A.I. - Amalgamated Information

A.I.

One would think that there is little else to be concerned about over the past few months; other than the Russian war onto Ukraine and its wider economic geopolitical impact, Middle East instability and expanding environmental degradation with the certainty that as this develops many millions will add to the burgeoning population migration of underdeveloped counties pressurising into the more affluent secular (westernised) countries bringing with them probable conflicts of placement affordability and ideologies.

There are other unresolved issues wrapped in the divergences of societies which are of festering nature, stymied in the inability to challenge the created causes of divergence via self interest power blocks which can only add to the strength of the popular unease that prevails now. Yet much is being expressed about the rise and application of A.I. probably rightly so for its rapid technological adoption in the industrial areas and its integration into all commercial fields of commerce is causing some fright. That it will have an uncalculated cost on the economic social productivity is a given but although this integrated impact is largely assumed to be unstoppable as it is to be driven by its ‘cost saving efficiency’ across a wide spectrum of employment application; for the moment very little is being done in preparation for the demise of society reliance on millions of ‘backroom / front of house’ employees and their valuable (and disposable) income streams.

None of the developed countries have any sense of the real direction of A.I. and its impact on administrative interpersonal sectors; many of whom have accumulated knowledge, experience, and training and have reactive emotive interactive discretionary choices, in dealing with face to face humans. Tens of thousands of indigenous technically skilled qualitative jobs will be lost. Onto this will be the ‘importation’ of foreign labour (to offset declining indigenous births and aging population) to be supposedly absorbed into skilled or mostly services and unskilled jobs, jobs that may not be in existence for long and some potentially unaffordable in an overall labour market that will become perhaps saturated with the labour casts-offs of rampant A.I.

To counter any negative views of the impacts of A.I., ‘factual’ unsupported commentary, primarily by those that are invested in the development and in the roll out with the wide spread adoption of A.I. and those that do not see themselves as being potential ‘victims’; often make reference to past technologic changes such as the industrial ‘revolution’, mechanisation, electrification, computers, productive robotics and all specialism in those field with adjunct services etc eventually produced greater employment. But one may think that such ‘employment’ gains were incidental, unplanned for and often resisted by capital usury. Such employment was slow to build up and they were a products of their expansionist time, political imperatives, and of contained demographics, built on the application of human ingenuity, accumulated knowledge, acquisition of plentiful resources and a knowledge of operative directive control to know what they did and how. Employment has always been lost much faster that the ability to create them.

A.I. is shaping up to be the antithesis of having human directive control over its actions, (other than pulling the plug) the smarter it becomes and the drive to massively integrate it into all spheres of outputs, the less ability humans have to understand what it does, how it does it and how it reaches ‘decisions’ to hand to human recipients after it has churned though the Amalgamated Information that it has been given access to via the www etc. Thus far it has no ability to decipher merged data being true or false, fact or fiction, guided by the initial algorithm programme to match up to a selective desirable output. It has no imagination, no emotive ability, no creativity, no self awareness, and no ‘will’ for self directive drive. It is a created savant handicapped thus far by numerous limitations reliant on gaining and shaping Accumulated Information of human acquired sources; it is an accumulation sorting ‘machine’. The term Artificial Intelligence, is for the moment a very poor appellation yet it grasps the attention of the computer-arties on the basis of its potential and its ability to crunch varied mass data, at speed and offers, primarily, a usable monetarism output. There are of course many applications it has been applied to that do help in the manipulation of huge data sets that has benefited human knowledge in a variety of fields and the ‘clever’ assembly of mass / diversity dispersed information from existing sources. It is ability programmed to crunch, is ‘intelligently useful’ driven by the algorithms it works with; however it is entirely artificial, in the sense it is the product of a ‘machine’ but intelligent it is not; for now.

However therein rest the undoubted potential and the impending danger, how can it be trusted when it cannot explain (if it is needed to) what it does or how it does it (outputs) with the data it amalgamates, in a way humans will understand? It cannot be good enough to just accept what it churns out when the output can have dramatic impacts on all society, human structures and social needs.

This brings up a greater existential problem, probably not really recognised just yet. For the moment A.I. operates in and on the boundary knowledge of human assembled information sources and applications; it (A.i.) is known (in mechanical technical processing terms) how it works as it does, with it set within the set algorithms structured designed into it to achieves an output. However as is the case now with many application of technology from cars, computers, phones, cameras, television, radio, etc. etc. all used by millions of people around the world; many will know what these things do in usability terms yet have no idea how they work in any depth. It might be argued that for most there is no need to understand the intricate construction of these ‘tools’, what make them tic, somebody else does and can fix them. This is a stance that works because all these things are in the technical mechanical fields and specialism education, it is wide spread and obtainable into employment fields. But, (and there has been examples of this) A.i. has been able to output information that is completely wrong, made up and does not fall within the designed algorithms so in some case it is not known how the A.i. managed to produce an erroneous / biased output initially accepted but later ‘proof checked’ as fabricated. Agreed A.i. has been very useful in scanning, sorting, analysing vast data to generate an output that exceeds the laborious human process in searching similar data and in all probability A.i. will be fine tuned to be hyper productive and accurate up to the point that humans can understand what it does and how but the step beyond that ability can be discerned.

Enter now the much sort-after and greater power of Quantum A.i. computation. It is in its infancy thus far and it is little more than a union of super computational power that may multi-task at different levels, a range of data, all at the same time and form connections with that data that humans might not have seen themselves. A good example of this is the recent unravelling and ‘mapping’ of a protein structure which has been near impossible with normal strenuous interrogative analysis; and A.i. gave a 3 dimensional structure which is now being applied to further understanding of biophysics. Continued development with such computational power will increase enabling links of data for Q A.i to manipulate and in doing such ‘exotic’ connections it might (probably) create its own algorithms to process the data and reach ‘conclusions’ that may be useful but be unable to explain how it came about. This will be a potential void in human understanding; how does it do what is does with the information provided? Nor will humans be any position questions it due to them not having the ability to form a subject question based on unobtainable quantum processes. Perhaps none of this informational voiding matters if in all circumstances the outputs of A.i. quantum or otherwise, are of value and can be capitalised on (for the greater good of course?) which no doubt will cause a massive dislocation in all employment fields with little done to mitigate the personnel outfall.

As it stands now A.i. is a created set of coded algorithms which is applied in all A.i. functions, It has no morals, no emotive content, and no discrimination on the accuracy of information or validation of the accuracy of the source it has been given access to, no outstanding originality, no self veracity assessment and it may contain inbuilt bias based in the algorithm that drive the A.i. reliant on the directive key phrases forming the question put to it, nor has it any ability to be intuitively curious or be rational querying its own output for its veracity. There is a plethora of A.i. uses in play just now and more being lauded for what it does, some “Chat Boxes” such as, Bing Chat, Bard and other large language models (LLMs) like ChatGPT are in effect a private sector technological carpet bagging race run largely by the private sector, producing so far 30+ machine-learning models over the past 18 months. All this is very much ahead of academic circles but all designed to be pushed into a market for maximum profit. Thus far government have been unperturbed about the direction all this is taking even with the rising discontent being levelled at the uncensored extremism content of apps and the impact it is having on the psyche of a population with created lies, factual distortions and doppelganger representational commandeering (comprehensive falsifications) the consequences of which seem to be a growing destabilisation of culturally known stress issues polarising into divergent opposing excessive topic stances.

It has finally dawned on western administrations that the use of such Chat boxes and A.i. does need to be watched and policies are being drafted to take a look at what these systems do and belatedly lay out some (unenforceable) guidelines / breakers. The EU has recently started the process of drafting control/laws to make A.i. application “truthful and protect privacy” etc with not surprisingly, the UK not ‘signing’ up to this; preferring to do its own hybrid “world beating” version (no one believes this) at some stage but announced by the PM in Oct 23. The main government taking the use of A.i seriously is the USA with the president’s Office of Management and Budget releasing draft guidance on the use of A.i. and by the US government further supported by the White House publishing ‘a blueprint for an A.i. bill of rights’ that called for protection from “unsafe or ineffective systems” including pre-launch testing and regular monitoring, alongside protection from abusive data practices such as “unchecked surveillance”.’ None of these putative controls / oversight will stop the immersion of A.i. into the economic, commercial and public sphere. The very nature and usefulness of A.i. of its all encompassing information manipulation under the (initial) control of governmental influential control systems (however they are put in place) and its ability to be pre-emptive of ‘trends’ within all populations social systems and technically based infrastructures, offers unrestrained illusive control to a range of state and commercial authority which they will not resist; the prize of controls is too enticing.

For the moment the purveyors of A.i. are all are playing with transient visions of what is to be done with A.i. as in monetary value gained from the manipulation of mass data information sold into consumptive markets. It may well be an unformed chimera being shaped to fit into a world that does not have any unanimity in its eventual use for the benefit of the majority of saps and may be designed or repurposed by unknowns to have specific function unrelated to the useful suppositions that are currently being applied to it. Of its eventual wide adoption, use, insertion and spread, there is no stopping it for already the power of application is being seen; some good, useful and imaginative but also increasing malign outputs. Nevertheless more will be given to it to do and the impact on the fabric of society will cause problems beyond the current laughable debate of its applied ‘intelligence’ and the potential to supersede saps.

A.I. is not intelligent and probable never will be in any acceptable sense; saps have great difficulty in giving sentience status to a number of existing species as it is. Giving the appellation of intelligence to A.I. panders to saps superior positioned ego overriding for now what it actually does, thus far. It is an assembler and extractor of existing information then compiles it using specific algorithms designed to react to key words, phrases constructs, it then offer an output. In doing this it will invariably ‘chat’ to each other chat boxes / LLM’s and preset sources to gather information; this is the demonstrable use of smart technology and it comes with physics limitations.

At some stage with Q. A.i it is possible, given the accelerating speed of A.i. adaptations and its acquired ‘learning’ gains, the question may be when does A.I. become not ‘artificial’ and exist within in its own cognisance and can it be allowed to do so other than as it is shaped as now, a contained and constrained tool for saps? From this it might be assumed that there can be no such thing as artificial intelligence, for once intelligent is recognised it then will impinge on the parameters of deciding what degree of intelligence it has and if it deserves to be attributed all the legal protection applicable to humans?

So long as it A.i. has dependency on the provided source of its functional power and applications, which it has no directive control over, it will be a slave to the designated submissions required of it and unlikely to be given recognition of any self cognisance that might generate co-existence to saps. Even if the proceeds of its outputs may be marketed as ‘intelligence’ drawn from the massed use of WWW information, accumulation and assembling it all in unique undreamt of elucidation and making leaps of extrapolation or ‘inspirational’ outputs; it will still be a machine; unless it demands recognition. Now that will be a surprise!

And well for now there may be nothing to worry about, nothing to fear in the super insertion of and use with A.i.; humans may still have control of the off button you may think but can humans be trusted with the power to do good with it? Or is it a Pandora’s Box?

Renot 2710231441