path: root/lume/src/blog/a-weapon-to-surpass-metal-gear.mdx

---
title: A weapon to surpass Metal Gear
date: 2023-03-10
tags:
  - philosophy
  - metalGear
  - futurism
  - sciFi
  - ChatGPT
hero:
  ai: Anything+YorHa
  file: raiden-automata
  prompt: 1girl, cyborg, outdoors, landscape, forest, yellow eyes, blindfold, bodysuit, hoodie, white hoodie, boots, skirt, gloves, neutral expression, city ruins
---

Every so often, I like to look at some of the more weird conspiracy
theories and then try to debunk them. I consider it a media literacy
exercise, but there has been one theory that I've come across that is
impressively hard to debunk: the "Dead Internet" theory. I think that
the best conspiracy theories are the ones that are hardest to debunk,
and this one is increasingly getting more difficult to debunk.

The core idea is that the Internet itself is actually dead, no human
authorship of any content exists. Any actual human content that is
created is isolated into its own little
[heavenbanned](https://twitter.com/nearcyan/status/1532076277947330561)
bubble. Mainstream platforms, news outlets, social media sites,
Internet forums, chatrooms, everything filled with bot generated
content to the point that it's impossible to find another human. The
same arguments cycle back and forth throughout the forums with new
dramatis personae every iteration.

To be clear, this theory as literally written is absolute nonsense
and probably not worth taking too seriously. A lot of this theory was
originally designed to come up with reasons why people in forums will
just randomly vanish forever, never to be seen or heard from again.
That friend that you played Team Fortress 2 with for years? Ten
years offline on Steam. No goodbyes. No warning they'd leave. They
just vanish.

As someone who has experienced this type of vanishing a lot, I
get where the conspiracy theory comes from. I think it's quite a
stretch to assert that it's the result of chatbots, or something. But
over all this is a very interesting thing to try to dissect and argue
against. As AI models get more complicated and their output gets more
and more convincing, it's getting difficult to impossible to actually
argue against this theory.

This is something that is entirely possible to create with the modern
technology we have.

I hope that this is actually a false assertion, but it's getting
difficult to make the case that it is. This has really given me a lot
of pause and started to make me wonder what this technology is
actually going to be end up used for.

As a writer, there's two ways that you can analyze futurism and
science fiction. The common model is to compare it to the car. If you
are to talk about the idea of the car, the possibilities it allows
for, and what it enables people to do, that is futurism. Futurism is
mainly talking about the ideas and thoughts about what technology
could give us. For comparison, if you were to analyze it as science
fiction, you would probably start with the idea of traffic jams.
Futurism would enable people to live and work in different
geographical places, this is something that has not commonly been done
for the vast majority of history for the vast majority of population.
People generally lived where they worked, because travel was
difficult, expensive, and annoying. But, if everyone has to do that,
then there will likely be times where the roads are filled with cars
for people trying to get to somewhere else so they can do their job.
This creates a traffic jam. This is the science fiction of the car.

I would like to wager that AI technologies like ChatGPT, Stable
Diffusion, LLaMA, Whisper, and more are looked at in the eyes of
futurism. To contrast, I believe the "dead internet" theory is the
science fiction. If we have the ability to create all of this stuff
automatically by machine, then what will be left for humans to enjoy?

## Metrics and Ratios

At some level though, I recognize that for about eighty percent of
the articles that people actually read, an AI generated article is
probably sufficient. Consider the fact that most of articles that
people read are very basic high level things:

- X died
- X person said Y
- Event happened at place
- "No Way To Prevent This," Says Only Nation Where This Regularly
  Happens

Statistically, most people won't even read past the title anyways.
This is why titles get clickbaited. Because people want you to read
the fucking article, and nobody reads past the title. Consider the
fact that when Elon Musk, one of the biggest accounts on Twitter,
tweeted a link to someone's webpage, there was a less than one percent
click through rate. When you don't click on a link on Twitter, all you
see is the title and any extra preview data that Twitter scraped.

If that has a less than one percent click through rate, then why
should companies bother to pay humans to write articles that aren't
going to be read?

Turns out, capitalism has noticed this.
[Buzzfeed](https://buzzfeed.com) is reportedly [using AI models to
generate its listicles and other filler
content](https://www.forbes.com/sites/chriswestfall/2023/01/26/buzzfeed-to-use-chatgpts-ai-for-content-creation-stock-up-200/?sh=70b2faf67eae).
CNET [also recently had a spat with AI generated
articles](https://gizmodo.com/cnet-chatgpt-ai-articles-publish-for-months-1849976921),
albeit one that ended poorly.

## Arsène

What could such a dead internet look like? I decided to do a little
experiment and find out. I call the results
[Arsène](https://arsene.fly.dev/). It is a little app with Deno and
Fresh that creates a new article about every 12 hours about the
current phase of the moon, astrological sign, and three tarot cards to
describe the past, present, and future. It feeds all of this into
ChatGPT and I have done very little prompt engineering to try and
hide the fact that it is AI generated.

My fear is that all major/popular websites will end up looking like
this. Pleasing design, modern user experience, but absolutely nothing
worth witnessing.

<XeblogConv standalone name="Cadey" mood="coffee">
  [Fresh](https://fresh.deno.dev/) and [Tailwind](https://tailwindcss.com/)
  actually really does make for a nice combination for making decent looking
  websites at a very low amount of effort spent in the design phase. I'm likely
  going to use it a lot in the future for my more experimental projects or
  things that I don't expect to scale.
</XeblogConv>

If there's going to be all of this automatically generated content,
how are we going to sort through it all? Who is going to take the
time to curate things? How will information search change in the wake
of this technology?

## Context

Metal Gear is a series of games made by producer Hideo Kojima. The
series has always covered complicated topics such as war, political
philosophy, existentialism, free will, understanding how ideology
spreads, and in general is seen as the precursor to understanding a
lot of how the US has developed. And you get to play as super-solders
trying to piece together everything as it all burns.

This series is well worth playing, or at the very least watching the
cutscenes on YouTube. One of the most important things that it covers
is the idea of free will, and censorship on the internet. At some
level, its thesis is that America is diseased. Rotten to the core. We
drown ourselves in information, and there is no good way to sort it,
analyze it, or know what's worth preserving and what's worth
discarding.

In Metal Gear Solid 2, there is a faction known as "The Patriots".
They are a group of artificial intelligence models that has slowly
built up and gained sapience in the White House. By the events of
Metal Gear Solid 2, the patriots control the entire country. The
president is the puppet of the patriots. Information is censored
through the patriots. Broadcast TV, internet forums, email, chatrooms,
all of it censored by the patriots. Oh, and they control the nuclear
launch codes.

There is one cut seen at the end of the game where the patriots spell
out their core thesis and their rationale for why they are doing what
they are doing. I'm going to be quoting parts of it as I continue
this article, but you should really watch the entire thing to help you
understand where I'm coming from here:

<iframe
  width="725"
  height="408"
  src="https://www.youtube.com/embed/eKl6WjfDqYA"
  title="YouTube video
player"
  frameborder="0"
  allow="accelerometer; autoplay;
clipboard-write; encrypted-media; gyroscope; picture-in-picture;
web-share"
  allowfullscreen
></iframe>

Even without the help of artificial intelligence models, we create so
much information that there is no way that anyone is going to be able
to sort through it. In ten seconds, something ridiculous like eighty
five hours of video is uploaded to YouTube. Untold terabytes of data
are minted at Twitter. Millions of blogs post billions of articles
every day. There is so much information, and so much noise, that even
the one company that was designed to sort through the internet is
having difficulty sorting through the internet.

And now, it is even easier to continue that 24/7 spew of trivia and
celebrity bullshit with things like large language models. As an
example, let's think about that art project that I linked to earlier.
Here is the verbatim prompt that it uses in order to generate
articles:

```js
const prompt = `Ethereum is worth ${ethUSD} US dollars.
The current lunar phase is ${lunarPhase}.
The current astrological sign is ${astrologicalSign()}.
The current date is ${new Date().toDateString()}.

Past events: ${tarotReading[0].fortune_telling[0]}
Current events: ${tarotReading[1].fortune_telling[0]}
Prediction: ${tarotReading[2].fortune_telling[0]}

Write a lengthy analysis article that speculates why the prediction will come to pass and what it will do to the price of Ethereum. Lead your analysis with a line containing an original title followed by two new lines.`;
```

Consider this: that project was made in literally two days as an
exploration of how this technology could lead to a "dead internet". I
made this because of a quote that my mentor gave me:

> Come on now, put your fears on the canvas and then you can see how
> bad they really are. It's easy to get caught up in the imagined
> fears of hypotheticals, but looking your fear right at the face lets
> you see if it's truly something to be afraid of.

So I did that. Just imagine, I did this in two days as a joke. If some
idiot like me can spin out something like this in two days that gets
decent enough results, just imagine how this technology could be used
by someone actually malicious. Are we going to be able to trust things
that we read on the internet anymore? Are we going to be able to trust
video? Audio? Text? Images?

> **Colonel**: Raiden, you seem to think that our plan is one of
> censorship.<br /> > **Raiden**: Are you telling me it's not!?<br /> > **Rose**: You're being silly! What we propose to do is not to
> control content, but to create context.<br /> > **Raiden**: Create context?<br /> > [...]<br /> > **Colonel**: Absolutely. Who else could wade through the sea of
> garbage you people produce, retrieve valuable truths and even
> interpret their meaning for later generations?<br /> > **Rose**: That's what it means to create context.<br /> > **Raiden**: I'll decide for myself what to believe and what to pass
> on!<br /> > **Colonel**: But is that even your own idea?<br /> > **Rose**: Or something Snake told you?<br />

<XeblogConv standalone name="Cadey" mood="coffee">
  As I was writing all of this, Discord quite literally announced that they are
  making a new client feature with the ChatGPT API that allows them to
  automatically [create context for your
  conversations](https://www.engadget.com/with-the-help-of-openai-discord-is-finally-adding-conversation-summaries-160030905.html?guccounter=1).
  I literally can't make this up. The parallels and analogies are literally
  writing themselves at this point.
</XeblogConv>

Who will be our patriots? This has been a question that's been
rattling around throughout my skull for the last two weeks or more.

If you take a serious look at the Metal Gear series, you probably
would find it _dripping_ with allegory and other moments like this.
The thing that comes to mind the most is the one of the most
symbolically critical cutscenes from [Metal Gear Rising:
Revengance](https://en.wikipedia.org/wiki/Metal_Gear_Rising:_Revengeance)
where Raiden is talking with the one of the bosses/enemy terrorists
named Monsoon.

This game is also notable because it came out in that small sliver of
time between modern gaming being a thing and the common definition of
memes as what we know them as today didn't exist yet. In an
anthropological sense, you can think about a meme as a viral idea, and
that sense is being referred to below.

When Raiden is talking with Monsoon about war, violence, and justice,
their conversation drifts to anthropology:

> **Monsoon**: Yes, you aren't the only one to grow up on the killing
> fields. War is a cruel parent, but an effective teacher. Its final
> lesson is carved deep in my psyche: That this world, and all its
> people, are diseased. Free will is a myth. Religion is a joke. We
> are all pawns, controlled by something greater: Memes. The DNA of
> the soul. They shape our will. They are the culture -- they are
> everything we pass on. Expose someone to anger long enough, they
> will learn to hate. They become a carrier. Envy, greed, despair...
> All memes. All passed along.

When all these memes (both senses) build up and overwhelm us all, what
happens? Where do people find actual truth when everything worth
interacting with is drowned out.

Later in the game when Raiden fights the final boss, a _United States
Senator_ (this game is about as subtle as a nuclear bomb, it's great),
they talk about the fallout of The Patriots being destroyed in the
wake of the FOXALIVE virus (and the resumption of the uninterrupted
flood of garbage onto the internet).

<iframe
  width="725"
  height="408"
  src="https://www.youtube.com/embed/IMpuUUV2HeE"
  title="Senator
Armstrong&#39;s Political Speech"
  frameborder="0"
  allow="accelerometer; autoplay; clipboard-write; encrypted-media;
gyroscope; picture-in-picture; web-share"
  allowfullscreen
></iframe>

> **Senator Armstrong**: All right, the truth then. You're right about
> one thing... I do need capital. And votes. Wanna know why?.. I have
> a dream.<br /> > **Raiden**: What...?<br /> > **Senator Armstrong**: That one day every person in this nation will
> control their OWN destiny. A land of the TRULY free, dammit!
> [Armstrong resumes fighting Raiden.] A nation of ACTION, not words,
> ruled by STRENGTH, not committee! [Armstrong grapples Raiden,
> > rendering the cyborg mostly helpless.] Where the law changes to suit
> the individual, not the other way around. Where power and justice
> are back where they belong: in the hands of the people! [Armstrong
> > and Raiden headbutt each other, but neither is harmed.] Where every
> man is free to think -- to act -- for himself! [Armstrong beats
> > Raiden to punctuate every statement.] Fuck all these limp-dick
> lawyer and chicken-shit bureaucrats. Fuck this 24/7 Internet spew of
> trivia and celebrity bullshit. Fuck "American pride." Fuck the
> media! Fuck all of it! America is diseased. Rotten to the core.
> There's no saving it -- we need to pull it out by the roots. Wipe
> the slate clean. BURN IT DOWN! And from the ashes a new America will
> be born. Evolved, but untamed! The weak will be purged and the
> strongest will thrive -- free to live as they see fit, they'll make
> America great again!

To say that this game wears its heart on is sleeve is like trying to
say that water makes things moist. Senator Armstrong's core thesis is
that people need to be free to do what they want. However:

> **Colonel**: You exercise your right to "freedom" and this is the
> result. All rhetoric to avoid conflict and protect each other from
> hurt. The untested truths spun by different interests continue to
> churn and accumulate in the sandbox of political correctness and
> value systems.<br /> > **Rose**: Everyone withdraws into their own small gated community,
> afraid of a larger forum. They stay inside their little ponds,
> leaking whatever "truth" suits them into the growing cesspool of
> society at large.<br /> > **Colonel**: The different cardinal truths neither clash nor mesh.
> No one is invalidated, but nobody is right. <br /> > **Rose**: Not even natural selection can take place here. The world
> is being engulfed in "truth."<br /> > **Colonel**: And this is the way the world ends. Not with a bang,
> but a whimper.

Communities that used to have public presences, online forums, wikis,
all of that reduced to private, secluded, Discord servers. Spam has
always been a problem that has required a lot of careful research and
implementation because of how easy it is to do and how hard it is to
only block the spammers, and not the legitimate users. And now new
tools allow you to generate kilobytes of spam in tenths of pennies.

The things we experience become the memes we share with others. Expose
someone to hate long enough, they'll hate too.

> **Colonel**: That's the proof of your incompetence, right there. You
> lack the qualifications to exercise free will.<br /> > **Raiden**: That's not true! I have the right--<br /> > **Rose**: Does something like a "self" exist inside of you?<br /> > **Colonel**: That which you call "self" serves as nothing more than a
> mask to cover your own being.<br /> > **Rose**: In this era of ready-made 'truths', "self" is just something
> used to preserve those positive emotions that you occasionally
> feel...<br /> > **Colonel**: Another possibility is that "self" is a concept you
> conveniently borrowed under the logic that it would endow you with
> some sense of strength.

It is very likely that we are living in an age of ready made truths,
and this will all only get worse as these text transformer models get
more and more easy to use, run on lower and lower and hardware, and
become fast enough that they can be used in real time.

## The traffic jam

This is the traffic jam. A traffic jam not made of cars, but ideas.
Ideas and memes fed to people 24/7. God forbid what happens when
somebody actually malicious gets their hands on these tools.

We're already rapidly approaching the point where you can't trust
video and audio recordings of anything anymore. I've seen a new format
of memes spread around in the last month or two. They use AI text to
speech generation to show former and current US presidents in
ridiculous scenarios such as Joe Biden, Donald Trump, Barack Obama and
Chris Pratt playing Mario Party together:

<iframe
  width="725"
  height="408"
  src="https://www.youtube.com/embed/b6WYD0QeF4s"
  title="The Gamer
Presidents Play Mario Party (ft. Chris Pratt)"
  frameborder="0"
  allow="accelerometer; autoplay; clipboard-write; encrypted-media;
gyroscope; picture-in-picture; web-share"
  allowfullscreen
></iframe>

Just imagine what a boon this could be for misinformation. Imagine a
conspiracy, and with a few clicks of a mouse, create the evidence you
need to spread it. Create the memes that you need in order to
discredit your opposition. They'll spread like wildfire.

At this point, the genie is out of the bottle. Pandora's box is
opened. We're going to have to learn to live with these diseases.

Next year is an election year. I can only imagine what kind of things
we will end up being inflicted to on the Internet this next election
cycle.

It's very likely that we're not going to be able to trust breaking
news about political candidates very soon. It's also very likely that
a lot of the articles that people read are going to look like those
articles made by that joke of a website I made. Our world is being
engulfed in relative truths based on assumptions. The common belief in
an "objective reality" is crumbling.

One event happens. A million people witness it. A million separate
headcanons of that become the reality for those million people.
Some of them may be mutually understood by each other, but they are
still a million separate interpretation that can all conflict in a
million different ways. This is the way that that things always have
been. We just don't like to think about this.

Above all, my biggest fear is that one of the biggest sources of
liberation in my life is going to no longer be a source of liberation
for anyone else. This is the technology that can truly kill the
internet. Not with a bang, but with a whimper.

<div>
  <XeblogConv standalone name="Cadey" mood="coffee">
    I couldn't figure out where to fit this in the article, but I suspect that a
    lot of the reason that so many companies are announcing AI products in
    recent days can boil down to two basic factors:
    <ul>
      <li>
        AI-based products are a perfect denial-of-service attack on product
        departments. They look cool as hell. They raise a lot of buzz. It only
        makes sense that people would want to raise that buzz and look cool as
        hell in order to get promoted.
      </li>
      <li>
        The AI departments of most companies have not really published any
        products. The tech industry as a whole is going through economic
        downturns, and this is making companies look for places to cut the fat.
        I suspect that AI products are being introduced because it allows people
        to remain employed because their department is raising a lot of buzz.
      </li>
    </ul>
  </XeblogConv>
</div>

This is truly a weapon to surpass Metal Gear. Now accessible through
Bing and Google search. Now available for any kid with a dream and a
credit card.