Super Blog

taking a closer look at the tactical and strategi architecture of concurrent virtual achine networks

Abstract: I 24 single core virtual machines with approx. 250MB of GPU /network on the same virtual network segment.

Clever utility of the underlying physical hardware

Exposes the time shared landlord multiple occupancy substrate.

e.g . 24 small specialised ai models working concurrently, over time can ‘evolve’ synthetic datum.

other examples :

24 node neuronal control surface.

AI per strand of control structure.

The back pain cluster of AIs controller uses big AI modelling to work through the 24x AI datum streams.

Working together this would be able to command control upto 24 degrees of freedom robotics or machine control strutures.

Offers potential 24/7/365 operation of machines.

Now multiply the lines of communique over a warehouse or industrial operations and you can see the intelligence value.

Using virtual machine networks is a mathematical more valuable utilisation of computational resources across time domains.

Degrees of axis of freedoms in feedback based robotics systems.

Used intelligently could drastically improve manufacturing logistics part and parcel integrations could massively accelerate the day to operation reality of logistics operations.

The potential of this hierarchical arrangement can spin up topology and work distributions networks more effectively.

The theory of cybernetics control structures still hold weigh and impresses the academic debate around the topic. But its important messages have to be taken into consideration today even with the forthcoming revolution in technology which is on the other side of this information revolution.

I believe the true transformation point of all of this very very disruptive technological revolution i would say is still to be felt and perhaps the other side of the disruptive wave as it were wouldn’t be perhaps fully realised until 20 or 30 years into the future before the world fully resolved and adapted to the consequences.

This is fascinating stuff and still gives us a good mathematical grounding in todays world of emergent AI robotics and Intelligence intersectional space.

The topoloical arrangemet of this present gen structue, virtual machine, pci bridge gpu, and some space and time for computations and comparisons.

Much like a recording studio or synthesiser the whole mess of cables and connections with neuronal computational structures can be geometrically transposed on the ‘bodies’ of robots or manufacturing machines.

A sort of virtual machine central nervous system.

Photonics has also come to the fore of the discussion in some circles making a key point for the forthcoming pci7 system upgrade.

With nano metres comes super density super lightweight computational units.

However that being said signals and sensing, perhaps even in non-human domains of space and time. Giving us insides like never before. Expanding the human lenses with which we view knowledge and science with.

After initial stock issuance a company valued at 10,000 GBP in the initial registration has to move to 24/7/365 coverage.

An always on AI is probably the way to go.

The way things are its now definitely 24 7 365 is the new business model to compete in all markets all over the globe.

New things evolve, recently Honk Kong surpassed Switzerland as a wealth and asset manager class now the biggest in the world in relative dollar strength.

The structure is changing into the always on blockchain smart contract financial instrument effectively, this means as a direct competitive measure, to be a participant in these markets demands always on globally connected always on tech infrastructure.

Economically, and geo strategically, things have gone that way, the game has sort of evolved into that, over the term.

However so has organised crime we are all aware of this big hack of usd from Binance exchange, the biggest most profitable new financial exchange.

That being said its par for the course, we expect it, not from the biggest player in the crypto space but that’s how it goes.

Right now a geo strategic asset is a server farm. Always on, always accessible, always making money selling digital subscriber services.

We pretty much have to offer safety security, accessibility, traditional business computing services, of course Artificial Intelligence, and traditional websites and webstores with e-commerce, perhaps as a reseller to bigger providers.

We have to hedge our bets over the term with ownership of traditional assets, and new crypto assets and perhaps behave as a crypto treasury style company, perhaps at division level.

On top of the power supply and communications fundamentals.

An always on computational service layer, will be able to act as a foundation to higher operating assets layer.

Finding profitability in hosting crypto, blockchain and smart contracts is technically demanding and quite risky but eventually at an achievable scale this becomes profitable.

Slots on the blockchain provisioning layer ought to be considered a key business asset. Remember if your infrastructure books the blockchain in that incident you receive the rewards in terns if crypto pay-backs.

There are inherent possibilities that can change mathematically but due to variance in the network, although highly competitive, can be lucrative.

At some fundamental level, power supply, rent and rates and various operating costs have to be covered, here’s were we use the money from the initial stock sale covers this, but with continuing delivery of a variety of infrastructure on top service layers should allow us to operate and profit in this essential market segment.

There are a few models out there and several various business case studies that argue strongly the network effect of technology.

There is a technical necessity to updraft technologically into the server farm game as when needs requirements for more technical projects, the SAAS/SAP Validation and technical design services we have occasionally been required to fulfil, we will have capable infrastructure for these big computational jobs.

Obviously, in terms of capital and contract licencing and manufacturing fees, we can take a prototype and design all the way through to validation.

However as we are aware the actual capex costs to compete at that level will be way out of most peoples reach’s we never the less still need to maintain excellence in that technical competency should the nee arise or we indeed business in these layers.

Pain job today, once again for the hundredth time transferring everything , plugins etc to new studio computer. Its a good computer right enough but already disk almost full. This is one of the most painful expenditure of time doing the same thing for the umpteenth time. Anyhow, back on with the job.

So we have had a problem for a long time so weve decided to back propogate the OS version.

There is an compatability issue with the latest versions of linux nvida kernel drivers arising from the fact they have been built using different version of the cpp compiler.

This is causing an ABI mismatch fault. A security requirement needs both Linux kernel and NVidia kernel device drivers to have been built using the same generation of compiler.

The solution is to recompile everything, both Linux kernel and nvidia device drivers using the same toolchain.

However weve decided were going to back propagate the ubuntu to os level 18. A crazy decision but the reason is that the nvidia toolsets work fine at that generational fault-line.

We have about 75% of the available toolsets, but its just some key device drivers at kernel level we need to be truly independent.

So basically, weve been using ollama, nemtron 3 nano with open UI website front end, but mostly from the command line and the results have been mind blowing, really amazing, cuda works, the nsight toolsets work and we can compile and build cuda its just that we need that few extra capabilities so we can run everything compatibility of toolsets is the issue.

Bringing everything back a few versions, the ABI ( the application Binary Interface) obviously hasn’t caught up with all the different versions and releases and there is also the secure device driver UEFI/BIOS thing going on here aswell which is another source of compatibility and security issues.

SO basically its full rebuild from the ground up using slightly ex gen behind the curve versions.

SO back to the begining again.

ollama, nemotron 3 nano and webUI with python-env.

Current job downloading installing configuring and local hosting.

Will update with progression report as and when.

(note this is a bit of a throw away project as its a bit of a voyage of discovery I got a funny feeling this might not work out too good as ive only got 8gb NVIDIA GPU and the model is a 24gb model.)

Time taken – around 3 – 4 hours.

Latest NVIDIA propaganda at Berlin GTC is most impressive.

Its a bit slow for European autumn / winter. But the coherence of AI altogether is much improved. Of course code only, or software only actually misses out on the other half of the NVIDIA world. Robotics.

Yes Robotics, which in this case means anything not software, basically.

We’ve found the cheapest NVIDIA compatible robot at around thousand gbp , so basically a grand, for a compatible robotics platform that you can then sort of remote into the NVIDIA software, field of agents, AI and library ecosystem.

So there is small buy in, relative to the standards of the recommended technological skill set in the emergent field of cybernetics next chapter.

Of course, what are robotics, what is AI ? Well anything really that takes advantage of this client server architecture, but far more than just lines of circuits or lines of codes because it generates what is called intelligence. And that is mind blowing when you think about it.

It looks like NVIDIA have really improved the interconnect, the fabric, the service back pane, which is responsible for the gains up front.

Artificial Intelligence Assistance, is only say the consumer front end of the battle for AI.

The infrastructure layer is something that interconnects front end and back end stuff.

But its more intelligent than that. Its a somethings extra, perhaps the ineffable is being delivered ?

There’s a whole ways to go, moving forward, but presently it seems the states of affairs around NVIDIA, robotics, Cuda/Python, and now the services infrastructure is getting online, the ever increasing sort of pre-cambrian explosion of AI models and agency ( which is even worse than javascript and blockchain coins and tokens ).

Much more than even smart watches and all of that gym bro wearable tech.

Yes, as I have stated previously, the increase are a sort of self improvement thing for the industry as the cyclical development proceeds accordingly.

In terms of business and industry I think electronics and mechanical engineering as well as battery and power generation are going to be big going forward.

There’s lots of politics and philosophies surrounding the AI question but we tend to stay away from that as cybernetics is a technical subject, so we would rather keep to the technological fundamentals without necessarily expanding the gambit, the umbrella too far.

NVIDIA propaganda impressive, backed by real results. And the universal applicability of the product sort of unifies all fields and as a such its a very compatible business, especially amongst business to business sales. An extremely high degree of affinity, perhaps mostly for the science, technology and mathematics (STEM) first and foremost but also certainly any production downstream off that. Remarkable affinity.

Building a simple application with C++ /openGL and build environment for windows.

Project time : 2 hours

---

Goal : Build cycle to evolve an openGL based application for windows platform.

The idea is to construct the compile build deploy (release) cycle using openGL as a 3d graphics library.

The idea is to start out small, simple, and once we have got a build cycle going expand out and build up our 3d graphics based application on windows.

Correctly configuring the build cycle is key. Once this is done we will build up out with breadth scope and depth, so that over time we can build up our 3d graphics application and take it into a variety of directions and eventualities.

Next step : Network server awareness with update capacity

This next step is important for development. We must create an update mechanism. This is preferable at large scale. Although not strictly necessary yet in the early days of a 3d graphics project on windows. Once we avail ourselves of an automatic updates capability we no longer have to keep building from scratch and downloading the full project. Using delta updates (which we discussed in a previous post) this procedure will allow us to keep the size and time taken of updates to a minimum.

Having network server with networked awareness will allow all sorts of value added product features, such as communications and multiplayer/team play scenarios.

Project Negatives

Although therefore we would be probably be better just using unreal engine, or Unity, we are doing these quick projects in C++ with OpenGL do demonstrate principles, yes there is a lot of re inventing the wheel, and going over old ground, but that is the point of training and practice.

Project Advantages

There is a minimum compact form advantage to doing this project, the absence of technicalities makes the project small in terms of data size, and lessens complexity and interdependence, which can take up a lot of time ensuring all of the libraries, frameworks, includes, etc etc are up to date and protected against introducing outside errors and bugs and the difficulties of keeping everything up to date.

---

Alright here we go again! Looking good so far so will be keeping up with the shows workshops demos lectures etc.

News is travelling fast across the internet regards PHP security bug. It already seems to be addresses. BY using

sudo apt update && upgrade –y from ubunutu command line terminal should install the latest up to date version.

requires reboot.

dont know too much about this one but seems to be fixed.

SO NVIDIA TIME AGAIN AND COD TIME AGAIN.

Some good sales going on in video games i guest they’ve got to get some sales while everyone on the planet gets the latest COD.

Looking forward to the NIVIDIA Work SHops.

However we need to keep spending time on freeCAD and KiCAD as were losing pace by keeping up with all these Shows and Conferences, although essential to keep up with the information loop we also lose the time to manufacturing design productivity so we need to re addresse that in-balance.

SO the keeping up is not training and excersise, its not practice, the all essential experience generator.

For example a humble relatively simple motherboard would take a skilled human operator, perhaps highly skilled technician if truth be told, about 60 hours against the computer working on CADCAD software.

That’s like a month of employment perhaps, several thousand pounds worth of work.

However , we also have to add the actually charges and costs including testing and delivery, of the finished product. SO the manufacturing and logistics costs to actually have a product in the market space is quite a lot.

This is quite the evolutionary development, however there is a natural slowdown occurring with most manufacturers across the spectrum have markedly increased prices due to the current price wars that are occurring sociologically.

SO it looks like well have to tread water so to speak by using just present gen or ex gen equipment. My guess this will slow down sales all across the economy.

Its going to be an interesting time going forwards.

This is a fascinating emergent field, were the combination of technical problems in both traditional software and block-ware block chain enhanced software, whereby obstacles are faced as regards updates, wither live memory or installed software.

Just think how problematic it is for hundreds of thousands of users, or even at the level of operating systems and their corresponding updates.

As you can see its a major engineering challenge facing computer science today.

Enter the emergence of new specialism the field of updating. Although it doesn’t sound very exotic, especially compared to cybernetics or even the various AI or Super Intelligence fields.

Its an important phase transition to truelly handling global data volume.

Whilst mainly used in computer video games the research and development that has gone into to successfully managing mega online data volumes at multi-region scale.

More to follow.

you will need to tape some radio shows or copy and paste some podcast content into a folder for the basic stems to train the ai model on.

In fact this is a dual model ai.

We will also build up a small library of text, so we can run the text to speech AND the speech to text bi-directionally .

Finally, we will build the voice commands for a speech to text, to command processor.

Although this will only be able to achieve a small handful of voice commands but the skills learned here will contribute to far more complex AI projects further down the line.

However that’s still not so bad for our first AI model lets proceed to the projet.

If you want to record the podcast or lecture yourself you can use an x-box controller, plug in a video game headset and plug this into the computer. You will need to use OBS streamer and give windows permissions to use the microphone somewhere within windows settings.

One of the more common models from Nvidia ai will allow you to work speech to text against the Nvidia AI model but you will need to pay about a dollar an hour to use it from the Nvidia cloud services.

Just been finding out about this one, absolutely fantastic.

If your starting out mixing and mastering then this is definitely the one. Fantastic some old school features such as EQ gates, something Ive been looking for for a awhile.

Anyway, Im convinced will probably buy it in a month or two.

Also convinced to do final mastering using ReaperDAW aswell.

Presently SSL 360 with the amazing and cheap plug ins that come with the full fat subscription, hosted in ReaperDAW and actually remoting over the network into and out of and then back into Ableton(which is on another computer).

Its funny because there is a bit of AI convergence as were now using one pc, the new one (ultra7rtx5060) for the AI and Nvidia stuff which is remote controlling the old studio computer (i5rtx3060). (We have it dual boot with windows and Ubuntu)

So there’s a sort of phase transition going on here of sorts, similarities and differences aside, between the two fields. With an emergent fabric of some level of complexity, of course we only have finite resources, to share between all of the different work flows. AI, 3d Graphics, Music Production, Video games, Editing and Post etc etc).

There have been Intel NUC pcie cards which have been a micro PC on a additional upgrade card so we might scavenge one or two of them to add to the load out.

Before I forget we had a heavy few days just recently.

Ubuntu Desktop with visual studio code and nvidia cuda compiler.

We also attempted to compile build and install unreal engine from the github for linux.

Failed on unreal engine, well due to ubuntu not having the right GPU drivers, so built and ready to go waiting on an GPU device driver update (so totally out of our hands).

It was a heavy 18 hours session.

Some use of ReaperDAW aswell out of interest.

This prompted us today to get a windows 11 OS loadout so back to that.

However programing and using ai agents is way easier on ubuntu

Also did some work using Blender on ubuntu aswell.

The thing is you need both windows and ubuntu for the different areas of specialisation.

However window is naff these days were just forced to use it because all of the applications, plug ins, DAW etc etc work easily with it.

That being said we build a basic cuda application that runs on the nvidia GPU and we correctly configured the InSight develper tool from nvidia aswell which can attach into the gpu accelerating applications and track trace and debug cuda applications.

Interestingly, we were using the intel onboard GPU for the monitor, and we were using the nvidia gpu in mult tenancy mode, so whereby we can run maybe ten or so ai agents as virtual machines with each being given a slice of the gpu graphics memory, so for example we could divide the space by 8, have 8 virtual machines with 2 vcores, 2gb ram and 1gb of nvidai gpu capability aswell.

Using this swarm of agents, or field of agents, methodology we can the use these agents to create work as the manipulate unreal engine, ableton, blender,visual studio code, etc so you could generate a large corpus of work in the digital realm with relative ease.

Heavy going setting it all up.

The new studio PC is pretty decent we’ve only had it up and running just a few days now and its bothersome transferring everything across so that’s still going to take more time over the next couple of days over the coming week or so if we can get the timetable slots available.

Wow we’ve covered a lot.

intstrument and drum racks

automation

soundcreation

house keeping

templates

musical ideation

song structure

generative and random methodologies

cranking out bangers

loops, beats, breaks, fills.

chord progressions

lead harmony melody counterpoint

style and structure

progressions

saving working ideas for future ease of use

ableton deep dive / hidden or often overlooked areas of production

keyboard shortcuts

sidechaining and dynamic eqaulisation

recording

sample reworking

macros and control

user interface ideas and workflow shortcuts

Max For Live

Using some cybernetic control flow methodolgies

the ableton note application on iphone/apple tablet

using sample libraries

quick ways through search and narrowing down options quickly

Working with different scales

Working within different musical genres

BPM and music style

additional materials:

Using ReaperDAW for a standalone mixing and mastering studio computer.

Attaching ReaperDAW with SSL 360 over the ethernet onto your Ableton computer for a final Mixing and Mastering.

Using ethernet wifi and midi CC over the network so mixing and mastering is done on another computer within the same network allowing a dual system approach.

Creating and bundling sample packs for AKAI mpc.

Creating a finished product for publishing to online music platforms using DANTE. Focusrite RedNET x2p combined with a Tascam 250NR Dante enabled solid state recorder (out of the box standalone final recorder).

Live Performance and Studio Tracking Sessions will be covered in a future session.

Final notes : Heck of a lot of ground to cover and a pretty decent amount of work has to go into music production. Expect many hours of production time to go into actually making something worthwhile that has a high production value. However, that being said we should by now have a lot of templates etc already set up so we can quickly get to it without always having to start from scratch each and every time. Using these handy production housekeeping and saving and making available the various stuff we build within Ableton we can, once up and running correctly, save ourselves a lot of future re building from scratch everytime.

The separating out of mixing and mastering the SSL/bundle remotely controlling the DAW through ReaperDAW and having each and every musical sound already brought up to a quality sonic level frees up our immediate concentration on cranking out bangers.

All of the jobs have to be done so the more we get achieved up front the quicker we can work through the motions to deliver the final product within reasonable timeframe and schedule.

Some interesting add ons going on with blender MCP.

Now time to download and compile the source code.

Also using Reaper DAW so we got a pretty handy sort of start out load out going in terms of studio and creative work.

It will take a bit to load out the agents or MCPs to validate through blender unreal engine ReaperDAW and compilation cycles between C++ and CUDA but I think shortly we will have everything off the ground and starting to develop out.

Blender isnt rendering like it should at the minute.

So finally moved over to the new desktop.

This one is an Ultra7 270k, 16gb ddr5 ram 2.5tb disks and an nvidia rtx5060.

Took a bit to get the nvidia GPU sorted on the present gen of ubunutu desktop. The installation app had been discontinued in ubunutu so had to adapt the installation scripts to include the new version of ‘auto-install’.

The nvidia stuff is all up and running including the NVIDIA visual studio tools within visual studio code.

We’ve got it setup as a virtual GPU server so the graphics are handled by the onboard Intel graphics whilst the whole graphics card is available to ai models and cuda enabled apps and python libraries.

Its fast, really snappy and brings up to date for intel OneAPI, and Nvidia Cuda Compiler.

Usual toolset, Blender, KiCAD, FreeCAD etc and well impressed with the system.

Still to do : install a variety of AI models and compile the latest python libraries.

As regards development Ubuntu and Nvidia is the industry standard, The underlying substrate handles the workflows topologically. Where as windows struggles with that even with windows subsystem for linux.

Going forward- great software AI development environment so will revisit the stuff on developer.nvidia and further explore the ecosystem, Issac sim for robots, omniverse, etc.

Will be installing and operating field of agents to see if we can get a bit of a boost in terms of productivity applications outputs.

So these days we are mixing and mastering entirely in the box using plugins.

The plugins we are using are the SSL, Slate Digital, and Harrison, plugin bundles, with occasionally dipping back into native instruments or Izotope plugins.

The value for money and the technical capabilities, mimicking pro gear that would of otherwise costs many tens of thousands in a purely analogue, real world mixing and mastering technical capabilty.

Something that in the ‘real world’ extremely costly.

SO I guess this is as good as it gets really, the dsp and math algorythms are probably not going to be better by much.

The ‘Being Digital’ era is now maybe 30 years old ? Perhaps? Definitely mainstream. For relatively small effort and costs, and massive conveniences of doing all this ‘in the box’, and the general maximum availability and wide spread through internet cultural bases, such as Ableton or max for live, and the great software industry behind hardware and software products.

The music industry is at a cross roads, after streaming and all of that. Post AI, or at least Co-Existant with the AI thing, there is much difference in opinion into how things should evolve from here.

With the the internet pre-cambrian explosion, and the shear accessibility and market access that all this affords, combined with the earning potential if your product or software subscription reaches critical masses of subscriber numbers.

Culture has reached an accept or reject cultural moment with the coincidence and collision of all these lines of communique and relative and respective lines of development.

So I think that the capitalisation of music tech business and the results of certain busts and booms in terms of corporate groups that hold the intellectual property and global sales markets of the music industry and business, there is also a search for character, personality, originality, perhaps the emergence of a new punk to counteract the same all old corporate cultural tropes in the industry.

Newer products seem to be characterised by old school look and feels harkening back to the pre-digital era.

Too late to say we are post digital / post ai ?

To early to say we abandon the whole digital ai thing?

Do we go back to oil and canvas to make our album art ?

Certainly the very aesthetic of digital futurism, is incredible and perhaps hardly achievable by mere human comic artists,

This is really is a quite a historic moment into the future of culture, was Judge Dredd and 2000AD right all along?

Certainly technical capability along creative studios and the capability of the output, in terms of cultural gigabytes, perhaps, we increasingly ‘see’ inside of our selves, delving further and further into creative imaginery space, pulling out of the deep ever more crazy conceptualisations and modalites, perceiving or creating further along the lines of our imagination, incredibly performant.

Perhaps humankind always saw this, but lacked the cybernetic means to express the thought, or see the sign, and certainly melded with technical capability, which will be the biggest toolkit at our disposal, in terms of technical performant creative complexity within industrial architectural goal oriented deign , discovering new spaces for exploration again in kind returning a new level, a new place to evolve creatively from.

Perhaps old ways of looking at things might not serve us too well in such days ?

To be or not to be?

O before i forget some of these latest plugins have an easy ‘AI’ button that can set all the settings for you which saves time just getting something down ASAP.

I still know you need a mix engineer in the chair, in terms of cybernetics, a full fat analogue mixing desk, or the equivalent control service, works infinitely better faster, with immediacy and simultaneity (as in simultaenously)is still, as of time of writing, faster than computers to do the job.

Granted a ‘field of agents’ or a ‘swarm of agents’ could hit out at every circumstance massively in parallel, and i guess with a few basic rules could do everything quickly but without the human in the loop i don’t think it would get to the master achievement of a complete well done mixing and mastering. You know the human in the loop I think is still more capable, but a ‘field of agents’ could certainly bring everything up to speed as regards the leg work of setting up a recording session on computer or mixing desk and equipment.

Avenues still to be explored – perhaps using playstation VR or the Apple headset and using that to express virtually the architectural space of a mixing and mastering recording studio with a 3d version of the acoustic space.

Certainly 3d and spacial workflows for video games with the appropriate gaming style headphone would mean you could work entirely in VR/XR scenario, so that’s an interesting permutation to counter the complete lack of space, even when working with multiple monitors, and would loads of annoying zooming the keyboard and mouse around to a particular plugin , waiting for the screen to be revealed and then of course tweaking the setting. All of these tens of thousands of micro, sub-routings, or micro interactions, touch points upon the mix, each cost digital computational powers to beam in and beam out of each of the several parts.

Just a technical note, I probably could just get the ssl daw control surface, which all in is around 2 thousand pounds, bet even then would only put us about half way between keyboard and mouse, and the actual real world nuts and bolts mixing console and out of the box equipment, in terms of operator speed, but it would be putting us ten times faster than operating the production service using just a keyboard and mouse.

The final technical alternative would be mixing using apple pods with an apple monitor with Wavelab 13. This would be about 2 thousand pounds including the cost of the Wavelab software, which would be the cheapest possible atmos capability.

Any way that all for now.

Usual maintenance updates.

A quick review of AI used in Ableton, Max for live, and now these new ableton extensions.

Personally I don’t think there is anything wrong with it, it has really helped a lot of people get over the steep and long term learning curve which is only a good thing.

I think it is about to get blown wide open, but still musical aesthetic considerations will always relate to a human operator or human performance or human intuition and creativity.

Its sometime can be a heck of a lot of work to get your head around the tactility of using computers for music and the learning cycles for manipulating music software, plugins, modalities and ways of doing things or arriving at results.

There is still a good few ‘errors’ or ‘faults’ that come through ai generated music stems for example, but never the less helping get things going.

The end point of music streams and payments from online content hosting platforms isn’t the primary creative goal, and is a purely business inspired decision.

There are a lot of people out there that aren’t in it for the money. Never have, never will be although money would be nice.

SO for non commercial inspired endpoints, and also music for people to profit from online platforms, which admittedly are completely overwhelmed with AI “conjurations”, I still don’t have a problem really with ai generation.

In the end I think well evolve and adapt and move forward and come to include this additionally creative power as an enhancement or assistance in what is ultimately, from an artistic point of view, a quest for aesthetics and relisations through the sonic academy.

Pressure increases for business offerings via the techno verse.

Increasing G, AI swarm of agents and new Landlord AI pricing, perhaps being wrapped up conveniently on smart contract platforms.

There is perhaps the scope for the emergence of smart AI crypto currencies build from fresh say circa 2028. Certainly by that time it is probabilistically realistic to anticipate the emergence of a smart AI crypto currency and technical service coverage platform.

This would make it most simple to design and deploy from the customers perspective.

The infrastructure side will continue to develop. A sort of technical emergence of digital perhaps cyberspace landlordism, a sort of reverse SecondLife.

It looks like a new branch of cybernetics is emerging into the scientific research arena.

mathematical field control is emerging field effects basically bio-domes geodesic biosphere but achieved not by organic natural means by flora and fauna but by technical field effects using post silicon post quantum methodologies.

It is an emergent conceptualisation of a post quantum post ai scientific technical capability of reflexive intelligent feilds.

As an aside , this formative science-fiction discussion, is emerging in procedural content generation, the concept is the underlying ai enhanced computer program will formulate emergent worlds into the conscious perception, as a fore runner to particle accelerators, or quantum smashers , that will be able to , by the ai and the quantum circuitry computer controller, field effect the laws of physics and the quantum level that will effectively construct, worlds.

In essence we are LARPing this potential development by developing and using video games to explorer the concept in silicon, of emerging virtual worlds.

The coupling together of the three developments will in effect pave the way for the geo-spatial field effects in emergent worlds.

This would represent an almost holographical capability, one which we have, in terms of world building, within computer programmes, such as video games consoles and such like.

Its quite comic book were a super intelligent mega brain would be connected to the technical apparatus and could sentiently create worlds through a process of emergence by controlling the substrate beneath the laws of physics. SO effectively, perhaps a level of control, ultimately intimately connected to the opposite side of industrial processes, materials science and crystallographic processes such as microchip manufacturing, sort, of behind the curtain or the other side of the veil of the laws of physics themselves.

Its an interesting science fiction concept and we are , in a way practicing this through video games etc, on the other side of the mirror of the layers of the laws of physics.

This of course would be a super super technology and would be an impossible technology viewing things from present perspective.

We occupy geo spatial and technical dimensions of process and thus device improvement and manufacturing capability general moves along these two dimensions. improving with the progression of time and the recursive cycling through the generations.

The laws of physics and mathematics are both at the same time the limitation and the enabler, a strange duality were human ingenuity and technical skill constantly smash and bend and restructure materials science and manufacturing processes that bring out our technical capabilities.

Cost per Computer Comparison

Price Potential Break Down

between

24/7/365 coverage

Hourly rate (e.g. Nvidia GPU)

Model Agent Pairs

Pay Per CLick

Smartcontract

There’s also fixed costs to consider in the pricing. There has to be a differential pair provisioning pricing to take into consideration.

These foundational expenses have to be studied and understood aswell.

For example wrapping everything up in a smart contract has to come at a certain price you pay for the server.

Were talking a globally aware trillion point information architecture overlayed on the real world.

With super compute accessed at different logical topologies and hierarchies of information presentation, interrogation , and psychic strategic geo-temporal location.

The perceiving agent becomes most important in this equation.

One could argue that the network itself is the hierarchy intelligently superimposed to cover intelligence observations and conductors.

For example a spider intelligently builds its eight legs because it intelligent enough to know need and perceive the utilities of for example spinning a web. Its an interesting change of direction in the cause of things or the hierarchy of being.

To superimpose intelligence backwards in time although counter intuited the computational process systematic reflexes actually is more intelligent and observable phenomena.

In robotic factories the principle is also applied and the same with self driving robots cars and taxis.

Have been nagging on about this for a while so here we go.

Luckily got a rtx 5060 a good while back and also an oldish RAID NVME card that I can repurpose into the new build.

Still expensive, a mix of parts new and used brings the price down a good bit really.

The RAID NVME was from a Xeon server job a maybe 3 of four years ago and the discs have hardly been used.

The price back then, just going from memory was like 140 gbp around that. The cost to build now is something like 400 quid, seriously, how can the replacement cost be that much more expensive ?

Luckily I built up the component parts up over time using micro financing and even the cpu, this time its an Ultra7 270k has gone up by 20 pound, not much but still, and the gpu is about 45 pound more expensive now than when i got it around 6 months ago.

Long overdue, still rattling the i5 12thgen with rtx 3060 so that’s going into the second hand market to recover a few quid.

I might have mentioned it. But the primary reason is to get an IntelOne API and and Nvidia CUDA compiler compatibility line up.

So this will be able to compile software, tools, pcb validation and python and ai stuff with the present gen cloud compatible compilers and toolchains. So basically small jobs can be built on the box and if there is a need for a big job then the software can be drag and dropped either into intel Tiber cloud or build.nvidia.

Also to note some short courses completed over at build.nvidia regards ai agent workflows

However didnt get to far because of lack of windows subsytem for linux. Its hardly worth bothering on anything other than Ubuntu for nvidia AI.

I anticipate not much difference for unreal engine , blender, ableton , or windows based CAD/CAM (slight uptick in performance)

So for usual desktop workflows like DaVinci resolve or what have not expecting much.

However this AI models and Agents work flows look like you are better of on ubuntu. (i think 20.04 is nvidia present support line)

So there’s definitely a sort of case for a AI partner to assist the compiler or even for smart contract work aswell.

Back doing some audio post production so got on this plu in bundle.

Its a good value for money really the sonic acoustics are really good, clear and bright.

Its been, well say back to the Izotope plugin bundle I have version 9 and have just been using those previously, with the Rx Audio repair tool still one of my favourites.

Will post some new music soon. Have been busy these past years not really got time for it as have been doing web3 and some technical stuff like electronic engineering.

Feels good to be mixing and mastering music again its really chilling.