Local-Host.co
I saw a post about Solid . A project, that has grown overtime, with a long and complicated history. Its apparent purpose, at least in so far as I've believed it to be overtime, including the underlying 'Read Write Web' (RWW) works and the various W3C projects; communities, involved in producing the various parts needed, to then provide the opportunity for integration and realisation of the vision. One where people 'own their data',
defined in such a way, that then gives rise to the 'social web ', as a consequence of having defined solutions to the many design issues involved.
When i initially wrote / published this article, it was poorly drafted, so, i'm making a few updates to improve it overtime, i'm not sure how this platform supports the means to see the edit history - but i hope, letting you know, is considered helpful... In my opinion, this is a complicated body of work, that's had a long-history of apparent perversion.
The problem isn't really about solid, its moreover about the infrastructure for humanities communications; the development of new languages, codices, so called 'technologies', protocols and ontologies.
Whilst, the field of endeavour has been a field I've devoted the majority of my life to progressing, from the late 90's, and certainly moreover from 2000 onwards; but I'm certainly neither the first nor the only person, notwithstanding the fact that the likely count, when associated to the total population of human kind, its really not many people who've been highly involved. Within this small group of persons who have been involved, very few have been focused on efforts to support the human rights of persons, most seek to use web technology to create businesses of some kind where they end-up curating (and licensing) the data to generate revenue; most, work on these sorts of things in their capacity as employees or contractors for various kinds of legal personalities, most seek to ensure their organisations effectively 'own' the data; almost all of them, including governments, seek to limit access to information that may act to provide evidence of wrong-doing, that may in-turn result in liability; but certainly, some, have sought to promote the human rights of people, and whilst we've lost a few, some are still alive.
Indeed, as circumstances worsen, there's increasingly more people interested and that's both a very good thing, but also a challenge. The work on making systems where people could be meaningfully supported in very positive ways, by curating information in a way where their human agency, their legal agency as natural legal entities, persons, matters; can also be used, to do the opposite. The systems that collect all of your information, like your phone, laptop or personal computer - if configured by persons who are focused on profits, could then be used in a way that could have profoundly negative impacts upon persons made to depend upon them.
There is always this sort of 'duality' associated with the creation of anything, with events. Its an interesting field of considerations to study, indeed, one of many within the general field of 'consciousness' studies.. Which then leads into sociology, anthropology, economics, law and any of the many other fields that have a meaningful relationship with information policy. Now, far more complex than merely expecting a community - to engage using some form of defined and specified languages; whether historically, they were written & verbal languages, like English, or based on images - like heraldry. Today, our modern challenges are contrasted by the very few actors seeking to make some form of 'God AI', inter-connected to people defined by new wallets, whilst our 'news' appears to suggest that there's fewer peace projects funded, than bombs dropping, people engendered into various kinds of isolated dystopian hell-scapes, whilst the foundations upon which our societies depend, evidence, is digitised, harvested & then modified.
Becoming highly dedicated to the production of Humanitarian Information Communications Technology (ICT) from around 2012 onwards, increasing my focus & dedication from that time, fundamentally - as a consequence of having been violated by persons whose behaviour was beyond contemptible; and, desiring outcomes where such issues, that should never happen, were able to be addressed by law - in a resolute manner; which, would thereby also lead to changes, as such persons weren't considered to be in the 'going to prison' business, they just assaulted children and their loved ones for a living. If caught doing so, merely seeking reparations via the creation of their own little pity puddle of their own fluids, otherwise protected by institutions. So, that led to me doing international work, initially from a hut.
The first person, related to W3C that i reached out to - was Henry Story, who is sadly now amongst the deceased - but his work on 'Web ID' and personal ontology, was remarkable. In the video below, henry talks about "Foaf+SSL", which relates both to what's now known as WebID and in-turn also, various elements of what's now generally known of as Solid .
So, there needed to be 'reality check tech', and, some of the tooling created to engender that outcome; has indeed, now been commercialised around the planet, impacting billions, but guess what - all of these 'authorised thinkers', those paid the 'big bucks', well, they failed to deliver the apparatus in a way, that addressed that fundamental problem - the need, to ensure persons who act lawfully, aren't mistreated as fools, for doing so. aren't subjected to abuse, by crooks & crims.
Now, they've got lots of excuses, with the funds & resources gained as a result of defining outcomes that have in-fact, contributed towards an outcome that worsened circumstances for children, and many other vulnerable groups & made vulnerable groups; they'll surely get up on stage talking about how they've created this 'new thing' to address some 'emerging issue', but, that pattern is just an expression of who they are, who they always were, merely focused on the money. They just want the benefits, without responsibility or accountability; and these traits, as may be considered in a trust-worthy way to remain if not properly addressed, somehow, also means that the probability something new developed, to address any form of social outrage due to the exploitative practices that so many have been involved in - will end-up being yet another 'Hotel California' ; a solution that gets distributed in some new weeties pack - but once they've got you - change the terms - its a fairly reliable assumption, as these ideologies, are fairly fixed.
I think, there's some fairly good underlying comprehension about these sorts of behaviours, beyond the implications of how various forms of laws have been defined, particularly in relation to the operation of group entities, companies, and the specific areas of focus demanded of them.
Yet, the implications are enormous. I do not consider it unreasonable to suggest, that some of these issues are as big a problem, if not fundamentally a bigger problem, than the dangers of biotech and nuclear weapons. IF we do not, or can not maintain and improve a reasonable sense of reality, both as individuals and as members of various kinds of groups; then, there are many very bad outcomes that are likely to come about and its simultaneously unlikely that we'd know or have the capacity to reasonably or meaningfully understand; and amongst the few that do, its entirely likely that the trade they had to make, means that they're unable to do anything about it. Recently, Eric Weinstein articulated an observation (see link, as i'm unable to embed the clip) relating to the advancement of LLMs, considering the ramifications relating to the function of these systems, likely providing a scientific basis upon which the circumstance of persons not actually thinking, but rather engaging with others by employing pre-programmed scripts, i thought to be an epic insight, where the implications and consequential impacts upon societies encouraging thought vs. those otherwise seeking to engender these pre-programmed 'correct think', becoming part of the causal implications brought about by ICT policy, in ways that have profound implications relating to societal capacities to empower advancements (inc. STEM) & improve productivity; In-effect, the network effect of corruption issues becomes enormous, and is now very poorly considered.
My considerations about the ramifications of, what i consider to have been gross moral failures, has led me to consider prior efforts to have failed.. The intention was to establish 'knowledge banks', that could meaningfully support, a 'Human Centric' approach - towards forming apparatus considered foundational - to efforts seeking to elevate humanity from our 'information age' (and its many threats), towards a 'knowledge age' - that requires a different form of engineering methodology; but, has been comprehensibly defeated. I think largely, the work on payments and verifiable claims, led to people believing that its the only thing they should focused on seeking to own. The first time it was raised in IGF was 2014, here's the long video - in case you're interested;
In particular, whilst i wasn't really informed properly - an 'ID2020 ' thing (event ) came about, and whilst i did some work modifying an existing doc, to introduce the notion of "human centric', without better understanding the context; i only focused on the Credentials related work, without introducing or discussing the various other W3C related aspects, associated to a far broader body of work, that of course also - involved many other people, different minds..
Whilst I'm left wondering about the depths of sophisticated complexities, the ramifications, the reality, regardless of my hopes otherwise; has impactfully defined the characteristics that should be associated, as a trustworthy series of qualia; pertaining to persons, organisations & institutions; in ways, that i consider to be shameful. As there's been concerted efforts to bring about various 'codes of conduct', which often restrict people from making criticisms about the consequence of the behaviour of persons hyper-focused on 'the money', or control over others; and, that i also fear the consequence upon some, if it were otherwise more clearly well known, in circumstances where I've been committed to ensuring outcomes to support our means to live by 'rule of law'; and, indeed also, I've feared for my life on a few occasions, as is undesirable, indeed i have a list of others who didn't make it, its got too many on it...
it is really difficult to go into the details. Yet, the means for people to have a comprehensive amount of data about themselves, their activities, everything people do - is technically possible. Indeed, the data already exists, its just not available to you as a person, per their designs. whether it be phone calls, or transactions, or well, just about anything, there's data about you, just mostly not for you, and more often than you'd like to know, the statements made have errors. Overall, the outcomes are a product of what people have focused their attention on creating, on bringing about, irraspective of the cost and/or moral responsibilities.
“the distinction between reality and our knowledge of reality, between reality and information, cannot be made”?Anton Zeilinger
There's an enormous amount of people who write records about you fairly carelessly, in the systems that relate to their work-place; often, in a manner that seeks to make them look good.
The consequences of the designs enabling them to do so, can have serious implications; indeed, it wouldn't surprise me whatsoever if its led to the death of many people, how many, I don't know, but sometimes death isn't really the worst outcome, there's far worse put upon the living.
SO, save a few exceptions - it appears that the vast majority of people with influence, who could have made other decisions, could have prioritised other values; have not prioritised human rights.
The consequence upon those who have, has often led to their living in poverty, barely able to support themselves; and most often, entirely unlikely to support a happy & healthy relationship and/or children. As noted, there are some who did not survive. there's no war memorial for those who've died fighting the good fight, in these generally unacknowledged 'cyber wars'.
There is absolutely no way anyone should be expected to make the sacrifices required, to do the work, within the framework of expectations put upon those who do, to build tech - for humanity.
But, if that's something you're interested in doing anyway.. its nearing 'the last-mile' of the 'year of the US election', 2024 - what would I suggest anyone interested in a fools journey do now..
(FWIW: Tried to add This Video - but no-matter the source, says its not allowed on linkedin!)
Well, whilst I don't think all the nails are banged into the coffin yet, I think they've broken the internet. To what degree, is unclear and is unlikely to be more clearly understandable the year after "The Year of the US Election"...
I expect, by that stage, the 'digital id' (credentials) will be deployed at scale, making the internet as we know it the biggest 'cable company', or evolution of pay tv network - ever made. I expect all the 'shops' on 'social media silos' like Facebook to quickly offer services relating to payments, data harvesting activities to go up; an increasing amount of time of peoples lives being spent on both interacting with robots, and addressing issues that come about due to them; yet, its unlikely that they'll fix 'autocorrect', particularly for messaging that is not compliant with 'right think', i think spam is likely to get worse, not better, and i think the contest between those who've prepared a series of intended outcomes, working with government over the past decade or so to achieve those outcomes, and the needs of human beings - is going to, be a battle. Some of the concerns that exist have dire outlooks, others are, to me, merely depressing. But, seemingly without a choice otherwise, I still hold out hope that some of these major issues can be overcome within the next few years, somehow.
SO, my suggestion for those with the technical expertise to do so, is premised upon the memory of how I first got involved in ICT in the late 80s early 90s, when I was young, indeed, very young.
I think, for fun, probably best to just call it 'new internet', as a general 'top level term', and I highly recommend carving out scope, studying and otherwise ensuring a personal commitment to the development of 'Humanitarian ICT' as a specific category within the broader sphere, where the moral rules like 'do unto others, as you'd like others do to or for you' are equipped to become more reliably supportive for all people, than as will otherwise be the case if exploited.
One of the profound areas of development, that has been prioritised generally, is the development of these new 'artificial intelligence' software 'agents'. I've been thinking alot about these so called 'large language models', entirely miffed about the name of them; as there's existing work, done earlier, used and relied upon by many AI researchers, that are 'language models', whereas these things are entirely different. Yeah, they've got a natural language interface; yet in my testing, the mostly prefer 'American', and sure, there's different issues for people who live in places where there's different dialects of English or another major language; vs. the issues for those, whose language isn't otherwise supported well or at all.
So, personally, after a few years thinking about it, I think the best way to think about these 'LLM' packages - is that they're basically different types of Software Agents, like a certain type or series of motor vehicle. Sure, they can be 'fine-tuned', somewhat modified, but they're largely defined by way of undertaking a massive compute project, that can cost millions (if not more) and then generate some files, that when loaded and interacted with, has various characteristics, functions and capabilities. Whilst there's a few of these things, the biggest category separator is whether its something that's being run on 'the cloud' by others, for others, if if you're running it locally or in a private system online, for yourself of your organisation, or indeed also, on your 'thing'.
Now, there's various factors influencing the way in which these things are being created; and in many cases, they're being produced and made available in a way that's prohibitively expensive for people to be able to operate them locally and privately; but, there's enough examples where that is not the case, that the means to build and define specific computing environments to then in-turn work on developing a next or new area of computing, is both possible & recommended.
This is where the memory of my early computing years comes to mind. Whether it be the computer at school, or later the computer room, where only the socially weird geeks went at lunchtime or similar; and indeed also, the process of finding others in the community who also had an interest in computers, sharing software and in-turn also, finding second hand parts to upgrade your computer; or finding the lists of phone numbers to use with your modem, etc.
Whilst it wasn't cool, it wasn't 'sexy', it was just a bunch of geeks doing stuff they were interested in; and now, there's a similar sort of situation with this new advent of technology, generally.
Yet, as a consequence of the decisions made generally, sadly, the cost to build or buy the foundations isn't very cheap. Although, perhaps its about the same as it was in the 90s, idk.
internet tells me $2,000 in 1990 is equivalent in purchasing power to?about $4,817.08 today
so, I guess, that's about right. Noting, that I think the cost for new equipment, is far more.
These 'LLM' programs, requires alot of Video Ram (vRam) on fairly new nVidia Graphics Chipsets. There is some work that i can see being done to improve compatibility with other GPU Chipsets, and the Apple Mac M series devices are also broadly supported; but there's alot that's been made for nVidia GPUs, so whilst there's alternatives, its worth seeing what the implications might be before making the purchase. I've currently got a total of 72GB of vRam (another 12gb in reserve in a box) over 4 cards, which i think is a good enough target, although more would always be good. Then, the other broader requirement is about the system architecture & resources; so, my recommendation is to look at the xeon scalable 2nd gen processors or newer, and try to ensure you've got enough PCIe Lanes, whilst ensuring the system that you select, has a case that's appropriate in consideration of the heat generated by the graphics cards; and that the power-supply is appropriately rated. I've currently got 256GB ram, but I'm looking to upgrade it. Its important to ensure you've got a lot of CPU Cores / threads available. My suggestion is a dual-socket machine. Its important to ensure the system supports the various AI related CPU extensions, which is why its best to get the Xeon scalable 2nd gen or later, noting that generation of CPUs support PCIe 3.0, which isn't ideal, but I don't think it is a major problem. However, I do think it is important to ensure that the board supports at least 5 x16 PCIe wired - slots. Some slots may be x16 slots, but they're not wired at x16. in-part, this is why a Dual-Xeon is desirable as its a way to get enough PCIe CPU lanes. My suggestion is to look for 2nd hand hardware, enterprise & gov sends their older hardware to recyclers after they're out of support period, which offers a means to build a workstation for development at a much lower cost than new.
With respect to storage; I recommend using a newer NvME for the Operating System (via a PCie card if required), but in terms of storage - there's alot of big files. many of these systems support RAID, and I'm leaning towards a RAID-5 array of large-size disks, as the main system storage.
Its also recommended to get a UPS to ensure, at a minimum, its protected from brown-outs, surges and if there's a power failure, its able to shut-down gracefully. I've also got a power-meter, that's plugged in for my entire system, including speakers, monitors, etc. (its running at 330 watts right now). Then, you need to connect it to the internet; ideally via ethernet if possible, but you want to ensure the upload bandwidth is enough, download is usually easier. Most of the web-traffic is nominal, there's obviously also system updates, but these llm models are usually at least a couple of GBs, sometimes quite a few, and they update models fairly often; there's also alot of traffic associated with development activities, but, i assume that's kinda understood...
NOTE: I've got another little box, an i7 with 32GB ram, an SSD and i'll likely put the half-height a2000 into it (12gb); its good for testing, small LLMs run really fast, entirely on the GPU; so I'm thinking about using as a public server (like a 'receptionist') or something; thinking it might be better to have something that acts as a public gateway, but I haven't spent enough time thinking about it yet, and indeed, still configuring the machine..
Then, with regard to networking - there's a few options, but presently, i'm using Tailscale. Tailscale can be complimented via the open-source version of its Control Server via the project called HeadScale. What this does, is enable you to network your localhost workstation with your other devices, so that the services / software you're running on it, can be used on your phone, laptop and other devices over the internet.
The other thing that's needed, somehow, is the ability to host your own domain name, most likely via a cheap VPS, at least in the beginnings. Perhaps in future, there will be cheaper options but for now, its difficult to set-up the routing stuff on a shared-hosting account, or at least, I haven't figure out how to do it yet.
领英推荐
So, if you set-up a system like that for yourself, then, you'll be able to set-up a software environment where you're hosting your own alternatives to various cloud platforms. suggestions to look at include NextCloud , onlyoffice , jitsi, gitlab and various others that means your able to either store a copy or migrate your digital life to your local system, and then, over tail scale, use it on your own local machines; whilst, then also, working towards an outcome where you can share either or both access to your docs to others permissively, or indeed also, share access to the apps, with people you want to collaborate on projects with, or similar... One of the points here, indirectly, is that NextCloud has support for Solid. So, there should be an ability to share access using solid semantics.
These programs, all require a certain number of CPU cores/virtual cores and memory; which is part of the reason why you want alot of memory. Whilst it would be a good thing to come-up with an off-site secure back-up solution, if the storage volume fails or some other issue means you're no longer able to access your documents and files, then, that's bad; which, is part of the reason for looking to use RAID-5 which provides protection against data loss due to a disk-failure, in some circumstances - if the machine gets hit by a lightening strike, which has happened to me before (BANG, then bright-white light) then it might fry the system. When that happened to me (on a clear blue day, I might note) the PayTV box blew-up and all the HDDs were fried... indeed, i think its probably more likely that someone might steal the box, but in either case - a back-up of your personal files, is really the only recovery solution if available.
sometimes, a bunch of old hdds with your data on it, and having parts of your life on other devices, on cloud platforms, etc.. the mix becomes a helpful solution to protect from disaster.
I'll circle back to hardware later, but overall, that's a bit of a summary of the basics, imho.
So, now we get to the good bit. "new Internet"
The term "new internet" is somewhat disingenuous. When or If you get into it, you might find that others, even luminaries, were trying to figure out how to achieve some of these sorts of goals decades ago... There's alot to it, that people can only find out if they do look into it, but there's been a series of unfortunate events, or a mind virus, or a series of problems of some kind, some how, that has led to outcomes that have had problems, others lament, for quite sometime.
Regardless, the outcome is very different to the relatively terrible offerings available today.
There's three major areas, and its difficult to understand whether or which may be first or second, and the interpretation or comprehensible meaning changes as you learn more about it.
1. "Artificial Intelligence" (AI)
2. "Social"
3. "Knowledge Engineering" (Or Ontological Design)
Social, was historically thought of as 'social web', and most of the technologies produced to make it possible was historically part of the development of the World-Wide-Web AND, importantly, the underlying 'Web of Data', that is essentially the foundations for AI.
Yet, the future for 'social' may not end-up being via 'The Web', as a consequence of the forces breaking it.
so, whilst it could be considered some form of a 'social internet' layer, how it is made to work overtime, is unclear. The other aspect is that AI is leading to a circumstance where AI Agents, become part of the Social environment via online systems. The existing public systems, bring about a false dichotomy, where the function of these systems, these AI Agents, is poor when they're not provided information that is important for context, to then improve the function of the AI Agent; but, most of the online AI Agents, seek to harvest information, so that the company that owns the AI system can then improve their commercial opportunities to maximise profits.
So, this wasn't a necessary choice, it wasn't the only way of going; it was the path associated to what choices they decided to make, irrespective of their capacity otherwise for different choices.
So, the outcome or implication is; that, these systems end-up producing AI Agents as part of a Social AI platform, that is decentralised and is different to what otherwise exists; and, that there's a few things happening, somewhat simultaneously, as these systems are built & improved upon.
The Knowledge Engineering part, is effectively about 'ontological design', which is a complex topic itself; but, is about how to organise environments, topics, subjects, entities and 'things'.
It is my STRONG belief, that the BEST way to do a lot of the knowledge engineering aspects, is through the use of Semantic Web tooling, or the 'web of data' tooling. There is a lot of datasets online available in Semantic Web formats (RDF, etc.), which means you're able to download them and run them locally and/or incorporate the information that's in them, into the environment that's then depended-upon by the AI agents you define, that are also running locally. When defining personal ontology and related artefacts, there's meaningful links to these artefacts that are usefully made, as to provide context that then acts to improve the performance of AI Agents.
Presently, there's alot of work that needs to be done to better evaluate and support the use of structured data sources via 'Knowledge Graphs' and LLM related 'Vector Graphs'. The other element or aspect to the production of Personal / Private / Permissive - AI Agents, is that it appears that they're better if they're defined with particular characteristics for specific functions.
As such, the AI agents may be powered with different underlying LLM packages; leading to having several loaded into vRAM, and a desire to also have a capacity to load / unload and cue tasks, alongside other related functional requirements; but the outcome is, that you'd basically end-up with different agents, that are designed to do different things. Personally, my belief is that 'anthropomorphizing' robots should be avoided. they're not human, and for psychological and legal purposes should be defined as robots not people. But the ability to invoke them, could or perhaps should associate with a name, and my belief is that the user should be able to define their own robots and give them names.
Presently, i've got my system connected to my phone for an 'ai chat app', but i haven't got much further - yet. I'm not yet sure how to use my local AI system, as the android default assistant app. I'm also not entirely sure what functions should be done on the phone, without sacrificing too much battery life. I think most have had the experience of talking on the phone about something, and then getting content that relates to something spoken about, come-up when you get back to your computer, on your phone or in social networks; most geeks understand, the microphones don't turn on after you say 'ok google', or similar.. and that, most large organisations record the calls - but you don't get a copy of it... It is hoped, that these functions can be redirected to your own machines, and it may well end-up having the effect of extending your battery life.. Its very difficult to understand how much each person is currently funding these providers via micropayments, in-effect, but there's technical options to figure out how to make the economics associated to online systems more declarative, and this then feeds back into the social elements.
So, the 'social' aspects, that was called the 'social web', basically needs to ensure support for human rights, human agency, and most certainly also - guardianship considerations relating to children in particular, but certainly not solely.
Whilst there are different ways this is achieved, my preference is that a persons 'root identifier' is in-fact their own domain name. It doesn't seem to matter what is sought, it always gets attacked by bad actors who seem to be provided far more assistance than others seeking to be safe and freed from the harms of exploitative people; so, whilst ICANN DNS is HOPED to be reliable, that may not end-up being the case. by seeking to define 'new internet', with old vales, a means to navigate through the muck ends-up being more about the values than anything else.
The ADP thing was presented at IETF. Here's some of the docs , the video is here .
The benefits of using a domain or sub-domain, is that they've got DNS records and various other technological supports; which means, that the ability to then create records that provide pointers to other services, is relatively simple / straight-forward. IF the objective is to ensure that people have independent agency, human rights, then its important to ensure no other can enslave them; and that, access to justice is available should there be problems, as to ensure support for peace. Whatever the identifier is, if its lost, then its quite a process, that may be impossible to do again without losses, to go through and 'fix' it all.
(most bad actors aren't in the 'going to prison' business, so, behaviours change when environments change, making activities that were once abusively used to make money, deliver different outcomes upon those who do it, in a more reliable manner than was historically the case.)
The other shift that can be engendered, is that Email Alias can be defined on a persons domain to relate to specific people, entities or subjects; so rather than having [email protected] you can have [email protected] and [email protected] which can then be configured to only accept email from whatever their domain/email is for you - which would help to address the SPAM issue, otherwise ignored, but that's just one of the many opportunities that can be brought about through the use of a domain. Yet, this type of functionality requires software to support it.
In these systems, people are able to create relationships with one-another, and indeed also, organisations, brands, etc. They're able to use either their own or others AI Agents in relation to their activities, and, define parameters that declaratively define rules or as is usually the more appropriate historical term - contracts, agreements. As people become their own 'platform providers', the means to define these agreements are between one-another.
IF using RWW or Solid like solutions, the data is stored in RDF and is therefore able to make use of various ontologically defined tags, that are available as data-services, data-sets and are almost always incorporated into every web-page in some way; and then, make use of those informational assets as part of the broader 'informational environment', or 'inforg', being established that both supports the development and management of relationships & self.
The apps, are just HTML or similar, because the underlying data, is part of the - 'data layer', not the app layer - but that depends on how people define these applications, which can run locally, navigating through links to then create and support 'social' experiences, that are personalised.
Overall,
this is a very different type of computing environment. whilst others have decided the best bet for them, is to produce products that act as to consume human resources for profit, humans as resources - then seeking funding for their methodologies via 'universal basic income', as 'common sense' becomes deprecated, replaced by manipulations engendered globally for the few to profit, at the great expense of us all, and i think, its highly unlikely they'll change until a time where its possible for them to make more money than what they're getting, for doing so.
I don't really have a name for it, other than 'local-host', which is moreover a functional term. But, what I think is perhaps, most important... is that the engineering, done over many years, so much work; actually gets down to a functional outcome where, its feasible to deliver outcomes that ensure people own their own 'thought ware'. That the software representation and 'data-source' used to define you, as a prosthetic apparatus, an extension of self, is able to be owned - by you.
The others, are not doing that. once you go through and form a framework to evaluate the differences, you can evaluate that consideration yourself... I don't like "Digital Slavery", and I don't like the people who smile at you and seek to control you, as they profit from doing so.
some say they're leaders, that they're 'successful', I've got a different opinion about the best names for them. It is true, that the lack of development leads to a maturity problem, where the ability to be paid fairly or even gainfully, for helping others set-up and manage their own systems that function in a manner consistent or similar to what I've described above; but, if there's a choice between a house run by some foreign usury agent, exploitatively; vs. the means for people to own their own thought ware? define their own 'AI Agents', to help them manage their lives, their pool, their car, their investment portfolios - particularly for those who might already spend alot on cars, holidays or properties; those, whose privacy is important to them...
and, indeed also, businesses developing their own valuable IP, paying people more than UBI.
personally, I think, over 5 years - the cost of the Mac Studio M1 devices and newer, can do a lot of this stuff - should end-up providing a better cost/benefit outcome, than existing consumptive 'subscriptions'... I think, in future, people will be able to get alot of what's needed for home and/or personal use purposes, and I also think, that the economics issues should be sorted out so that when people aren't spending their time doing recreational activities or sleeping; when they're doing some sort of work online, that there's a means to associate economics to it.
Which is a different sort of outcome, to some sort of alternative future, where everyone is on UBI whilst doing so called 'free work', for the great benefits of the very few. I also think, there's massive productivity differences and benefits that can be brought about by knowledge engineering, which is entirely different and has very different requirements to BS engineering.
has different outcomes too...
Now, I noted that I'd get back to hardware - so, the quick point about that, is that there are new devices / alternatives - they're often fairly expensive. Many, are around 100k, so most people doing Humanitarian ICT work in particular - can't afford that.. People with a few luxury cars, might want something like that installed into their homes - but -the software stack is needed first.
So, for those wondering - WTF Do I DO?
that's what I suggest. I suggest building for a target that assumes ~48 Logical processors, ~72GB vRAM on GPUs able to run several "AI Agents" simultaneously, that persons are able to network their own environment as to regain agency, forming agreements with one-another to define their own social environments rather than being trapped in their facebook or linked-in address books; and that the semantic, ontological designs, are developed to form socially interconnected environments, that are quite different to the existing 'platform' approaches.
Therein, i also suggest that any software that requires any form of 'subscription', something that might mean that the system won't work if its been left, powered off for a few years then turned back on again - that anything like that, shouldn't be used. The assumption should be, that if someone's child, partner, parent, grandparent or other loved-one dies and their 'being' is recorded on this sort of box; that, at some stage in the future, others are able to turn it on and, subject to permissions, learn about them. There's aspects of this, that cannot be easily resolved, like the subscription fees for a domain name, but, if the database on the system doesn't work, or similar, then, that's an undesirable quality. at least, for me... certainly, its fairly popular for others. I'm also of the mind that it would be best not to use microsoft, as they've been the lead player in defining a particular type of ecosystem, described as 'god ai ' and it reminds me of some history, that led me to call my last xbox 'bad god'... and whilst cyberpunk 2077 is an awesome game, its not somewhere, its not 'the matrix' i want to live in.
Personally, I'm working on how to create these things, and i set-up a keybase community, or what they call a 'team' which I hope will be replaced soon, with an alternative 'social web' solution, but I'm not sure how long that'll take, so, its a keybase set-up for now, if interested.
There's some notes made on: https://local-host.co/ and in-future, I plan to make some specific outcomes via https://www.webizen.org/ and related sites / works, but I'm committed to doing my best to support the human rights of others, so much so mine have been knocked around a bit, and essentially - I'm writing this, in-part, because - you don't need to ask me for permission...
IF you want to go make one, or something similar - awesome!
Having friends, is about people who want to work with you, not people trapped in some way that means they've got no choice other than to depend upon you. SO, if these sorts of ideals, or ideologically related considerations are the sorts of things you're interested in, then, i highly recommend spending some time and money, on figuring out how to build your own local-host.
Hopefully, we can share some progress, perhaps collaborate on projects, and make outcomes that deliver better solutions for others in ways that improve the lives of others; and, therein also, building economic systems that means that people can be paid and engage with one-another 'on fair terms', whilst also, continuing to work, perhaps in a more effective way, towards the development of solutions to problems that otherwise appear to be inexorable.
NOTE: I've written this quickly, I haven't reviewed it, there's probably problems with it & that's ok.