16.018
in the AI landscape and China, despite all

00:11:16.048 --> 00:11:19.548
the problems they should in theory have
with getting graphic

00:11:19.608 --> 00:11:23.508
cards, GPUs,
plays a super important role and

00:11:23.568 --> 00:11:26.907
will be an important competitor in the
year to come.

00:11:26.928 --> 00:11:30.867
And on the less, as mentioned,
we'll have to see if AI

00:11:30.887 --> 00:11:34.348
models improve like this
or maybe a bit flatter

00:11:34.548 --> 00:11:38.428
like this. Of course,
large language models aren't everything,

00:11:38.588 --> 00:11:40.928
text generation isn't everything.

00:11:41.008 --> 00:11:44.867
We saw pretty amazing improvements this
year when it

00:11:44.948 --> 00:11:48.127
comes to image and video generation.

00:11:48.168 --> 00:11:51.708
I mean,
if you compare the kind of videos you can

00:11:52.028 --> 00:11:55.588
View 3 by Google or of Sora 2 by OpenAI

00:11:56.147 --> 00:11:59.428
with what was possible a year ago,
that's mind-blowing.

00:11:59.468 --> 00:12:03.448
It's so much better, so much, uh,
better quality, higher

00:12:03.568 --> 00:12:07.208
character consistency, audio is better,
and that's just the

00:12:07.308 --> 00:12:11.137
US models.
If you take a look at some Chinese models

00:12:11.208 --> 00:12:14.564
like the Kling models-Um, you can see some

00:12:14.574 --> 00:12:17.264
amazing development and results there too.

00:12:17.274 --> 00:12:19.624
And it's not just the case for video.

00:12:19.704 --> 00:12:23.544
Image generation was revolutionized first
by

00:12:23.584 --> 00:12:27.504
OpenAI's GPT image model, but then
that model

00:12:27.544 --> 00:12:31.344
now looks quite old compared to Google's
NanoBanana,

00:12:31.384 --> 00:12:35.364
and now its NanoBanana Pro model,
which excels at

00:12:35.424 --> 00:12:39.054
generating all kinds of images, including,
uh,

00:12:39.084 --> 00:12:43.064
images that include a lot of text like
this beautiful infographic

00:12:43.074 --> 00:12:47.024
here, which was created with, uh, my site,
buildmygraphic.com,

00:12:47.064 --> 00:12:50.504
which gives you a nice feature-rich editor
and beautiful

00:12:50.904 --> 00:12:53.224
style templates to build graphics like
this.

00:12:54.004 --> 00:12:57.404
So there is lots of development in the AI
space when it comes

00:12:57.464 --> 00:13:01.264
to other modalities other than text,
and of

00:13:01.324 --> 00:13:04.964
course, I'm also convinced
that we'll see more

00:13:05.164 --> 00:13:08.464
useful tools being developed on top of AI
models.

00:13:08.564 --> 00:13:12.534
Even if those AI models shouldn't become
much smarter, we will

00:13:12.534 --> 00:13:15.864
definitely see better
and better tools

00:13:15.904 --> 00:13:18.894
models and therefore deliver more value.

00:13:19.044 --> 00:13:22.684
Just like Cursor is not just some AI model
being

00:13:22.804 --> 00:13:26.714
used in its raw form,
but instead an AI model that is

00:13:26.744 --> 00:13:30.554
heavily adjusted by engineering the
context in a

00:13:30.564 --> 00:13:34.524
clever way, having good system prompts
and all that stuff that

00:13:34.604 --> 00:13:38.584
leads to better results than if you
were just to use

00:13:38.624 --> 00:13:42.144
the raw model through the OpenAI API
or whichever

00:13:42.154 --> 00:13:45.874
provider you are using. So that was a

00:13:45.984 --> 00:13:49.494
lot of talk about AI. I'm aware of that,
as I mentioned.

00:13:49.524 --> 00:13:53.464
I'm sorry.
But it is obviously an important part of

00:13:53.524 --> 00:13:55.524
we're living in if we like it or not.

00:13:55.544 --> 00:13:59.364
And believe me, I would be very happy
if we were not

00:13:59.464 --> 00:14:03.304
living in a world that's full of AI slop
and all

00:14:03.384 --> 00:14:07.284
that AI hype. But it is what it is.
And of course,

00:14:07.364 --> 00:14:11.004
I'm fully aware that this also was and is

00:14:11.104 --> 00:14:14.964
another very difficult year for software
developers,

00:14:15.024 --> 00:14:18.204
especially for web developers,
especially also for junior

00:14:18.304 --> 00:14:22.244
developers because we're coming still from

00:14:22.264 --> 00:14:25.853
that all-time high during the pandemic
where all the

00:14:25.924 --> 00:14:29.264
companies couldn't get enough developers.

00:14:29.284 --> 00:14:33.044
Well, turns out they had too many.
So you all probably

00:14:33.104 --> 00:14:36.384
know that chart about the software
developer jobs.

00:14:36.444 --> 00:14:40.084
And, well, at least
that seems to finally have reached the

00:14:40.104 --> 00:14:42.293
bottom, but now of course there is AI.

00:14:42.364 --> 00:14:46.064
And as I mentioned before, there
are these discussions

00:14:46.104 --> 00:14:49.884
or the pressure inside of companies about

00:14:49.924 --> 00:14:53.224
using AI instead of developers for certain
tasks.

00:14:53.233 --> 00:14:56.944
And especially junior developers,
therefore, are not in high demand because,

00:14:56.984 --> 00:15:00.824
hey, AI can do it, right? Wrong,
totally wrong.

00:15:00.834 --> 00:15:04.584
And who's going to be the senior of
tomorrow if there is no junior being

00:15:04.644 --> 00:15:08.124
hired today? But yeah, it is what it is,
and therefore,

00:15:08.164 --> 00:15:11.584
unfortunately, this
was another difficult year, and next year

00:15:11.664 --> 00:15:14.884
is probably going to be diff-
difficult too,

00:15:14.944 --> 00:15:18.864
which is really bad. There is no other

00:15:18.884 --> 00:15:22.594
way of putting that. Now despite AI,
thankfully,

00:15:22.624 --> 00:15:26.444
there are also some other interesting
topics to talk

00:15:26.524 --> 00:15:28.704
about when looking back at this year.

00:15:28.724 --> 00:15:32.384
For example, we, of course,
saw new versions being released for

00:15:32.464 --> 00:15:36.344
Next.js, React, Angular, and

00:15:36.444 --> 00:15:40.264
all those other frameworks
and library stats are important for

00:15:40.344 --> 00:15:44.244
us developers. It's also fair to say
that it

00:15:44.264 --> 00:15:48.124
looks like React and TypeScript
and Next.js won

00:15:48.624 --> 00:15:51.984
the AI race. I mean,
take a look at the React

00:15:51.994 --> 00:15:55.214
downloads. AI clearly has a favorite.

00:15:55.264 --> 00:15:59.013
And I've talked about this in other
episodes before,

00:15:59.084 --> 00:16:02.904
this is a huge problem in my opinion
because no matter if

00:16:02.964 --> 00:16:06.744
you like, um, the fact
that we had times where a

00:16:06.804 --> 00:16:10.644
new framework seemed to come out every day
or not, it's

00:16:10.654 --> 00:16:14.524
definitely not good to have one clear
winning stack,

00:16:14.564 --> 00:16:17.884
which is React, TypeScript, Next.js,
ChatGPT, and Tailwind.

00:16:17.944 --> 00:16:21.644
That's it. That is definitely not good,
and I

00:16:21.724 --> 00:16:24.924
don't see that changing, unfortunately.

00:16:24.964 --> 00:16:28.484
And that is not to say
that any of these technologies would be

00:16:28.504 --> 00:16:31.664
They're not. They're great.
Competition and

00:16:31.744 --> 00:16:35.324
innovation just also is good,
and it looks like that's going to get

00:16:35.464 --> 00:16:38.484
harder with AI. But hey, okay, th-
there was AI again.

00:16:38.504 --> 00:16:42.334
Let's, let's move away from AI. So yeah,
we saw these new versions bringing

00:16:42.384 --> 00:16:45.744
us new features, but especially React

00:16:46.704 --> 00:16:50.684
and to some degree Next.js, but it
was React, had

00:16:50.724 --> 00:16:53.324
some problems, some security problems.

00:16:53.364 --> 00:16:57.244
It was just a few weeks ago that a

00:16:57.284 --> 00:17:01.164
couple of vulnerabilities
were reported related to React

00:17:01.224 --> 00:17:05.184
and React server components.
I did create a separate

00:17:05.264 --> 00:17:08.954
episode about that, so take a look at
that if you wanna learn more.

00:17:08.954 --> 00:17:12.724
That, unfortunately,
wasn't the only security issue

00:17:12.844 --> 00:17:15.604
related to web development we saw this
year.

00:17:15.644 --> 00:17:19.464
Instead, we also had multiple NPM-related

00:17:19.983 --> 00:17:23.884
attacks, node modules-related attacks,
where

00:17:23.983 --> 00:17:27.614
some packages were infiltrated with

00:17:27.644 --> 00:17:31.164
malicious code in a way that
that code would

00:17:31.224 --> 00:17:34.624
spread automatically to other packages and

00:17:34.704 --> 00:17:38.224
ultimately potentially reach your machine
to then

00:17:38.464 --> 00:17:41.704
steal crypto or crypto

00:17:41.744 --> 00:17:45.684
credentials and stuff like
that from your machine.

00:17:45.724 --> 00:17:49.324
And of course, that all is probably

00:17:49.424 --> 00:17:52.944
just the beginning. I mean,
vulnerabilities and

00:17:53.024 --> 00:17:56.784
attacks have always been a thing,
but especially with

00:17:56.864 --> 00:18:00.804
AI, I see a huge danger of that trend

00:18:00.924 --> 00:18:02.694
accelerating for two main reasons.

00:18:02.704 --> 00:18:06.294
For one, AI can, of course,
help with developing

00:18:06.324 --> 00:18:09.414
malicious code.
It can help run such attacks.

00:18:09.424 --> 00:18:11.704
It can help with social engineering.

00:18:11.744 --> 00:18:15.464
I mean,
you can create realistic voices

00:18:15.564 --> 00:18:18.764
days.
So that will be a huge problem going

00:18:18.824 --> 00:18:22.192
think.But of course,
the other problem potentially

00:18:22.312 --> 00:18:26.002
is that AI can also lead to gaping

00:18:26.032 --> 00:18:28.851
security holes in software. And just to be

00:18:28.932 --> 00:18:32.351
clear, the React vulnerability

00:18:32.532 --> 00:18:36.392
and the NPM attacks are very

00:18:36.492 --> 00:18:40.252
likely not linked to AI in any way.
I still

00:18:40.472 --> 00:18:44.432
see these problems coming up in the future
for the two

00:18:44.472 --> 00:18:48.112
reasons I just described.
And we saw with the React

00:18:48.172 --> 00:18:51.272
attacks and with the NPM attacks how

00:18:51.331 --> 00:18:55.252
vulnerable this web development,
this front-end web development

00:18:55.792 --> 00:18:59.581
ecosystem really is.
And especially in a world where more

00:18:59.591 --> 00:19:03.472
and more applications
are powered by React, well, it's pretty

00:19:03.572 --> 00:19:06.712
clear that that is a quite interesting

00:19:06.812 --> 00:19:10.292
target for an attack and for looking for

00:19:10.372 --> 00:19:14.212
vulnerabilities. And the same, of course,
is true for NPM and Node modules

00:19:14.272 --> 00:19:17.752
because even we experienced

00:19:17.792 --> 00:19:21.192
developers often just install packages.

00:19:21.212 --> 00:19:24.812
And of course,
we trust many of those established

00:19:24.831 --> 00:19:28.012
Why wouldn't we? So even we are very

00:19:28.172 --> 00:19:32.091
vulnerable to these attacks.
Now take it a step further

00:19:32.152 --> 00:19:36.052
and think about people
that don't know anything about coding

00:19:36.091 --> 00:19:39.292
too much about it
and just live code their way through the

00:19:39.331 --> 00:19:43.152
world. Yeah, these are quite

00:19:43.232 --> 00:19:47.212
interesting targets, I'd say. So yeah,
that was a thing, that will

00:19:47.232 --> 00:19:50.732
be a thing, I think.
Speaking of problems and

00:19:51.052 --> 00:19:54.972
attacks, not really an attack
but still kind of worrying

00:19:55.572 --> 00:19:59.232
was that we had two big outages this

00:19:59.292 --> 00:20:02.952
fall, AWS and Cloudflare. AWS

00:20:03.012 --> 00:20:06.952
took significant chunks of the internet
down this October for

00:20:07.012 --> 00:20:10.522
multiple hours, almost a day overall.

00:20:10.572 --> 00:20:14.052
Cloudflare kind of did the same on a
slightly smaller

00:20:14.172 --> 00:20:17.622
scale. Again, is AI a problem here?

00:20:17.652 --> 00:20:21.492
Were those engineers using AI for these
configurations

00:20:21.532 --> 00:20:24.232
that blew, uh, things up? I don't know.

00:20:24.252 --> 00:20:28.152
Probably not, but we don't know. But it's

00:20:28.211 --> 00:20:31.772
kind of worrying how vulnerable the

00:20:31.851 --> 00:20:35.682
overall internet is
and how dependent it is on a

00:20:36.032 --> 00:20:39.762
few big providers. And we knew
that before.

00:20:39.792 --> 00:20:42.972
It's not new and it's not the first AWS
outage.

00:20:42.992 --> 00:20:46.851
But these two big incidents
which happened only a few

00:20:46.912 --> 00:20:50.752
weeks apart, yeah, they really show us

00:20:51.152 --> 00:20:55.072
what the internet has become.
It's been that place of

00:20:55.112 --> 00:20:58.752
decentralization and it kind of still is,

00:20:58.851 --> 00:21:01.572
but on the other hand,
it kind of still isn't.

00:21:01.672 --> 00:21:05.512
If AWS, if one big region of AWS

00:21:06.232 --> 00:21:09.672
gets taken out, boom. Here go

00:21:09.932 --> 00:21:13.492
30% of all websites as it seems. So yeah,
that is

00:21:14.072 --> 00:21:17.572
definitely also something we saw this
year, that happened this year

00:21:17.632 --> 00:21:20.371
implications in one way or the other.

00:21:20.392 --> 00:21:23.532
And that is kind of, well, showing us

00:21:23.912 --> 00:21:26.272
what the internet has become, I guess.

00:21:26.292 --> 00:21:30.272
Speaking of what the internet has become,
and here I have to go back to AI

00:21:30.331 --> 00:21:34.292
briefly. I apologize.
Significant portions of all traffic

00:21:34.371 --> 00:21:38.072
on the internet and of all content on the
internet are

00:21:38.192 --> 00:21:42.152
already generated by AI.
So we got bots reading

00:21:42.292 --> 00:21:45.532
content generated by other bots generated
by AI.

00:21:45.572 --> 00:21:48.831
Great new world, I guess. So that, of

00:21:48.932 --> 00:21:52.232
course, is also something
that accelerated this year.

00:21:52.272 --> 00:21:56.192
I mean, just two or three years ago,
we had almost no

00:21:56.252 --> 00:21:59.472
AI-generated content on the internet
because it

00:21:59.612 --> 00:22:03.112
sucked. But now it's, it's everywhere.

00:22:03.192 --> 00:22:06.561
And guess what? That will also only

00:22:06.672 --> 00:22:09.972
accelerate. I mean,
if you're on the same bubble as I am on

00:22:10.351 --> 00:22:14.152
X,
everybody's just building AI slop

00:22:14.192 --> 00:22:17.892
there. Great. That is, uh,
a lovely new future.

00:22:17.902 --> 00:22:21.672
And don't get me wrong,
you can generate very useful stuff with

00:22:21.712 --> 00:22:25.392
AI.
I love that I'm able to generate images or

00:22:25.512 --> 00:22:29.292
icons with AI. Of course, I am aware
that we

00:22:29.351 --> 00:22:33.252
also have some problems here
if we think about copyright and so

00:22:33.312 --> 00:22:36.512
on. But it is useful, I will admit

00:22:36.522 --> 00:22:40.331
that. And the same is true for some text,
being able to

00:22:40.351 --> 00:22:44.201
summarize a PDF document. There
are many use cases where AI is

00:22:44.272 --> 00:22:46.382
very useful.
I wanna be very clear about that.

00:22:46.412 --> 00:22:50.092
But the AI slop and the internet becoming
more of an AI

00:22:50.132 --> 00:22:52.992
machine full of bots
and AI-generated content,

00:22:54.232 --> 00:22:58.162
yeah, I've said it before, it's, it
is what

00:22:58.192 --> 00:23:02.112
it is.
I'm not super pumped about it though, to

00:23:02.132 --> 00:23:05.022
put it like this. Now,
I could go on for much longer.

00:23:05.052 --> 00:23:08.871
I could talk about TanStack,
which has a little

00:23:08.912 --> 00:23:12.172
wave of popularity,
at least in my bubble now at the end of

00:23:12.252 --> 00:23:16.032
I could talk about the TypeScript port to
Go

00:23:16.092 --> 00:23:19.732
instead of Rust, which
was announced at the beginning of the year

00:23:20.371 --> 00:23:24.272
soon going to be complete.
And you can already try the Go version of

00:23:24.312 --> 00:23:27.342
TypeScript, it looks like. It's awesome.
It's much faster.

00:23:27.452 --> 00:23:30.532
I could talk about the, the trend back to

00:23:30.632 --> 00:23:34.132
monoliths and VPS instead of

00:23:34.172 --> 00:23:36.442
microservices or managed services.

00:23:36.472 --> 00:23:40.201
And all these are just small trends,
small building blocks, of

00:23:40.292 --> 00:23:43.612
course. In large enterprises,
we're still working with

00:23:43.652 --> 00:23:46.002
microservices probably for good reasons.

00:23:46.052 --> 00:23:49.371
And as I mentioned before, React
and Next is still king.

00:23:49.392 --> 00:23:53.172
But-These are all nice little things which

00:23:53.312 --> 00:23:57.262
still make this overall web development
world really

00:23:57.272 --> 00:24:01.252
nice to be in, to put it like this.
I'm super excited about all these

00:24:01.272 --> 00:24:05.072
things. I'm excited about the fact
that we got BetterAuth, an

00:24:05.152 --> 00:24:09.052
amazing authentication library
which became super popular this year,

00:24:09.072 --> 00:24:12.992
which I'm using in all my projects,
which makes authentication such

00:24:13.072 --> 00:24:16.872
a breeze. I enjoy working with Bun
and Hono.

00:24:17.052 --> 00:24:20.932
I really love what you can do with
PostgreSQL, that you can

00:24:21.032 --> 00:24:24.612
use extensions like pgvector to build a
vector

00:24:24.712 --> 00:24:28.252
database and build RAG applications,
because again, there are

00:24:28.392 --> 00:24:32.032
useful things you can do with AI.
I also very much like the

00:24:32.072 --> 00:24:35.732
fact that you can use Cursor
and Cloud Kode or OpenCode

00:24:35.772 --> 00:24:39.592
or all these other AI coding assistants to

00:24:39.652 --> 00:24:43.092
be more productive
and to outsource some tasks to AI.

00:24:43.102 --> 00:24:47.062
I'm absolutely loving the fact
that we got a bunch of

00:24:47.112 --> 00:24:50.612
amazing new features built into browsers,
like the

00:24:50.652 --> 00:24:54.251
popover API and many, many other cool

00:24:54.292 --> 00:24:58.232
APIs, and also CSS properties
and values that you can

00:24:58.332 --> 00:25:01.752
use today in all major browsers to make
your life

00:25:02.352 --> 00:25:05.952
easier. The problem there just is that AI

00:25:06.112 --> 00:25:09.732
probably doesn't know about these features
yet, and maybe never if it

00:25:09.812 --> 00:25:13.372
just is that machine that builds React
and Next

00:25:13.472 --> 00:25:16.372
applications,
and then learns based on what it built

00:25:16.392 --> 00:25:19.472
But we'll see how that turns out.
I just wanna

00:25:19.482 --> 00:25:22.552
say, web development is, in

00:25:22.752 --> 00:25:26.262
theory, in great shape. We got all these

00:25:26.312 --> 00:25:28.682
amazing features
that we can use in browsers today.

00:25:28.772 --> 00:25:32.092
Many things are much easier to build than
in the past.

00:25:32.112 --> 00:25:35.512
We got mature frameworks.
We got innovation through

00:25:35.632 --> 00:25:38.652
frameworks like TanStack Start
or the BetterAuth

00:25:38.732 --> 00:25:42.532
library.
We got so many amazing stuff going on

00:25:42.792 --> 00:25:46.012
can be such a useful tool. I

00:25:46.272 --> 00:25:50.172
really, really hope that we'll be able to

00:25:50.252 --> 00:25:54.132
make good stuff out of all these building
blocks, that we'll be able

00:25:54.432 --> 00:25:57.772
to build up on this great foundation in
the

00:25:57.832 --> 00:26:01.732
future,
and that we'll not get consumed by all the

00:26:01.872 --> 00:26:05.632
AI slop and all the BS around AI, and

00:26:05.752 --> 00:26:09.492
all the problems that come from AI and

00:26:09.612 --> 00:26:13.452
companies trying to force AI on us in
every aspect of our

00:26:13.512 --> 00:26:17.312
life. I really hope that executives

00:26:17.352 --> 00:26:21.282
will learn that AI is a great assistant,
but

00:26:21.412 --> 00:26:25.332
not a great replacement for developers,
and that

00:26:25.372 --> 00:26:29.361
we can leverage the potential of all these
amazing

00:26:29.472 --> 00:26:33.372
things we got today,
because I truly believe that this is the

00:26:33.412 --> 00:26:36.742
best time ever to be a developer and a web

00:26:36.812 --> 00:26:40.632
developer.
But it's also one of the most difficult

00:26:40.672 --> 00:26:44.472
times, unfortunately. So yeah, that
was my look back.

00:26:44.592 --> 00:26:47.292
I might have missed some things you found
important too.

00:26:47.332 --> 00:26:49.382
Please let me know. Please let me know.

00:26:49.512 --> 00:26:52.932
I try to be comprehensive,
and this video is already super long.

00:26:52.952 --> 00:26:56.942
I will also create a separate video where
I take a look at the next year, and

00:26:57.432 --> 00:27:00.641
I create a bunch of videos where I share
opinions and thoughts about everything all

00:27:00.672 --> 00:27:04.652
the time. So yeah, if you like that,
follow me and

00:27:04.712 --> 00:27:07.852
have a great end of the year
and a great start into the new

00:27:07.972 --> 00:27:08.712
year.