AI as a Service

The product development starts soon, from the initials done over the last few weeks. An AI which has the aim of being more performant per unit cost. This is to be done by adding in “special functional units” optimized for effects that are better done by these instead of a pure neural network.

So apart from mildly funny AaaS selling jokes, this is a serious project initiative. The initial tests when available will compare the resources used to achieve a level of functional equivalence. In this regard, I am not expecting superlative leaps forward, although this would be nice, but gains in the general trend to AI for specific tasks to start.

By extending the already available sources (quite a few) with flexible licences, the building of easy to use AI with some modifications and perhaps extensions to open standards such as ONNX, and onto maybe VHDL FPGA and maybe ASIC.

Simon Jackson, Director.

Pat. Pending: GB1905300.8, GB1905339.6

Pico ZX Spectrum

I’ve started making my own edits of Pico ZX Spectrum as I found it interesting. A bit of preparation work was done. Like adding a custom port and adding save state routines, a new interrupt mode for some logical sense (finally) on the “Speccy”, and a counter register in the sound chip for video sync from the vertical blanking by observation polling for specific count values. The timing is too tight for beam following.

There’s still more to do. I did a video mode enhancement (256*128 in 8 colours), perhaps a new ROM, architecture extensions and some extra Z80 instructions. 


“Azgaz. Propane and protein suppliers. Beans and ‘as beens for beings plus gas. … This has to be the best winter supply name.”

“Could later expand into biodigester funerals, propane tank cold boxes, efficient air Peltier blower rings, eggs and a coffee shop library warm room.”

I don’t think it’s as daft as it sounds. Basic base staple and maybe some quality brown wholemeal seedy bread. Perhaps with a carbohydrate warning. I suppose the funniest part of the vision dream was the “Posh beenegi”. Not so much the content, but the price list in LED, and a rolling supply demand accounting for running the cooking and service. Not your basic beans just run at supply cost and the miracle of prophetic markup weekly spot pricing.

You don’t know what was on the time since cum joke laugher discount line.

Looks like some kind of summer supplies line is needed. Onto thoughts of efficient freeze-dry desalination. So pressure in the hot tank is lower than in the cold tank. So the dual hydraulic opposed piston is balanced on an escapement of area. The compression hot end is routed through the hot tank, and vice-versa for transfer of heat by the oscillation of the escapement.

The final complexity is the salt accumulation in the hot tank. Some kind of screw extraction for the sale of “complete dried salt”. Likely requires some desalinate, clean cycle. So salt and distilled water? Or should that be remineralized water? I wonder how the “pump end inversion” with “routing to the other tank” affects efficiency? The S-balanced escapement could be thrown by a solar power linear actuator and some simple control electrics. With the sides of the tanks, it might look like a posh $ sign.

Accounts Time Again

Did you have a COVID gap year? I did. Accounts have been all “out out” this year and not in a good way, but not excessive in the throws of the corporate vision. I can’t do any services for clients who can’t resist fraud scams as there is talk that the government is pondering that service providers might be levied to assist banks with their fraud losses. Unfortunately due to the nature of scam design probing and thought ending “… but that’s just defrauding the tech company.”, it will not be possible to fund the “fraudy” habits of customers unable to use product or service.

Protected Modes

Ah, the classic is data execution prevented? It stands to reason that return-oriented programming would like to execute data as code by placement on the stack for a return. The obvious solution to this is a split of the stack into two stacks, one for data and one for code. Tagging all registers in the processor as containing data or code, and having a protected mode “make code” instruction, along with enough instructions forcing an automatic demotion to data would then place on the right stack as the store register would have a target kind tag.

This exceeds regular page virtual memory protection, and the supervisor bit makes for 4 stack pointers. This might be suitable for incorporation into a RISC-V processor but that all depends on the specifics of what register is the stack pointer (convention dictates sp is x2 but this can’t be certain in hacker code assemblies). And precisely how is target setting the alternate stack pointer is achieved.

In this way, it would be impossible to return to a data stack item and hence perform code address arithmetic on stack items. As per the RV64G extension practice of RISC-V, the custom-0 space is assumed until any ratification process solidifies the instruction encoding and so SXi is the extension name. 

It might be possible to fit the whole functional extension into the M extension space funct3 remaining RV64M combinations. 001, 010 and 011. These are all dual input single target register operations in R-type format. Knowing that demotion of code pointers to data pointers occurs on arithmetic and some other instructions (to be determined) automatically it needs documenting but no extra demotion instructions are necessary. They all set an internal state such that mcause is set to illegal instruction (2) when not in a safe supervisor mode and there is an indeterminate condition on which of the code or data pointer registers is to be used.

So each register is marked as containing code or data by a bit, and each integer register is duplicated for when it is used as a load or store address. It is chosen to address from the pair (regular and code shadow) based on the bit classification of the datum source or target value register (regular only).

001: This works opposite to 010 by moving a code pointer to a data pointer with an offset (rs2 slot). This is not protected obviously as it becomes data.

010: The instruction works like an add where rs1 is the data register, and rd is the code register with rs2’s slot being a 2’s complement sign-extended offset immediate to add. It is protected by supervisor exception as certain code targets are what need protecting.

011: This is a convenience instruction to move a code pointer to a code pointer (shadowed register for each general register when used as an address for all loads and stores while the stored register is flagged as code and not data), with an offset in rs2’s slot. It is protected by supervisor exception.

Yep, the 32 classification bits need to be stored somewhere, and an extra status register is needed. It has to be protected of course and should be called mdata.

Memory protection through the page table then isolates the differing actions of code and data pointers used for stacks. x0 still returns zero when used as an address (from the shadow value). JAL and JALR then use the stacking information too to target a PC as code and the correct stack. It could be considered a supervisor exception state if the link register is data.

Looking for open security analysis.

Gradient Optimization

So if a gradient descent hyper-parameter controlling the learning rate is the usual way, how can this possibly be improved? Considering that in some way the approximation of future gradient alterations is distributed depending on the batch, the stability via an average gives a more stable basis to then infer an accelerated projection of the future descent.

The biggest problem to consider is bound oscillation. When the accelerated projection is magnifying the learning delta to apply such that locality is an asymptotic non-convergent (reverse symmetry in summation acceleration by considering the divergent terms as “merging toward” the first term limit). This then would converge as a metaseries in some instances, but not all. It then becomes essential to scale the approximations by inverse power weighting to make a convergent for highly entropic unstable weights. It may also indicate that weight decomposition may be an effective strategy to obtain a neuron split into the stable (time aligned) and the unstable (time inverted) partitions of a signal.

Assuming the unstable partition has a repellor (opposite to an attractor in chaos), modelling could be used to invert the accelerated projection to the repellor. If the accelerated series is approximated by an integral, the unstable inverse acceleration would perhaps be a reversal of the limits of integration? Or a sign reversal of the limits?

In a sense the splitting of the network into a composition of multiple networks based on partitions related to the number of critical negative signs (or more precisely the number of things that could have negative signs). In this case just 1 sign for a time is like hyper-parameter convergence property. The algorithm after decomposition can then be specifically optimized per partition.  

Pi Pico

So I got a cheap Pi Pico as the postage and packaging were more expensive than adding in a Pi Pico to exceed the minimum free P&P limit. It’s about a 2 MB flash drive with a dual computing core and some simple GPIO. It looks like it can do about 500k sample/s ADC over 3 channels at 12 bits, so some audio project seems like a good idea to try.

It has some Amiga “copper” style coprocessors for IO too, so making a video raster scan is likely easy too. It does have enough power to simulate an 8-bit core with a spare CPU left for other purposes. At 133 Mhz that’s quite an efficient bit of silicon area. A 1980’s super-computer with slightly less vector parallelism and a bigger (smaller) storage media. Bargain!

Moonshine Elliptical

Moonshine Elliptical represents my latest combination of commentary and finding on the massively impressive tome of knowledge about elliptic curve theory (useful in cryptography and generally understanding space and time). Fields of character (the summing of multiplicative identity to equate to the additive identity) two, three and other along with factorization of parts of the world mechanic into finite simple groups (the extended concept of primes of which the primes are just one sequence).


Future Prediction by AI

So given that the future estimation could be trained on data from a delayed assumption state from the past prediction of the present, then what is missing? The missing seems to be based on the time factorization process NP problem and innovation stimulus which would cover things that are unknown within the net as well as time relevance which was not compensated for (the delay has an opportunity to sample lesser pasts for greater present prediction but produces nearer futures without doing Monte Carlo assumptions for a spread).

A subnet could be trained to do the estimations of the best assumption for such a predictive engine, leading to a trainability for an expected spread entropy (a situational requirement of MUST and or ANY as GOOD) given a similarity measure of an output of training to a random network spread RND classifier. https://www.youtube.com/watch?v=z4lAlVRwbrc is an interview with an author on an interesting paper about AI exploration. This covers the RND idea in a use case. Training a post RND latent space map to merge lingual or other equivalent factorizations of the novelty could be part of this.

The reevaluation of situational state novelty then can become a post addition of a trained residual based on the expected future estimation and the purpose to which the predicted estimator is to be put. Imagine on a stage pretending or on a real battlefield. The eventual motor actions of production to have for benefit?

Boltzmann-Fermi-Dirac Colour Charges

It’s a long shot but imagine if you will a gluon made of two halves. The halves can each be drawn from the two “weights” (low and high of a non-zero-sum field) if they broke symmetry and the two “charges” like from the zero-sum field.

Given cancellation and combination, 8 gluons happen.

L+L+, L-L-, L+H-, L-H+, L+H+, L-H-, H+H+, H-H-.

So there’s more green weight “sticky” and the Boltzmann distribution for the half Bose-Einstein as a Fermi-Dirac perhaps. The blue colour perhaps travels less far due to higher “mass” (if it splits), but as the energy input in the strong force makes more gluons at a critical threshold, the further interaction has more energy and a less gluey implementation in blue.

I wonder if the QCD simulation evaluations can take this all into account for better accuracy. I put the two yellow “charges” in there which technically would be massed green, but given the charge +/- cancellation without perhaps LH equality would suggest a kind of neutral weight dipole.

EDIT: As the energy increases moving into “small x QCD” the expansion of colour coverage, to prevent saturation by a UV gluon density catastrophe, the critical temperature is exceeded and the “cooper pair” effect on the half bosons is removed and Pauli colour saturation removes the gluon density within the nucleon. Yes, this paragraph is unproven, but there must be some effect stabilising the UV catastrophe. This would also lead to a cyclic order of colour based on mass expression above the critical temperature for some critically small x.

Minecraft Modding

And so I downloaded Forge for 1.18.1 BETA. I’ve got the basic details down, and hope to convert some tutorial module to 1.18.1 and then set about making some extras to try out. ExactFeather396 is the github repo, and it will develop slowly. I seem to have a Microsoft username.

It seems quite a lot is JSON these days, and sometimes the names of methods change a bit. Hopefully this is not too weird to convert and make into something. Some AI mob redstone thing? Who knows?

EDIT 2022-03-01: Ah, so that’s how potions are made! And the RegistryMap class might come in useful later. I’ll have to make some of the classes final and private or default some of the public classes.

Just analyzing the base code at the moment to make it adaptive and have minimal technical debt. I might abstract off some of the names.

Seems the rendering of Mobs is somewhat complex. So I’ll have to have a look see. A basic Zomie clone reskinned seems easiest. I’ve started on an AI exception mechanism. It starts with BaseCodeException which fires the instict emote() when it reaches the base code, which proxies the actionTry() and actionCatch(BaseCodeException) for consequential instinct.

Ah, exceptions and encapsulation within a RuntimeException to avoid the dreaded can’t override method with throws extension, and having type checking and catching.

EDIT 2022-03-09: So it has been simplified a bit and made more complex. I’m making it so that various Loaded classes via extension are invoked via a static and then an instance of self passed to it allows the invoke dynamic override while avoiding lots of casts to super classes.  This may look more complex but it does allow easier pull request merges by keeping things in seperate files to the mini module level.

Added in a simple potion system I’ve yet to test.

VCV Rack V2

VCV Rack version 2 was released in the new year and the SDK has also been released. I’m making KRTPluginA available for it (as an update 2.25.27), as it is a simple edit update and making some better documentation. I’m in the process of doing the simple API 2 migration. The text to write though is more of a complex decision.

When the phrase “one line description” bumps into a review platform wanting better and perhaps longer description with “\n” literal line splitting (as there is no tool tip wraping). Basically it’s just adding in tool tips and slightly better description and riddance of some spelling mistakes.

So thinking of what I can do for “Z” as I’m going to call it. Something related to some interesting maths so it gives me a reason to read a bit more.

The New Year

Now if only I could find out how to spend this Clubcard voucher I could have a belated New Year’s out off office, on my own special company bash. Seems both previous attempts have ended in “Not this product range”, and “It’s been used” and other fabulous excuses from the gobshites. I did get a I don’t know, and call a “busy” helpline as they also seem to be a fan of draining mobile batteries so you can’t use the Clubcard feature in the app when your phone goes off.

Alcatel Longer Term

So it does do the WiFi network “sign in”, but not always (with no listed error), and took an obsurd “got to do a spy search” amount of time to agree to format the recycled SD card. … Google picture uploads will be charging from 1st June. Coincidence or ET finger plug?

Proper Crimbo

Hi, and greetings from a warm Ikea after leaving the slowly sinking wet, cold boat. Let’s all hope for the “frost” not happening this year with wet feet marine one being a bit shit.

So I got a new battery with a phone around it. The old one had reached the end of service with cold failure imminent. So I got an Alcatel 1 (the cheap dual SIM unlocked one), and it is quite nice for the price, and the battery lasts as it should. I’d recommend it.

It fits a surprising amount in the 8GB, due to android Go and little bloat. You must check if you already have some Go versions installed of some of the old bloatier apps (like Maps). Bits of it are to the cheap end of the market standard, but functional. It is available in other versions with a better camera and more memory, but I don’t need those. A good calculator, YouTube and varous social app lite editions. It still has enough space to keep Chrome, and install Libby for audiobook reading.

I suppose I’ll find out how the settings have been reorganized, and how to enable debug mode. One of the 7 taps on the build number in the system settings. OK.

James Webb

They put a big telescope into space. Perhaps they’ll find some intelligent life. Or maybe the intelligent life will broadcast the bunny ear fingers of the thing galaxy wide for a bit of Crimbo fun.

QMK Link Time

So having got the new build process working as expected (some python ~/.local cache storage issue), link time optimization provides almost 4 kB for future features. Latest master build this is quite a lot of feature space when considering such a small microcontroller. A nice feature added to the more modern build options.

I should be able to put together a nice Crimbo update with such a gargantuan amount of code memory, and no need to duplicate features the computer should do easily. In some ways the difficulty is making something extra which might get used.

Docs for amperzand is a little experiment in server languages I’m thinking about. Anonymous inner class? Sounds like it might be cool. So there is still about 3 kB left after double striked letters (for drop capitals perhaps). So some 46 macros to define before I have to get a little more creative in what’s available.

And now added automatic South Korean sylibilic combintions.  

Ltd. Still and QMK.

Yes the QMK active branch and some news that my accounts are now filed. Zero in/zero out as a boring COVID an low contact availability year. So Ltd. status continues as far as I’m aware. I’ll keep you all informed.

So now the send_unicode_string() function is used for a macro system within the keymap latest coding. This is opposed to how the macro layer emits function key combinations, which is more in line for a tool on the computer handling it. I’ve also added repeated substring compression too. (“\\0” to “\\9”)

So more of a hard baked solution, but does allow more complex multi-character glyphs to be produced instead of just one Unicode code point. So with about 1100 bytes free about, that’s about thirty UTF-8 bytes per key action (over the 37 defined key action macros). Even “\\\\” is defined for emitting an initial backslash, just so backslash can be used as a prefix for more complex macro processing than just print until end of string.

Other macro features like tapping key codes for nested macros? Yes, if the keycode is added to the array so “\\A” will tap the first keycode.

QMK Keyboard Again

Latest 2021-11-22 commit goes for an 10 layer design with the language BQN built in and three further planes of Unicode glyphs. This leaves 2296 bytes of firmware left for further adaptions. Seems the lock key option consumes quite a bit and I don’t need it.

Altering the U’ ‘ defines such that a Unicode glyph is copied between the single quotes would add a Unicode character to the design.

The control iconographs for example:

[IAT] = U'⚠', [IA] = U'⟁', [IB] = U'🗚', [IC] = U'🗐',
[ID] = U'🔖', [IE] = U'🔎', [IF] = U'👍', [IG] = U'🔔',
[IH] = U'⌫', [II] = U'⭾', [IJ] = U'⏎', [IK] = U'⭿',
[IL] = U'📇', [IM] = U'✓', [IN] = U'🗋', [IO] = U'🗁',
[IP] = U'🐧', [IQ] = U'📤', [IR] = U'📥', [IS] = U'💾',
[IT] = U'🌱', [IU] = U'👎', [IV] = U'📋', [IW] = U'🔑',
[IX] = U'🗙', [IY] = U'🗜', [IZ] = U'⎌', [ILBR] = U'⎋',
[IBSL] = U'🌍', [IRBR] = U'☣', [ICAR] = U'⚗', [IUND] = U'☢',
Are picked to represent general principals of the control characters in a modern computer environment. Some of them may be difficult to understand at first but for example the last row could be considered ECO/BIO/CHEM/PHYS, on a atomic building.
Deciding on the extra control layer glyphs as they don’t have ASCII slots but are possible to type is a bit more complicated. I’ll give them a bit more thought. 
There’s an interesting VSCode crash 139 development (SEGMENTATION FAULT) just occurred which shouldn’t be happening but is a crash in the rendering process. Obviously some “bad code” in VSCode?
I’ve improved the shift mechanism for some of the extended layers, and filled in the ANSI control code layer to my satisfaction. I’ve finalized the Navigation and Macro layers to final satifaction and added a number lock on the Magenta shift of the Macro Yellow layer.

K Ring Technologies Unlimited and MOND Galactic Equivalence

Due to the facinating non-working companies house service for email notification for upcomming filings which are needed not working in the COVID period I have miss the accounting date and K Ring Technologies Ltd, is to be struck off the companies register. All business is therefore to become sole trader until such time as sense resumes.

But apart from that I’ve extending the quantum uncertainty in a gravitational field idea and come up with


As a dipole expansive explination for dark matter and thruough the singularity of the first term an eventual repulsive dark energy kind of force. In the dipole limit in a galaxy for example the force will aproximate 1/r and so is effectively a MOND on the small galactic scale.

Looks like the Net needs a reflective NAK uptake NAK shutdown.

And so some arbitary bollox began. Take self query and accept delay of self accessing you to provide NAk nothing or shutdown as your proto want DoS attck. End, happy music continues. If bought net percent, then for purpose?

Seems 17/9/2021 has a dislike of digital flux to elsewhere. It’s surprising how a 10% NAK request packet NAK happened before could force upon the uptake of the promotion of the internet vesrus those which refuse cache and just want DoS attack without in a sense showing provido.

Why do cells recycle approximately 10% of themselves toward an imune system which kills the not of body? Why does the engineering of telecoms having a standing 10% use of bandwidth to achive effective 90% use goals?

It’s not my number, just the IP/ISP stack. It could just be the sound of Musk and the super-titty highway, but not likely?

VCV Rack Again

Now that VCV Rack virtual modular synthesizer has stabalized I tried it out for developing C++ modules to add in to the rack. Haven’t decided on the exact nature of the modules yet (2021-7-14), but it does work for the developer better than the older version. First a full source compile is not required. There still appears to be some issues with control alignment in the module GUI which is auto generated which is easily fixed by adding 28 to the Y coordinate of the control. My advice is leave the control “circles” visible until the actual controls cover where you intend them placed.

KRT Plugin A is the repository. So module A seems to be a filter with some strange DSP and input option for a HPF/LPF ring modulation with a metalic on the corner frequency offsetable from the main Q frequency but with tracking added on independantly. Is this the most fun that can be had with 4-poles? An electronic DSP filter joke. So the test basically works, but some DC on none HPF path, but as HPF is by inverse LPF and DC cancels, there must be some DC injected somewhere. OK, found the obvious error at last. Filter A is finished.

And now polyphonic with SIMD. This makes it about 7 times more efficient. Now has graphics. Developing module μ for calculus purposes, then onto a few other niceties. module μ is finished too. Any errors in the calculus should be reported as this module is about a calculated sound, and errors sounding better are for other modules.

On version 1.2.3 already 😀 as it even took this long to define a suitable versioning number system. At the moment controls are virtually CNC’ed on a grid, and panel graphics are manually kerned (as auto(Tm)ating this is perhaps overkill. v1.3.4 release now include the fat T with some bad disharmony as well as lovely 4th and 5th sync sounds and stuff.

L;D and R are now in planning. And are completed for the 1.6.9 release (2021-07-27). Maybe some speed optimizations and next a more complex module. A nice website is also being made in markdown here as that’s what is expected by default .json tags.

The 1.8.13 release (2021-8-2) includes 8 working modules with the new ones being Ω (a clock distributer with randomness) and V (a VCA triplet). The 1.9.15-rc2 relaease (2021-08-05) includes F a morph filter. Hopefully goes live soon soon when compiled by VCV. The Y gate sequencer is almost ready for 16 channels of triggers.

15 machines up there now (2021-8-19) including some oversampled ones, and some utilities for helping out with problems you didn’t know you had. One minor fix for the F filter will be in the next release, and some slight improvements, plus another 3 modules. 


Recursive Predictive Neural Networks

Given that the output of a neural net can be represented as y derived from an input x and a feedback operator f(y) the network can be trained on which may include differential and integral operators in the operator f. As f(y) can be considered to be the feedback synchronization point which is clocked to transit the network forward in prediction, f(y) is delayed in y such as to be f(y(t-1, …, t-n)) is the applied feedback to stop “epileptic oscillation” of the forward net function.

The network itself can be programmed on the sequence to learn in an open loop gradient decent and the bias of x activation to f(y) remembrance by either weighting or digital percent application gating. The pattern to lock onto for an input can be trained independent of an input, and then offset by application of the triggering input to balance activation of one output versus another. The actual spreading and maximization of the output attractors becoming disjunct from instancing which attractor to present as output from input.

The “old” feedback from the “last” remembered thing introduces some chaos and mal-attractor effect. This can be removed a little by using an expected previous context training pre-sequence. This can then also introduce contextual recall. The “short term memory” being the contextual state of y, so programming the long term sequence prediction memory with context y and stimulus x.

The production of optimal context for stimulus itself become a network programming challenge. It represents the concept of changing predictive utility. As the forward transfer of the network produces the output to feedback, the network itself could produce the optimal context from the requirements delivered through part of x deciding the contextual decode mode. A separate net to organize the change of context in bulk would have specialization separation and generation of terms in parallel advantages. In utility though it would only be used to switch contexts, or cross imagine contexts to place the prediction net on a creative sequence.

This could have application when the context is considered a genetic algorithm process for tuning the network to produce some kind of granular attractor synthesis. The process of providing the scoring feedback in synthesis mode controlled by a hardwired concept of misadventure excursion in the prediction. Another network for bad state recognition to complement the entropy generative context granularization network? So the reality predictive network is contextualized, granulated and tested for productive futures. Then a final factorization of synthetic addition requirements of the imagined product can be performed by a final independent network.

Consciousness is within this last network as the self image of adding self as a possibility factor. The production of a threshold of motor action to produce an attempt at achieving the estimated reality granularization (subject to bounds constraints) being the primary motivator.

A Speech Action Co-ordination Domain

If the input x, and the output y with feedback descriptions, current “genetic” gene combinators and more can be serialized as a inter AI language, the projection of multiple “conscious” entities in the predictive net of reality simulation can engage in a factors for product optimization as well as other non zero sum optimizations. A net to process one internal representation to another with an acknowledge of simultaneous state with confusion feedback. At higher data rates a negative acknowledge protocol can take over with estimations of animism action between confirmation certainty with residual accidental error bounding.

A Survival Function

The selection basis of the context provided to the reality estimation can adapt to return a higher valuation of the survival “situation understanding” function. This in the real sense is the optimization function for selection of purpose. The reality function just attempts to maximize a correct simulation of reality. The context function attempts to maximize use of granular entropy to increase the coverage range of the reality simulation to increase options of consciousness to action. The action threshold function then decides if the likely action chosen is done, and in a way represents a kind of extrovert measure of the AI.

Component Parts

  • Reality simulation (estimation)

  • Reality factorization (situation)

  • Granular imagination (context)

  • Action selection (desire)

  • Input processing (percept)

Using some kind of Fibonacci growth connection in a surface topological toroid? That would be more on hardware interconnect optimization. Of more interest to the feedback in the reality simulator would be the parametrized operators building differential and integral representations from the feedback. Of the three forms of end point integral, all could be represented. The fact that the log kind has complex series to evaluate, and has no necessary complex log representation might be an added difficulty but would “lock” onto such functional time generatives.

Negative time offsets on the end point limit on such integrals when complex processing is applied introduce the idea of the 2*pi synchronous summand based on angle, as this maybe a better input controlled output representation of the complex domain for an N:1 mapping. A Gaussian distribution of error about the coefficient division.

Chaos Measure

The feedback operator f depends on calculation of differential and integral functions based on weighted sums of y at various t and so it could be said that any initializing or changing of the reality simulation to another play back “granule” has some new data placed in the feedback memory. This new data can have a varied impact based on the likely-hood estimation of the time samples having an impact on the calculated differential and integral values along with sensitivity to the feedback signal. This implies each memory bit has some measure of bit change (in a genetic algorithm mutation) on the divergence from the reality simulation. This then can be used to infer a focus mask. The use of gene crossing focus weighting or masking then synchronously produces a chaotic deviation from the training reality.

Modulation of the stored memory context would appear on some level equivalent to altering the coefficients of the estimates for differentials and integrals, but as the chaos measure is a deviation control from an exacting physical model of time evolution, it is thought better to keep the operator mathematics at a static precision, and deviate granularity by memory modulation.

For example 1, -9, 36, -84, 126, -126, 84, 36, 9 are the coefficient to predict the future next sample from the previous nine samples based on a zeroth differential estimate. In open loop training the feedback would introduce a delay step, but prediction of the future would in effect cancel this delay so that effectively the f(y) does not have to be calculated and y can be used. The large range would create some oscillation as the context shift registers were filled with data to feedback. This open- loop programming without reference to f allows pre-training without any feedback instability but with a later oscillation about the manifold.

Computational stability requirements are improved if the feedback f is amplified by default expectation, as this forces some non-linear mixing of x to reduce the net summand, moving the bode point of the feedback away from the inactive denormalized zero value. It also increases the net feedback applied to keep the reality simulator feed forward gain below one.

All n orders of differential can be cast as future predictions, and all the integral accelerated forms can be represented with future casting into any t with some renormalization possible but not essentially a necessity. In fact a rectangular offset in the y-axis integrates as a ramp addition to a monotonically increasing sum. Can the network learn a root finding algorithm for applied integral time when wired with learnable pass through of a variable integration time? This time offset from the future prediction time (integral offset time) u can be fed into the operator f and passed through as f(y(t(n)), u(t(n))) with some of the prediction y being used as u.

Alias Locking

In any synchronous DSP circuit with non-linear effects the requirement to keep x and f(y) within the frequency range where alias distortion would potentially present as false signal does indicate that the coefficients could be modified to provide an alias filter. But it maybe found that a small chaotic dither dithers the aliases further and leads to a wider band spreading about an alias. The detection of a coincidental alias may aid detection of the signal expected. This extra minimal noise could be extracted from the environment by deviations from expectation. An AI task of removing aliases may be considered as something that could be learnt, but also generating an inverse filter to supply the alias spectrum (excluding sub-harmonics of the clock rate).

Consciousness as the Correlated to Self Action

When the self action of the model produces a correlation in the reality simulation it could be said to have observed a correlation to self in the model. The relation to the situation factorization domain then becomes an obvious connection to equation of virtual actionals given the real actional set. This allows futures, and past observational training. The weighting function of physical error cutting a cookie of size survival plus some splurge.

So it seems “pain” or some milder proxy for bad function should increase situation recognition, reduce recent action, increase the accuracy of reality simulation, improve the percept and perhaps change the context toward know safe positives. An autonomic bypass from the percept to counter action is likely also “grown”.


The situation analysis net is likely better functional with some feedback. The purpose of this feedback in not time evolution estimation like in the reality simulation, but the use of the factorization of the situation in building a system of meta situational analysis which could include self consciousness. Technically the feedback could be nested recursively and be applied as part of the x input of the reality simulation, but that makes for more complex training. 

Considering that many factorization domains have a commutivity structure it implies that post convolution might be a good way of splitting the network result into “factors”. This is placing the convolution as the last layer and not the first layer.

Or FFT for that matter, and in some sense, this layer becomes the first layer of the action decision net of desire.








And the variational encoder ratio for optimal mixing of the networks?




Variational auto-encoder. Maximal representation of externality. Normalization average.


Time evolution feedback via calculus operators.


Produce genetic algorithm modification for estimation feedback.


Variational auto-encoder with post convolution or ideal order factorization of variation and causation tree.


Threshold action sequencer. Classifier with threshold.

The unity of consciousness as that identified with the knowing of multiple action paths in the imagination as capable of altering a future percept and certainty in achievment of a happy context and situation.

This extends on to the idea of emotive functor attractors as the controlled mechanism for genesis of output from the actional desire. This separates desire as an actional devoid of emotion, in complex with a driving emotion set. What has become of the splurge of biological evolute on the smudged cross product? Does it really assist functional understanding of the power efficiency of self action?

The situation analyser in performing a domain factorization, applying a feedback and estimation of a rule and a correlative later situation could in principal assist with modelling from rule followed by implication of rule. The Gödel incompleteness of the inferred logic controlled by “your stupid” and the implicant “fix yourself” as a splurge cull.

The convergence of the multiple series for different integral forms have bounds. These could be considered some sophisticated parallel to attractor convergence in fractals. As they have a possible intersection as well as a pseudo digital behaviour (time analytic of halting problem applied to divergence) they can be used to represent some digital manifold, while maintaining series differentiability. This implies c(y) and f(c(y)) more importantly be fed back to the estimation.

The separation of the percept before the estimation in a real sense is the great filter. Some post situation feedback would help. The log scaling is perhaps also quite important. Considering an exponential half life maybe controlled by production of an enzyme to remove the metastable precursor to reduce it, the multiplicative inverse is quite likely (Newton-Raphson approximant) and integration make for a log scaling possibility. Some feed forward of x provides entropy and some exponentiation or other series decompositions might be useful.

Templigeadicalogical Algebra

Well, where shall I start? All could see the sums were good, of those taught sums a protected right of the conversationally on. A heavy reign introduction to subtraction and hence divisional and rights of subdivision consensus elective protection from decimation.

Command hierarchies of the on for example a battle unit knowing it is one that is on, set to fight for subfeariors of command wants subjunct summative transmissive networks of optimization feedback of induction of sums?

The out wave of a famine “un” the handy past the time of fight show lean on the order of meditative? As the collective induced cook as woman work to collect the connection of hand fights to womb growth multipliers.

This might be fun at the food policy unit 😀

Fight coming up from blame of “monk” to obivatiate doubt on family protection, and thrust occasional mindsets to perimeters of risk reduction from onslaught foresightful of the time to tummy from mummy. As the di summand moved predecimand, the focus on god cycle before analytic deconstruction by sumandment of temple duty, became moot as the knowledge collapse in chaotic cycles not brought into feedback bode stability as the control hierarchy became argumentative dominant.

Bun fight!!!! Let then ate cake?? I’m ‘avin a go at integrands, might PID control feedback stuff if the boding camlmand hireachies summand with better cross information flow? Sumands like ? The first L of simon. Technically though it is a second order tension of l fighty fiesty hg an onset of bode instability so yicked up set delguage from a fWell, via a dissociable epigenetic panic.

Prove me hyper politely wrong on the abuse that extends from the fear of critique. I’m on, some of the nicest people I know are women. There on, but maybe not on on in the wordfield. The uncertainty potential of action in fold downs of understanding? I can say being a man I understand testosterone. The idiomatic fork as well extends from this therefore the competition between communicative and full active fully automatic via lack of information has its inductive effect.

Anti June could swear on summand a sailor! Error analysis in.  Idiomatic jokes are always a shitter. Control yourshelbves bint dat ladies. First orders, second orders power orders, summands and seais so ship? Lndend? Trade …

Accents for the poor? Accination programs? The enlightenment of the orbifold tonces. The dictum freeze from Oxford, an experiment in analytical management by saturnalian net. A distribute of multi-lingo automation?  Distribute, estimate, summand, perform error control minimization. Unlimnate uncertainty of position.

Parse Buffer Overflows? Dark Priorities.

Sounds like such fun. An irremovable or a point update fix on the press? https://github.com/jackokring/majar/blob/master/src/uk/co/kring/kodek/Generator.java sounds like fun too. Choices, choices? Amplified radial uncertainty of Δr.GMm.Δt≤ℏ.r2/2 was kind of the order of last night. Is it dark matter? Is tangential uncertainty in the same respect part of dark energy? The radial uncertainty in a sure instant of time, and the potential gravitational energy? A net inward force congruent with dark energy?

And a tangential version of the squared hypotenuse of radius and tangential uncertainty of radius resultant? That leads to a reduction of gravity at a large radius and is more like dark energy. More evidence for a spectrum of uncertainty amount hence the “less than equals” being simplistic on an actuality?

Oh, no I’ll have to investigate the last GET/POST before errors … how boring (last time an Indian) … guess who?

The Small Big G and Why Gravity?

As G the gravitational constant is small compared to other force constants this would make delta r be bigger in gravity for the same amplified ħ uncertainty. With the time accuracy of light arrival in the visible range, the radial uncertainty at a high radial distance integrates over the non-linearity of the 1/r^2 force, for a net inward. Tangentially, the integral would net a reduction in gravity.


So a partial reason for dark matter and dark energy to be explained by quantum gravity. It’s a simple formula and Δv/Δt as a substitute for Δp=mΔv using F=ma=GMm/r2 in ΔxΔp≤ℏ/2 so the answer is approximate an r±Δr might be more appropriate for exacting calculations, and r2+Δr2 as a tangential hypotenuse.

As https://en.wikipedia.org/wiki/Coulomb%27s_law is 20 orders of magnitude higher the dark coulomb force will be 10 orders of radius larger for the same effect.

As the Mass by the Cube, and the Uncertainty by the Square.

As the distance increases to the centre of a gravitational lens, the uncertainty of the mass radially becomes significant so effectively reducing the minimal acceleration due to gravity, and growing the volume bulk integral of mass in uncertainty. The force delta would be inverse cubic, countered by the cubic growth in integration volume. The force would therefore in isotropy become a fixed quantity effect.

This is not even considering the potential existence of a heavy graviton, or the concept of conservation of a mass information velocity that would have a dark energy effect. It still seems “conservation of acceleration” is not even a taught effect considering there are many wine glasses that would have loved to know about it.

As for the rapid running constant increase toward the unification energy and what inner sun horizons would do to a G magnification? Likely not that relevant? Only the EM force seems to increase in coupling as the energy of the system dilates in time. This would imply the other three standard forces decrease, so necessitating an increase in radial uncertainty on average. The strong force has a with distance effect below the confinement distance, and so as the radius reduces, a Δr.k.Δt≤ℏ/2r rule is likely which would lead to the most likely reciprocal isomorphism of dark matter and dark energy.

Due to quark mass differences, and k, therefore, being one of 15 = 6*(6-1)/2 constants depending on the quark pair a triad product pentad structuring of force to acceleration might occur, with further splitting by boson interactions with quarks. Maybe this is a long shot to infer the finality on the low energy quark set of 6. Likely a totient in there for an 8. That’s all in the phi line and golden, silver and forcing theorems. I wonder if forcing theorems have unforcing and further forcing propergatives?


You could be right.  So? It’s not as though it affected any of the local accelerators I don’t have. If it’s all about the bit not understood, then as a product constraint, it is where the action is at. As the maths might work, I am speculating the further equations will be in a less than form and so need fewer corrections? Premature optimization is the root? Any tiny effect would be on that side of equality perhaps. Maybe it was just a tilt on the suggestion of an inverse isomorphism. I couldn’t say, but that’s how it exited my mind.

Project “majar”

majar is a Java package. It is in development. The current focus is on a shell language. Also, abstraction interfaces for things will arrive, and then implementations.

  • KeyBase – a database based on the idea of a “BulkStream” (a Base) supporting a 5th normal form of relational database where internally everything inherited from Key can be stored. Fields are not stored in records, but each field value becomes a record so an object storing kind of database. Imagine a field query, and then obtain records for all tables and databases.
  • Kodek – the KODEK of K Ring naming fame. I’m sure I’ll get around to a KeyBase Store specialising in a Kodek supporting Class.
  • majar (intentionally lower case) – a scripting language with a bash launch script so that majar becomes the language from “ma.jar”, get it? 
  • Abstracting the script language to run web applications.
  • A Java Servlet extension class for easing some of the pain.
  • A port 287 IP proxy for localhost compression to the publication of “My Public Computer” interface. Also an application server for the browser client.

K Ring CODEC Existential Proof

When p=2q. L(0) is not equal L(1).

Find n such that (L(0)/L(1))^(2n+1) defines the number of bias elements for a certain bias exceeding 2:1. This is not the minimal number of bias elements but is a faster computation of a sufficient existential cardinal order. In fact, it’s erroneous. A more useful equation is


Showing an asymmetry on pq for even counts of containment between adding entropic pseudo-randomness. So if the direction is PQ biased detection and subsample control via horizontals and verticals position splitting? The bit quantity of clockwise parity XOR reflection count parity (CWRP) has an interesting binary sequence. Flipping the clockwise parity and the 12/6 o’clock location inverts the state for modulation.

So asymmetric baryogenesis, that process of some bias in antimatter and matter with an apparently identical mirror symmetry with each other. There must be an existential mechanism and in this mechanism a way of digitizing the process and finding the equivalents to matter and antimatter. Some way of utilizing a probabilistic asymmetry along with a time application to the statistic so that apparent opposites can be made to present a difference on some time presence count.

Proof of Topological Work

A cryptocoin mining strategy designed to reduce power consumption. The work is divided into tiny bits of work with bits of stall caused by data access congestion. The extensive nature of solutions and the variance of solution time reduce conflict as opposed to a single hash function solve. As joining a fork increases splitting of share focuses the tree spread into a chain this has to be considered. As the pull request ordering tokens can expire until a pull request is logged with a solution, this means pull request tokens have to be requested at intervals and also after expiry while any solution would need a valid pull request token to be included in the pull request such that the first solution on a time interval can invalidate later pull requests solving the same interval.

The pull request token contains an algorithmic random and the head random based on the solution of a previous time interval which must be used to perform the work burst. It, therefore, becomes stupid to issue pull request tokens for a future time interval as the head of the master branch has not been fixed and so the pull request token would not by a large order be checksum valid.

The master head address becomes the congestion point. The address is therefore published via a torrent-like mechanism with a clone performed by all slaves who wish to become the elected master. The slaves also have a duty to check the master for errors. This then involves pull-request submissions to the block-tree (as git is) on various forks from the slave pool.

This meta-algorithm therefore can limit work done per IP address by making the submission IP be part of the work specification. Some may like to call it proof of bureaucracy.

The Cryptoclock

As running a split network on a faster clock seems the most effective hack, the master must set the clock by signed publication. On a clock split the closest modulo hashed time plus block slave salt wins. The slave throne line is on the closest modulo hashed values for salt with signed publication. This ensures a corrupt master must keep all slave salts (or references) in the published blocks. A network join must demote the split via a clock moderation factor. This ensures that culling a small subnet to run at a higher rate to disadvantage the small subnet is punished by the majority of neutrals on the throne line in the master elective on the net reunion, by the punitive clock rate deviation from the majority. As you could split and run lower in an attempt to punify!

Estimated 50 pounds sterling 2021-3-30 in bitcoin for the company work done 😀

The Rebase Compaction Bounty (Bonus)

Designed to be a complex task a bounty is set to compress the blockchain structure to a rebased smaller data equivalent. This is done by effectively removing many earlier blocks and placing a special block of archival index terminals for non-transferred holdings in the ancient block history. This is bound to happen infrequently to never and set at a lotto rate depending on the mined percents. This would eventually cause a work spurt based on the expected gain. The ruling controlling the energy expenditure versus the archival cost could be integrated with the wallet stagnation (into the void) by setting a wallet timeout of the order of many years.

A form of lotto inheritance for the collective data duplication cost of historic irrelevance. A super computational only to be taken on by the supercomputer of the age. A method therefore of computational research as it were, and not something for everybody to do, but easy for everybody to check as they compact.