20090926

Titles of Papers not yet Written

Invent titles of ridiculous papers that you would like to exist.

I'm fairly certain that these people titled their paper before they did the research

Edit : I am turning this into some sort of sticky page where I will add in titles that sound cool, but have not been written. Some of these have already been used as titles for posts on this blog, I will try to keep them linked in as things change.


  • Circuit Lower Bounds are Hard because of The Man
  • Completing Your Graduate Thesis Through Applied Hobonics
  • Differential verses Integral complexity as measures of replicator sophistication
  • Growing up in Many Worlds: The Developmental effect of exposing children to The Many Worlds Hypothesis of Quantum Mechanics
  • An improved calibration and auto-ranging function for targeted microwave weapons permitting nonlethal use
  • (in)Sanity and Science.
  • The Journal of Applied Hobonics
  • A Mathematical model for the genetics of drug addiction
  • Metrics of stability in emergent agent based social models
  • Micro-exposure to Common House-Hold Hallucinogen and their Effect on Childhood Development
  • Modalities of a Decentralized Military-Industrial Complex in Globalization Protests.
  • Moduli spaces and non-unique souls.
  • A Novel Hobonic Approach to Food Acquisition
  • Numerical solution for audiovisual stimuli of maximum representational dissonance
  • The origin of cognitive impulses
  • Post-Human Rights in a Postmodern Polity
  • The potentials of Molecular Computation in non-coding genomic DNA
  • Putting the so called junk DNA to work as a computational medium
  • Reconstructing Extinct Genomes Through Mathematical Analysis
  • Resource availability drives rapid, reversible mutation in non-coding DNA regions
  • Sensory prostheses and novel Nonlethal Riot Control Methods
  • Social algorithms for clique self-assembly, and applications politics
  • Equilibria and Optimality in Regulation of Vice
  • Subjective Information : Philosophy of Individualism and Consensus Reality in Information Theoretic Terms
  • Toward a unified theory of the sensory-motor control loop and its pathologies


Sensory Prostheses and Nonlethal Police Methods

Not too long ago (seems like ages, but some of us are still serving community service for unauthorized protest), 20 delegates from the self appointed most-powerful nations descended on Pittsburgh. I'm not... exactly sure what, if anything was accomplished in this G-20 meeting, besides raining a whole hell of nuisance on the Pittsburgh community. I'm not going to go into the details of the general awfulness and terror that ensued, you can find that on elsewhere. Instead I will focus on the intersection of neuroscience, non-lethal police methods, and sensory prosthetics.

Police arrived in town with all manner of goodies and anticipating war. These included the usual pepper aerosol bombs and rubber coated bullets, but also the technologically sophisticated Long Range Acoustic Device (LRAD).

The number of technologically sophisticated devices designed for non-lethal control methods is increasing, and research into the human nervous system both guides the design of such weapons, and also provides clues about how to guard against them.

The principle use of nonlethal weapons is to achieve a tactical advantage and a measure of control without causing permanent injury. I will group said weapons loosely into three categories : painless immobilizing techniques, methods of inducing pain, and advanced methods of sensory incapacitation. The first two are familiar and historical. Advanced, painless, methods of sensory incapacitation is a subject of ongoing research.

We can have lots of fun with painless immobilization techniques ( handcuffs ? rope ? nets ? barricades ? paintballs filled with glue ? foam bombs ? ), and have most likely not exhausted this category. We can also inflict pain fairly easily ( blunt force, tasers, pain stimulating chemicals in various delivery systems, bashing people over the head (these also seem to have the nasty problem of occasionally just killing people)). We are only beginning to hack the sensory systems to achieve immobilization in the absence of either pain or physical restraint.

The LRAD falls somewhere on the boundary between sensory hacking and pain induction. With respect to sensory stimuli, the LRAD is quite simple. It is very, very, loud. It is so loud that it will almost certainly irreversibly damage hearing. Its primary mechanism of action to to simply cause extreme pain in the more sensitive components of the human auditory system, and it is still a crude device.

High intensity strobe weapons are a bit more curious. They generally do not exceed the luminance tolerance of the eye, and so pain is not their main effect. Instead, they overwhelm and incapacitate the visual system, allegedly causing nausea and vomiting in some subjects. This strobe light incapacitating ray may well be the first true sensory hacking weapon.

How does it work ? As far as I know, that information is not publicly available, but I can speculate. The lights change faster than the human visual system can adapt. In fact, they change so quickly that the computational architecture of the retina and primary visual cortex ceases to perform its usual computation. As a result, the visual system is overwhelmed by artifacts of the computational structure : at the right frequencies, vision breaks down into noise and simple geometric shapes. Given a particularly unpredictable and destabilizing visual stimulus, the discrepancy between vision and the other senses becomes large. This lack of sensory alignment can cause nausea of a form not entirely dissimilar to motion sickness.

Can we conceive of other such weapons in other sensory modalities ? Certainly touch and taste are inaccessible remotely. This leaves audition and olfaction. Frankly I can't think of any way to hack olfaction besides a really terrible smell (thank you Sea Shepard). LRAD is a very crude hack of the auditory system, and barely counts. Perhaps a loud signal mimicing the statistics of natural speech can block communication in crowd control situations ? Perhaps an as yet undiscovered auditory analog of the flicker hallucination effect exists ? Or, perhaps a combination of auditory and visual stimulation in a synchronized fashion can geometrically amplify the efficacy of the strobe light incapacitating ray.

You don't think I was just sitting at home thinking of new technologies for hurting people while my friends were getting gassed, do you ? No, for many incapacitating agents there exist countermeasures. Sensory prosthetic systems can act as a general purpose perceptual filter, removing noxious stimuli and passing only useful information on to the brain. We are not yet in the age of augmented reality of this sophistication, but military research may drive us there.

A heads-up display with expanded sensitivity to wavelengths outside the visible spectrum could conceivably dodge the jamming effects of the strobe-nausea ray. By a combination of shifting to wavelengths that are not being jammed, and filtering out the flickering signal, we can reconstruct a useful visual scene. In short, Mr. La Forge is probably immune to strobe weapons.

An auditory prosthesis ( e.g. hearing aid ) can be turned off to prevent pass-through of the LRAD. Now, an auditory prosthesis attached to a digital filtering device may be able to pass through some sounds while still blocking the LRAD channel. At the very least, the prosthesis could clamp the audio to a non-painful level, and return to standard pass-through in-between LRAD bursts.

I guess I don't really have anywhere I'm going with this, besides perhaps that being a cyborg will render you immune to some riot control tactics.


20090920

psychedelia, contradiction, and theorem proving

Suppose we have a system which generates new facts from a body of axioms, derivation rules, and previously-known facts, i.e. a theorem-proving system. For this to be done tractably, the system much have some heuristic for exploring the space of possible facts. As long as it is functioning properly, the "shape" of the set of known truths within this space will depend both on the formal system itself and on this exploration heuristic.

Now suppose we introduce a logical contradiction into the system's body of known facts. Now all points in our space of facts are reachable. If the system had unbounded computational power, the entire space would "light up" at once. In a finitely-powerful system, we see an expanding shape, now determined solely by the exploration heuristic, i.e. by the properties of the theorem prover and not the thing which it is supposed to be studying.

This is the sense in which psychedelia is "mind-manifesting".


20090916

The Ethics of Nanotechnology

It's time for some serious social science. The EU Commission has released guidelines for ethical nanotechnology. Much of it is pretty sensible advice as to openness and encouraging the use of nanotechology for socially beneficial goals, but there are some sections that deserve critical review.

First, 4.1.11, that research bodies should standardized terminology. Nomenclature is more than word games, it shapes the ways in which think about the world. Calling for standard terminology would be an early step in the closure of nanotechnology. Do we want standardization to allow clearer collaboration, or are the many definitions of nanotechnology necessary for a diverse and expanding research community?

In the prohibition section, there is a call for not violating fundamental ethical principles, with new viruses given as an example. We should strongly consider the implications of military nanotechnology, particularly for targeted assassinations and robotic soldiers. Their call for non-therapeutic human enhancement is more problematic. I believe that we are going to see a blurring of the lines between medical (health is a social construct), commercially necessary (can you afford not to be as smart as your nano-enhanced colleagues) and recreational body modification. We are already cyborgs, why are nano-cyborgisms so dangerous that even research into enhancement technology is prohibited?

An aside on the topic of drugs. Commonly available mechanosynthesis would be a massive boon to drug users. Right now, synthesis of illicit drugs is controlled by monitoring their precursors. Mechanosynthesis uses the same elemental building blocks for all its products, so it essentially putting a drug lab in every home. How can we prevent personal fabricators, both macro and nano, from undermining laws that right now we consider necessary for society?


20090902

Do It Yourself : How to Build Hallucinogenic Goggles

This post will describe how to construct a pair of goggles which can be used to induce geometric visual hallucinations (1 2 3) via strobe light patterns. This tutorial should be accessible to anyone familiar with Arduino, and I do not cover details of the electronics design. These goggles can be constructed for 25 to 50 dollars, depending on how good you are at scavenging parts.

WARNING : this and similar projects have been known to induce seizures in susceptible individuals.


Device Summary

This device consists of three major components : a physical interface to provide the visual stimulation, electronics to control the physical interface, and code which governs the behavior of the interface. The physical interface consists of ping pong balls in swimming goggles with LEDs inside. The electronics are an Arduino pro-mini, and a few additional interface parts. The code is Arduino SDK C style driver code.

Component 1 : Physical Interface
Update : This later post suggests there might be an easier, faster, and more durable way to construct the goggles. The design posted here works fine, but is tricky and time consuming to make, and also fragile.

parts :
  • 4 to 10 Ping pong balls
  • 2 RGB LEDs, frosted clear casing (this is important, sand down the outside if not frosted)
  • 2 4x1 male headers, .1" spacing
  • 2 4x1 female headers, .1" spacing
  • 1 8x1 female header
  • 1 Dolfino medium sized silicone adult swim goggles ( had to buy in a 3 pack )
  • 2-3 ft of elastic ribbon
  • 3-4 ft of ribbon cable, only 8 channels required. Other cables with 8 channels also work.
Description :

Ping-pong balls, cut as if by a plane penetrating approximately 15% of the ball diameter, and rejoined with with smaller section inverted to form a cup like structure. RGB LEDs are affixed via solder to male headers which penetrate the corner of the ping pong balls (near the joint of the two sections). Light is emitted from the LEDs, reflects off the back of the larger section of sphere, and creates uniform illumination in the smaller cup. Two ping pong balls are nestled in a modified pair of swimming goggles. A ribbon cable connector is affixed with female headers which interface between the male headers on the spheres, and the male header output from the electronics. Note that logos or text printed on the ping pong balls can usually be removed with acetone ( nail polish remover ).

Tools :
  • one minute epoxy
  • superglue
  • soldering iron
  • solder
  • sharp razor
  • scissors
  • medium to fine sandpaper
  • wire cutters
  • tweezers
  • pin
  • toothpick, etc. for mixing and applying epoxy
Instructions :

Construct (2x) ping pong ball shells which are mirror images of eachother:


  1. Imagine the section cut by a ray displaced 34-40 degrees from vertical and rotated around the z axis. Alternatively imagine the section of a circle cut by an arc of 70-80 degrees. This partition defines the sizes of the large and smaller sections which form the spherical diffuser. You will not be able to cut both sections from the same ball, since material is lost in cutting, and a 1-2 mm edge is required for overlap to bond the sections together. Additionally, neither side should have a company logo on it, since this will ruin light diffusion. Ping pong balls have a ridge where the two halves are joined in manufacture, avoid cutting through this ridge since it will create an uneven joint that will prevent the balls from being re-assembled. I don't have exact measurements, but on my model the diameter of the circle at the interface of the two sections is 1.365"
  2. Prepare the larger section first, as described above. With a razor, cut a crude circular hole in the ping pong ball, perhaps circumscribing the logo if one is present. Slowly and carefully expand this hold by cutting around its circumference with a pair of fine scissors. Stop approximately 2mm from the final desired hole. At this time lightly sand the hemisphere on a flat piece of medium to fine sandpaper to create a fine, flat interface.
  3. From a new ping pong ball, prepare the smaller section. Cut the ball crudely in half using a razor, then carefully trim one half down to the intended size of the smaller section, plus 3mm.
  4. The smaller section should rest in the larger cup, and be large enough not to fall inside. Do not glue the sections together yet.
  5. Using a pin, create evenly 0.1" spaced holes for the male header in the larger section as shown. You may want to practice on a spare bit of plastic first. Insert the short end of the male header through these holes, and super-glue the header in place. Trim the LED leads so that the LED rests as shown, and bend down the last 2mm of leads to align with the inner header pins. If you do not have frosted housing for the LEDs, lightly sand the exterior of the LED with fine sandpaper. Clear housing creates light that is too focused for uniform diffusion in the eyepiece. Tin both the LED leads and the header pins in advance. Solder the LED onto the header from the inside; do not to melt the plastic. Super-glue the smaller piece into the large piece to make a finished eyepiece. Once the super-glue thoroughly hardens, you may want to finish the joint in the plastic with additional careful trimming and fine sanding ( don't sand through the joint )
  6. The final pair of eyepieces should be mirror images of each other, which is just a matter of correctly positioning the LED leads :

Construct ribbon cable connector:

EDIT : this is a terrible, tedious, way to build a cable. The correct way involves some sort of headers that are actually designed to clamp into ribbon cable, or using these little header connectors that use pins which clamp onto the wire (pins sold separately ?). I will post a writeup if I build a pair using better technique.


I found that it was important to have a separate cable that would disconnect from the goggles under force. This prevents the inevitable accidents from destroying the tediously constructed eyepieces, and modularity makes the whole thing easier to repair. This step is open to improvisation. Here is what I did :
  1. Tear a band of 8 lines from a section of ribbon cable. The cable should be as long as you would like the strap from the electronics to the goggles to be. I think 3-4' is fairly good.
  2. Cut the ribbon cable diagonally such that the spacing between the lines matched the 0.1" spacing of the 8 pin female connector
  3. Strip 2mm bare wire of each line
  4. Solder the line to the 8 pin female connector. Tinning the contacts in advance helps.
  5. Apply 1 minute epoxy to the contact, to provide both insulate and structural stability. Make sure there are no shorts between lines before you do this.
  • EDIT : Hot glue works better here, for a number of reasons. Hot glue remains flexible once cool, which allows for smooth transfer of strain on the cable without breaking the contact. Epoxy hardens, which results in an inflexible interface which slowly cuts and degrades the cable. Breaking of the cable, as well as squishing of the ping-pong balls, seem to be the two most common failure modes of this design. If anyone knows of any commercial connectors that would be better for this design, let me know.
  • Tear the line in two for ~1.5', creating a split from 8 lines to two ribbons of 4 lines. Prepare 4-pin female headers similarly to the 8 pin female header, in a symmetric fashion as pictured below. I used a clip that came with the swimming goggles' strap to stabilize the point where the cable splits in two.
  • The assemblage of this connector cable with the eyepieces should have the indicated pinout at the 8 pin female header :

  • Modify swimming goggles and complete physical interface assembly :

    1. Locate suitable swimming goggles. This is harder than it sounds. The only goggles I found suitable were the mid-sized silicone pair in a three pack of Dolfino goggles. The goggles must be of a correct size to snugly fit the eyepieces, and be able to deform to the circular shape of the eyepieces. The goggle must also be able to hold together with the lenses removed. Many goggles are bridged by an attachment to the lenses, rendering them unsuitable. Ideally, you would also be able to affix a strap to the goggles even with the lenses removed. Due to the limited availability of suitable goggles, this step may require improvisation.
    2. Remove the lenses. In the pair I used, the lenses were held in with a weak silicone glue. It was difficult to remove the lenses without damaging the goggles. Superglue proved effective at repairing large accidental tears in the silicone goggles
    3. Attempt to insert the eyepieces. If necessary, create an opening in the silicone to feed the male headers though. I used either a razor, or a hole-punch, depending on the thickness of the silicone. Insert the eyepieces.
    4. Create a head-strap. I used elastic ribbon, threaded through the hole used for the header pins, held in place by plastic loops, and super-glued back on to itself. One end was folded and kept free to adjust tension.
    5. Attach ribbon cable headers to the eyepieces, check that you have oriented the ribbon cable pinouts correctly.
    6. If the eyepieces are loose, optionally super-glue them in place to the goggles. Note that this will make repairs and maintenance more difficult.


    Component 2 : Driving Electronics

    Parts :
    • 1 Arduino pro-mini ( and FTDI breakout for programming ) (other options 1 2 3 4)
    • 1 6x1 right angle male header
    • 1 8x1 right angle male header
    • 2 12x1 straight male headers
    • 6 Resistors for the 6 LED channels as determined by your board voltage and LED datasheet (voltage, current) specifications. Use this handy LED resitor calculator. For 3v boards, a resistor may be un-necessary for the (green, blue) channels.
    • hookup wire
    • 1.5"x2.5" radioshack protoboard
    • Battery Pack
    • Power Switch
    • optional : LED displays, pushbuttons for a hardware user interface. I used a 16 segment display for some of my models, and a couple designs have pushbuttons to cycle through the various strobe light patches.


    Description

    There are probably a million and one ways to make 6 LEDs blink quickly in a controlled fasion. You can drive your LED goggles however you wish. I used an Arduino because the programming interface is easy to use. I also hope to figure out the serial interface to the arduino so I might write a control sketch in processing, for real time tweaking of the waveform patterns. My construction consisted of an arduino board, with the 6 pulse-width-modulation ( PWM ) output pins attached to the LED goggles. I also attached a 16 segment display and some push buttons to the design, but you can experiment with whatever features you wish.

    Tools : Soldering iron, Solder, Soldering accessories

    Assembly of an example control board :
    1. Since the arduino chip rests on raised headers, and the 16-segment LED display has ~1.5mm clearance, we can hide some of the circuitry underneath these components. Since this is a 3 volt board, I only needed 56ohm resistors for the red channels. Your LEDs and board may have different constraints. Also solder on the 6x1 right angle male header to the Arduino pro-mini serial FTDI interface ( I think thats what those 6 pins are called anyway ).
    2. We then solder in place the arduino chip and LED diplay. The LED display is set up for multiplexing, so the corresponding segments of each digit are connected, and the display is driven by alternately drawing both digits, controlled by switching on and off the common cathodes. Since I was short on pins, several display pins also double as input pins for the switches. Every so often, the sketch switches the display pins into read mode and polls the state of the buttons.
    3. I used a lot of tedious surface-mount style wires on the back to keep the design clean. It took some practice for me to get used to this type of soldering. Attaching the battery pack and power switch is not shown.


    Assemble All components :

    This is open ended, Experiment !. Prototype your design on a larger Arduino and breadboard. Tweak the driver code to your preferences. Make a more permanent device using your favorite prototyping technique ( Or design and order a custom PCB ! Please tell me if you do, I'd probably buy a couple! ).







    Component 3 : Code

    I've put some files up on Sourceforge

    Arduino sketch

    ( and here, another example adapted for the AtTiny13a)