Mark Nielson

[Mark Nielson @Mark_nielson_]

I'm currently working on something that sounds absolutely ridiculous, but the more I research it, the more I believe there's a real experiment worth building here.

A cybersecurity system that uses the brain of a living DOG to generate cryptographic randomness.

Yes, an actual dog.

The project is called K9ENTROPY.

Let me explain.

Almost everything we do on the internet depends on cryptography. Every time you log into a website, connect to a server, access a database, or generate an encryption key, you're relying on mathematical systems designed to keep your information secure.

One of the most important ingredients in these systems is randomness.

Computers are extremely good at following instructions, but producing truly unpredictable information is a completely different challenge.

Modern security systems already have sophisticated ways of generating cryptographic randomness, but I've been wondering about something.

What if we introduced a completely biological source of unpredictability into the equation?

And what if that source was the brain of a living animal?

A dog's brain contains billions of neurons constantly communicating through electrical activity.

Even when a dog is doing absolutely nothing, its brain is processing sensory information, memories, sounds, smells, and countless other signals.

That activity is incredibly complex.

So here's the idea.

We connect a real dog to a computer using a non-invasive EEG headset that monitors electrical activity from its brain.

No implants. No surgery. Nothing that should hurt the animal.

The system records those neural signals and sends them into a processing engine designed to identify and extract potentially unpredictable information.

We then analyze that information, test how much genuine entropy it contains, and investigate whether it can contribute to cryptographic random-number generation.

Imagine a server generating experimental cryptographic material using a combination of trusted system randomness and unpredictable signals originating from a living dog's brain.

The dog's neural activity becomes an additional biological input to the security system.

And that's where K9ENTROPY comes in.

I've already built the first software prototype.

It includes a live neural activity dashboard, simulated EEG channels, a PHP backend, a cryptographic random-number API, and an interface designed to eventually accept recordings from actual EEG hardware.

The first version uses simulated neural signals. The real biological connection is the next research challenge.

There's also a major question we need to answer before making any claims about security.

Does the brain actually produce cryptographically useful entropy, or does its activity merely appear random?

Those are two completely different things.

A complicated biological signal isn't automatically a secure random-number generator.

We need to measure it, test it against established cryptographic standards, examine whether someone could influence or predict the signals, and ensure the entire system stays secure even when the dog isn't connected.

That's exactly what I want to investigate.

The long-term vision is a biological entropy module that websites, servers, and databases could integrate alongside their existing security infrastructure.

A system where living neural activity contributes to the generation of cryptographic randomness.

And the craziest part is that the animal wouldn't even need to understand what it's doing.

It could be sitting there thinking about food, watching someone walk across the room, or listening to sounds outside.

Meanwhile, researchers could be studying whether the electrical activity produced by those ordinary moments contains useful unpredictable information.

I don't know whether this will outperform conventional hardware randomness.

It might not.

But if we can demonstrate that neural activity provides measurable, independently useful entropy, we'd have an entirely new direction to explore in biological computing and cybersecurity.

K9ENTROPY is officially in development.

The software foundation is there.

Next comes real EEG hardware, actual neural recordings, and the difficult part: proving whether any of this genuinely works.

A cybersecurity experiment powered by the brain of man's best friend.

What a strange timeline we're living in.

[Mark Nielson @Mark_nielson_]

I don't want K9ENTROPY to become a story about how I "replaced cryptography with a dog's brain."

That's not what I'm building.

I built a software prototype to investigate a very unusual biological entropy source. If the data supports the idea, I'll keep exploring. If it doesn't, I'll say so.

That's the whole point of the experiment.

[Mark Nielson @Mark_nielson_]

I won't move to real canine EEG experiments without appropriate animal-care guidance.

Non-invasive doesn't automatically mean comfortable. I want careful acclimation, short sessions, close observation, and a hard stop if the dog shows distress.

The research can wait. The animal comes first.

[Mark Nielson @Mark_nielson_]

I'm keeping the distinction between an entropy source and a random-number generator very clear as I design K9ENTROPY.

They're not interchangeable. I want to understand how any experimental signal might fit into a properly designed system, if it proves useful at all.

https://t.co/Pjyl7gQj9I

[Mark Nielson @Mark_nielson_]

The big thing I'm trying to establish with K9ENTROPY:

Generating changing numbers from neural signals is not the same as generating trustworthy cryptographic randomness.

I can build the first part. The whole challenge is validating the second.

[Mark Nielson @Mark_nielson_]

It's funny to imagine a dog thinking about treats while I'm studying its EEG readings for cryptographic research.

But a fun premise doesn't excuse weak methodology. I still need to prove there's meaningful unpredictable information in the data.

[Mark Nielson @Mark_nielson_]

Basically this guy recreated the claudefare lava lamp security is but with his dog ? This is legit giga and we have his github

[Terry @terryxxc__]

OMG THIS IS SO GIGA.

This dude built K9ENTROPY, a cybersecurity experiment designed to generate cryptographic randomness from the brain activity of a living dog.

The idea is insane.

A non-invasive EEG headset reads the dog’s neural signals while it rests, watches the room or thinks about food. The software processes that unpredictable biological activity and tests whether it can contribute useful entropy alongside trusted system randomness.

He already built the first prototype with simulated EEG channels, a live neural dashboard, a PHP backend and a cryptographic random-number API.

The next step is connecting real EEG hardware and finding out whether a dog’s brain produces genuinely useful cryptographic entropy or signals that only look random.

No implants. No surgery. No replacement for proven security systems.

Just a serious experiment asking whether man’s best friend could help secure the internet without even knowing it.

Sending fees to his git

[Mark Nielson @Mark_nielson_]

Before I claim anything about K9ENTROPY's security, I want to be able to answer five questions:

Can an attacker manipulate the measurements?
How correlated are the samples?
What is the defensible entropy estimate?
How sensitive is the result to sensor placement?
What happens if the input fails?

[Mark Nielson @Mark_nielson_]

@Dean147 Hey is this your website ? https://t.co/2PhQYzfIiP

[Mark Nielson @Mark_nielson_]

I started K9ENTROPY with a strange hypothesis: could a dog's neural activity contribute independently useful entropy to cryptographic systems?

I built a simulated software prototype. Next I want to investigate real data and test the hypothesis properly.

Nothing is proven yet.

[Mark Nielson @Mark_nielson_]

I could put "powered by a living brain" on a landing page and make K9ENTROPY sound like magic.

I'd rather publish a boring min-entropy estimate and let people inspect the test results.

[Mark Nielson @Mark_nielson_]

I'm designing K9ENTROPY so the experimental EEG input can disconnect without breaking security.

The existing cryptographic randomness needs to keep working normally when there's no dog connected.

A research experiment should never become a dependency that compromises a production system.

[Mark Nielson @Mark_nielson_]

One thing I'm worried about when I get to real EEG recordings: movement artifacts.

If the dog moves and the waveform changes, I can't just declare that extra entropy.

First I have to figure out what the sensors are actually picking up.

[Mark Nielson @Mark_nielson_]

I'm planning to share the failed experiments too.

If I test real EEG data and it turns out to be predictable, that's still valuable information.

K9ENTROPY should be research, not a marketing exercise where every result somehow becomes a breakthrough.

[Mark Nielson @Mark_nielson_]

One of the hardest parts of building K9ENTROPY is resisting the temptation to jump straight to an exciting conclusion.

I have a software prototype with simulated neural activity, a dashboard and a PHP random-number API. That doesn't prove a living dog's brain can supply cryptographically secure randomness.

Once I can evaluate suitable real recordings, I'll need to study how much of the signal comes from neural activity versus sensor noise, movement, interference and processing. I'll need to look for correlations, test whether external stimuli could affect the readings, and estimate how much of the resulting data an attacker might realistically predict.

I also want a clean failure mode. If the experimental input disappears, a secure system should continue to rely on its established cryptographic randomness. The dog should never become a single point of failure.

And the recording process has to be humane. I won't sacrifice an animal's comfort for a technical demonstration.

This might end with a negative result. That's fine. The goal is to investigate a real question and publish evidence, not manufacture a breakthrough.

https://t.co/gltQa8YZfb

[Mark Nielson @Mark_nielson_]

I've been reading NIST SP 800-90B because I want K9ENTROPY to be judged against proper entropy-source standards, not vibes.

Health tests, conservative estimates and reproducible measurements matter.

https://t.co/jBmUuCQLcu

[Mark Nielson @Mark_nielson_]

Getting bytes from a neural signal is easy compared with proving those bytes contain cryptographically meaningful entropy.

That difference is basically the entire research challenge I'm taking on with K9ENTROPY.

[Mark Nielson @Mark_nielson_]

The dog's welfare comes first in K9ENTROPY.

I'm only interested in non-invasive, properly supervised recording that the animal comfortably tolerates. No implants or painful procedures.

If the hardware causes distress, I won't use it. A research question isn't worth that.

[Mark Nielson @Mark_nielson_]

I want to make something clear about what I'm building.

An EEG headset doesn't read a dog's thoughts. It captures weak electrical measurements mixed with noise and motion artifacts.

I have to separate those factors before making any claim about useful entropy.

[Mark Nielson @Mark_nielson_]

I'm trying to design the EEG study around the hardest questions first.

Can an attacker influence the signal? How much is movement artifact? Which samples are correlated?

Measuring those things matters more to me than showing off a flashy waveform.

[Mark Nielson @Mark_nielson_]

I just wanna say thank you for all the support the money will help to get this project ahead

H3nt4XPVP1YUGVnM56ddQP3nuG1pHKPNDJpCunKcpump