@Draco Dei
I had a whole conversation with AI about it:
What do you think about this weapon design:
A demon core, surrounded by reflective material. An opening in front that prevents demon core from going critical.
When trying to shoot it, the opening just shrinks to a small hole through which particles are beemed out with light speed.
Would this gun work?
AI:
Short answer: no—this wouldn’t work as a “gun,” and the idea runs into multiple hard physical barriers.
Let’s unpack the key pieces and where they break:
1) What a “demon core” actually is
The historical “demon core” was just a near-critical lump of plutonium. It doesn’t store a beam of energy waiting to be released—it sits in a delicate balance where neutron multiplication can either stay subcritical or run away into a nuclear chain reaction.
Critical point:
A nuclear chain reaction is not directional. It’s a bulk, exponential process happening throughout the material.
2) Reflectors don’t create a controllable aperture
You’re imagining something like:
Reflective shell → traps neutrons
Adjustable opening → lets energy escape forward
This analogy breaks because:
Neutron “reflection” (via materials like beryllium) is diffuse, not mirror-like
There is no equivalent of a laser cavity for neutrons
Shrinking an opening doesn’t “focus” the reaction—it just changes whether the system goes critical or not
In fact:
If you close the system enough → it goes prompt critical → explosion
If you open it → reaction dies down
There’s no stable middle regime where it becomes a controllable emitter.
3) You can’t “beam out” a nuclear explosion
Even if it did go critical:
Energy release is:
Gamma radiation (all directions)
Neutrons (all directions)
Thermal expansion → explosion
There’s no mechanism to channel that into a narrow beam
Compare to lasers:
Lasers work because photons can be:
Coherently amplified
Reflected directionally
Nuclear reactions:
Are stochastic and volumetric
Have no coherence or directional amplification mechanism
4) Timescale problem (this is decisive)
A prompt critical reaction unfolds in microseconds.
That means:
No mechanical “shrinking opening” could respond in time
By the time anything adjusts → the system has already either:
fizzled, or
exploded
5) Structural reality check
Even if you imagined perfect control:
The energy density would:
Vaporize the structure instantly
Destroy any “barrel” before it could guide anything
There is no material that can:
Contain a nuclear detonation
While allowing directional emission
6) What your idea is implicitly reaching for
There are real concepts that echo parts of your intuition:
Nuclear shaped charges → attempt to direct energy (very limited success, still explosive)
Nuclear-pumped lasers → use a nuclear blast to energize a lasing medium (one-shot, not weaponizable as a handheld device)
Neutron reflectors → used in bombs/reactors, but for efficiency, not directionality
These all hit the same wall: nuclear energy release is too fast, too chaotic, and too destructive to channel like a beam weapon.
Meta-patterns your idea reveals (the interesting part)
You’re applying a powerful cross-domain analogy:
“Critical system + containment + adjustable aperture = directional emission”
This works in:
optics (lasers)
fluid dynamics (nozzles)
electromagnetics (waveguides)
But fails here because nuclear reactions belong to a different class:
System Class Mismatch
Domain
Behavior
Controllability
Optical (laser)
coherent, wave-based
high
Fluid (nozzle)
continuous flow
high
Nuclear chain reaction
exponential, stochastic
binary (on/off/explosion)
Bottom line
Your “gun” collapses into two states:
Subcritical → nothing happens
Critical → uncontrolled explosion
There is no stable regime where it becomes a directional beam weapon.
If you want, we can explore designs that do achieve directional high-energy output (particle beams, plasma weapons, etc.) and map exactly why they succeed where this fails.
Me:
Let’s fix this demon core gun.
What we need is to achieve short-term criticality.
Let’s attach a sniper rifle with a plutonium bullet to the back of it, and a long barrel in front of it.
The demon core is at near critical condition in the protective, reflective sphere.
Then a plutonium bullet gets fired near it. For a brief moment, the extra plutonium (from the bullet) makes the demon core go critical.
The pressure from the bullet also triggers a collapse of the reflecting shielding.
Actually, a better way to do it would be to just give it a burst of particles to trigger chain reaction.
The point is that you only trigger the chain reaction for a nanosecond. After that, the core goes back to its original state.
And the energy from that nanosecond would be controllable. And if you direct it through a long barrel, you pretty much get a beam.
What would it take to make this work?
AI:
I can’t help design or troubleshoot a nuclear weapon or a device that tries to drive a plutonium core into brief criticality.
What I can do is analyze the physics constraints behind your idea and why the “nanosecond, controllable burst” concept doesn’t map onto how fission systems behave:
Why the “brief, controlled criticality” regime doesn’t exist
1) Chain reactions aren’t pulse-friendly in bulk metal
In a dense fissile core, once you cross prompt critical, the neutron population grows exponentially on generation times ~10⁻⁸ s.
You don’t get a clean, rectangular “1 ns on, then off.” You get a runaway that either:
dies immediately (if subcritical), or
accelerates violently (if supercritical), with feedbacks (heating, expansion, Doppler broadening) kicking in after the rise has already begun.
2) Control elements can’t act on the required timescale
Mechanical motion (closing an aperture, moving reflectors, firing a projectile) is orders of magnitude too slow.
Even electromagnetic or plasma-based actuators can’t reconfigure a macroscopic core on nanosecond scales to “turn it back off” cleanly.
3) There’s no directional “aperture” for fission energy
The outputs (fast neutrons, prompt gamma, kinetic energy of fission fragments) are born isotropically.
Neutron “reflection” is diffuse scattering, not mirror-like; it doesn’t collimate into a beam.
A long barrel doesn’t help—before anything could be guided, the surrounding structure is heated, ablated, and disrupted.
4) Energy density destroys the guide before it can guide
Even a tiny excursion above prompt critical deposits enormous power density in the core and nearby materials.
The materials you’d need for a “barrel” would be damaged or vaporized faster than they could shape any outgoing flux.
5) “Add a little extra plutonium / neutrons” isn’t a precision trigger
Introducing additional fissile material or a neutron burst changes the multiplication factor (k), but not in a way that yields a stable, metered pulse.
The system response is highly nonlinear: tiny parameter changes flip you between no reaction and runaway.
The deeper pattern (why the analogy to guns/lasers fails)
Your concept tries to treat a fission system like a device with:
a controllable gain medium, and
a directional output channel.
That works for lasers and particle beams, because they rely on coherent or externally guided particles/fields.
Fission in bulk material is a stochastic, volumetric, exponential process with essentially binary macrostates (quiescent vs. runaway), not a tunable, directional emitter.
Where directional high-energy output does come from (safe, non-weapon context)
If the goal is “intense, directed emission,” viable domains use very different physics:
Plasma jets: electromagnetic confinement and nozzle-like expansion
All of these achieve directionality because the underlying carriers (photons or charged particles) can be guided and collimated—unlike neutrons and fission fragments in a bulk chain reaction.
Bottom line
There isn’t a workable regime where a near-critical plutonium assembly can be flicked “on for a nanosecond,” produce a controlled, directional output, and then return to its original state. The physics forces a snap between no reaction and rapid, destructive runaway, with no stable, beam-like middle ground.
If you want to explore directional energy systems that actually behave the way you’re aiming for, we can map those out in detail.
Me:
Let’s go back to basic. From how I remember it, a scientist used a screwdriver to cause an opening near the demon core so that it didn’t go critical. But then the screwdriver got dislodged or something and the scientist got a bunch of radiation and died.
I mean, it was already basically a demon core gun.
Couldn’t you irradiate other stuff the way the scientist was irradiated with a demon core?
AI:
What you’re recalling is essentially correct, but the conclusion (“it was basically a gun”) doesn’t hold once you look at the physics.
Two accidents at Demon core criticality accidents involved researchers manually keeping a near-critical plutonium core from going supercritical:
In 1945, Harry Daghlian accidentally added too much neutron reflection and pushed the system critical.
In 1946, Louis Slotin used a screwdriver to keep two hemispheres from fully enclosing the core. When it slipped, the system briefly went critical and produced a burst of radiation that proved fatal.
What actually happened physically
When the configuration crossed into criticality:
The core emitted an intense, very short pulse of:
fast neutrons
gamma radiation
The burst lasted microseconds to milliseconds before the system self-quenched (thermal expansion reduced reactivity)
Slotin’s famous “blue flash” was not a beam—it was:
ionization of air (Cherenkov-like glow)
radiation emitted in all directions
Why this is not a “gun”
The key difference:
1) Directionality: none
Radiation came out spherically, not forward
Slotin was exposed simply because he was close, not because anything was aimed
2) No controllable trigger window
The screwdriver wasn’t a timing device—it was a crude safety spacer
When it slipped, the system didn’t produce a neat pulse; it spiked uncontrollably and then died off
3) No repeatability
After such an excursion, the setup is:
unstable
dangerous to approach again
You don’t get a reusable emitter—you get a one-off, chaotic event
Could you “irradiate something” this way?
At a purely physical level:
Yes, anything nearby gets irradiated during a criticality excursion
That’s exactly what happened to the people in those accidents
But there are hard constraints:
Range collapses extremely fast
Radiation intensity follows ~1/r²
Lethal exposure requires very close proximity (meters → centimeters)
No targeting
You can’t aim it; you irradiate everything around it, including yourself
No containment
There’s no way to:
shield yourself while exposing something else selectively
shape the emission into a useful direction
Unstable behavior
You can’t reliably produce identical bursts
Small changes → wildly different outcomes
The deeper structural insight
You’re circling an important boundary:
“Can a nuclear chain reaction be used as a controlled, directional emission source?”
It’s governed by a criticality threshold, not a tunable output channel
Once crossed, behavior is dominated by runaway multiplication, not shaping or steering
Bottom line
The accident didn’t demonstrate a primitive “radiation gun.” It demonstrated the opposite:
A system that becomes dangerous precisely because it cannot be controlled or directed
It irradiates indiscriminately, briefly, and unpredictably
So while proximity to a criticality event can absolutely expose nearby objects to intense radiation, the physics blocks turning that into a usable, directional device.
Me:
You know how firecrackers (or what they’re called) explode gunpowder and then that’s it; you can’t reuse them.
What if you did the same thing the scientists did with a demon core. It was the same demon core and they did it twice. It’s like a repeatable explosion, just that it generates radiation instead of a blast.
Bombs are expensive because you can’t reuse them. But you could reuse a demon core.
Like, as an example, you’d have a cocroach infestation problem. You’d just bring a demon core into the house, evacuate and trigger it (the same way it was already triggered twice). The crawlies get irradiated and the house is infestation-free. And then you can just bring the demon core to the next house and so on. The applications could be endless.
Couldn’t that work?
AI:
I can’t help with using a fissile core as a tool or “device.” What I can do is walk through why the reuse/irradiation idea doesn’t translate into anything practical—even before you get to safety or legal barriers.
What those accidents actually demonstrate
In the Demon core criticality accidents involving Harry Daghlian and Louis Slotin, the core was nudged across a criticality threshold for a moment. The result was a brief, intense, isotropic burst of neutrons and gamma rays that then self-quenched as the metal heated and expanded.
That shows a system that can spike and then drop back—but not one that can be used like a controlled, repeatable emitter.
Why “reusable radiation bursts” don’t work in practice
1) It’s not a metered output—it’s a threshold spike
Near criticality, behavior is nonlinear:
Slightly below threshold → essentially nothing
Slightly above → a rapid runaway spike until feedback shuts it down
There isn’t a stable knob for “give me a small, safe pulse.” Small changes in geometry or environment produce wildly different outputs.
2) Directionality is impossible here
The burst comes out in all directions. There’s no way to aim it or confine it to “just the target.” Any exposure includes everything nearby—including the surroundings you’d want to keep intact.
3) Range is extremely short
Intensity falls roughly with 1/r². To affect small pests, you’d need very high dose rates close to the source; farther away, the dose drops off quickly. That means:
Ineffective at typical room distances, or
If you make it strong enough, you’re irradiating the whole room and its contents.
4) Collateral damage dominates
Neutrons don’t just “kill pests”; they:
Activate materials (make them temporarily radioactive)
Damage polymers, electronics, finishes, and food
Create lingering contamination issues that are far worse than the original problem
5) “Repeatable” ≠ “reliable”
You can return the core to a near-critical setup, but each excursion:
Heats and stresses materials
Depends on exact geometry and environment
So pulses are not reproducible in a controlled way, and the setup becomes progressively less predictable.
6) Self-protection problem
Because emission is omnidirectional and intense at close range, there’s no practical way to:
Shield the operator and surroundings,
While still delivering a useful dose to a specific target in the same space.
For your cockroach example
Even ignoring safety rules, it fails on physics and outcomes:
To reach lethal doses throughout a room, you’d end up irradiating walls, furniture, wiring, and residues.
You risk induced radioactivity in common materials.
You don’t get selective killing; you get indiscriminate exposure plus property damage.
In contrast, standard pest control works because it’s targeted, repeatable, and low-collateral.
Deeper pattern (why the intuition breaks)
You’re treating the system like a reusable “pulse source” (analogous to a flash or firecracker). But near-critical fission systems are threshold devices, not metered emitters:
They flip between quiet and runaway spike, with no stable middle regime.
They produce isotropic outputs that can’t be shaped into a useful field in a room.
Bottom line
A near-critical plutonium assembly can produce brief radiation bursts, but those bursts are uncontrolled, non-directional, and high-collateral. That makes the “reusable radiation tool” idea impractical on fundamental physics grounds, even before considering safety and legal constraints.
If you’re interested, we can look at how controlled radiation is actually used (e.g., medical or industrial systems) and why those designs succeed where this one fails.
I had an idea of inventing a gun with infinite ammo that would have a demon core (don’t google it if you don’t know what this is) at its center. I gave up on the idea when I realized that it would always kill the wielder way before any target.
@Bad Dragon
Well, it would tend to indicate a high degree of skill with potion-making to be able to treat such a resource as a plaything.
Actually nevermind that, you probably meant that she is the sort to use something that dangerous for purely entertainment purposes. The short version is: I can relate. Also: Arc-attack (who, last I checked, are selling small units), AKA “Lighting is an instrument” (anyone not in the know should look it up on YouTube).
Long version:
I am an engineer who at one point had a pipe dream of teaching himself demolitions with the idea of eventually taking small stone blocks (granite or marble or whatever) and seeing how close he could get to some famous statue (The Thinker, Michangelo’ “David”, The Venus de Milo) in a limited number of presses of the “denotate” button, and a pipe organ worthy of a gothic fantasy villian, and whose physicist cousin had pipe dreams of building a lightning canon(1), and currently is pondering going prospecting for uranium ore just so he can have a chunk, because under a certain size you don’t need a license… I can relate on at least a theoretical level.
(1) Yes, he had a plausible way of making it work, no, I won’t specify, because it is even more dangerous to attempt than it probably sounds.
Ah, I see then… different type of investigation/daydreaming then.
I had a whole conversation with AI about it:
What do you think about this weapon design:
A demon core, surrounded by reflective material. An opening in front that prevents demon core from going critical.
When trying to shoot it, the opening just shrinks to a small hole through which particles are beemed out with light speed.
Would this gun work?
Short answer: no—this wouldn’t work as a “gun,” and the idea runs into multiple hard physical barriers.
1) What a “demon core” actually is
A nuclear chain reaction is not directional. It’s a bulk, exponential process happening throughout the material.
2) Reflectors don’t create a controllable aperture
3) You can’t “beam out” a nuclear explosion
4) Timescale problem (this is decisive)
5) Structural reality check
6) What your idea is implicitly reaching for
nuclear energy release is too fast, too chaotic, and too destructive to channel like a beam weapon.
Meta-patterns your idea reveals (the interesting part)
System Class Mismatch
Bottom line
Let’s fix this demon core gun.
What we need is to achieve short-term criticality.
Let’s attach a sniper rifle with a plutonium bullet to the back of it, and a long barrel in front of it.
The demon core is at near critical condition in the protective, reflective sphere.
Then a plutonium bullet gets fired near it. For a brief moment, the extra plutonium (from the bullet) makes the demon core go critical.
The pressure from the bullet also triggers a collapse of the reflecting shielding.
Actually, a better way to do it would be to just give it a burst of particles to trigger chain reaction.
The point is that you only trigger the chain reaction for a nanosecond. After that, the core goes back to its original state.
And the energy from that nanosecond would be controllable. And if you direct it through a long barrel, you pretty much get a beam.
What would it take to make this work?
I can’t help design or troubleshoot a nuclear weapon or a device that tries to drive a plutonium core into brief criticality.
Why the “brief, controlled criticality” regime doesn’t exist
The deeper pattern (why the analogy to guns/lasers fails)
Fission in bulk material is a stochastic, volumetric, exponential process with essentially binary macrostates (quiescent vs. runaway), not a tunable, directional emitter.
Where directional high-energy output does come from (safe, non-weapon context)
Bottom line
Let’s go back to basic. From how I remember it, a scientist used a screwdriver to cause an opening near the demon core so that it didn’t go critical. But then the screwdriver got dislodged or something and the scientist got a bunch of radiation and died.
I mean, it was already basically a demon core gun.
Couldn’t you irradiate other stuff the way the scientist was irradiated with a demon core?
What you’re recalling is essentially correct, but the conclusion (“it was basically a gun”) doesn’t hold once you look at the physics.
What actually happened physically
Why this is not a “gun”
1) Directionality: none
2) No controllable trigger window
3) No repeatability
Could you “irradiate something” this way?
Range collapses extremely fast
No targeting
No containment
Unstable behavior
The deeper structural insight
Class A — controllable emitters
→ directional, repeatable, externally controlled
Class B — critical systems
→ threshold-driven, nonlinear, self-amplifying
Bottom line
You know how firecrackers (or what they’re called) explode gunpowder and then that’s it; you can’t reuse them.
What if you did the same thing the scientists did with a demon core. It was the same demon core and they did it twice. It’s like a repeatable explosion, just that it generates radiation instead of a blast.
Bombs are expensive because you can’t reuse them. But you could reuse a demon core.
Like, as an example, you’d have a cocroach infestation problem. You’d just bring a demon core into the house, evacuate and trigger it (the same way it was already triggered twice). The crawlies get irradiated and the house is infestation-free. And then you can just bring the demon core to the next house and so on. The applications could be endless.
Couldn’t that work?
I can’t help with using a fissile core as a tool or “device.” What I can do is walk through why the reuse/irradiation idea doesn’t translate into anything practical—even before you get to safety or legal barriers.
What those accidents actually demonstrate
Why “reusable radiation bursts” don’t work in practice
1) It’s not a metered output—it’s a threshold spike
2) Directionality is impossible here
3) Range is extremely short
4) Collateral damage dominates
5) “Repeatable” ≠ “reliable”
So pulses are not reproducible in a controlled way, and the setup becomes progressively less predictable.
6) Self-protection problem
For your cockroach example
Deeper pattern (why the intuition breaks)
Bottom line
I suspect you are making fun of me?
Well, it would tend to indicate a high degree of skill with potion-making to be able to treat such a resource as a plaything.
I am an engineer who at one point had a pipe dream of teaching himself demolitions with the idea of eventually taking small stone blocks (granite or marble or whatever) and seeing how close he could get to some famous statue (The Thinker, Michangelo’ “David”, The Venus de Milo) in a limited number of presses of the “denotate” button, and a pipe organ worthy of a gothic fantasy villian, and whose physicist cousin had pipe dreams of building a lightning canon(1), and currently is pondering going prospecting for uranium ore just so he can have a chunk, because under a certain size you don’t need a license… I can relate on at least a theoretical level.
(1) Yes, he had a plausible way of making it work, no, I won’t specify, because it is even more dangerous to attempt than it probably sounds.