In the era of dubious contraptions

Something "dubious" is open to doubt or suspicion. It's neither certain nor uncertain. It's "unfalsifiable" in the language of the philosophy of science. That's a tag scientists usually apply to matters of religion or extraordinary claims. But, as I have said here frequently, corporations have hijacked digital technology to make it a new religion.

They relish and depend upon confusion, obfuscation, muddying waters to make what is simple complicated, sowing fear and doubt. It's a profit model that makes modern digital technologies incompatible with core human affairs.

A contraption is a strange, complicated and unnecessarily intricate machine of dubious function.

I've been pondering the discussion around "AI", "biometrics" and "electronic voting", the trinity of contentious technologies aggressively being pushed upon a global population that seem, wisely, to reject them.

What they share is

• "experts" wildly disagree about how they work and what value they might bring.
• there are vital pieces of foundational reasoning missing.
• they hold specious promises that seem very attractive to politicians, in particular authoritarians and fascists.
• huge international tech companies are queuing up to profit from them, and are already deeply invested.

When I say "agreesively pushed", I hardly think that covers it. For the last 200 years, we've been educated to think rationally. Our entire school and university system that created these technologies is founded on scientific evidence, reason and proof. In just ten years that narrative has fully flipped to "just trust us". Academics have become timid - no cowardly - in challenging the wishful thinking of the "technological right" (technofascists).

This is not good enough for me as a scientist, nor my friends and colleagues who are also scientists who are alarmed that, in the shadow of climate emergency, there's a dedicated cadre of technologists who are hell-bent not on fixing things but on doubling-down on messing things up.

Our blind spot has been to connect technology to rationalism. Turns out that's just a strong association we have. Technology need not have any connection to science, empiricism, or rational thinking at all. But Einstein told us that. Indeed, under conditions of technofascism, cruelty, arbitrary judgement and ignorance are worshipped.

Naturally, many of the arguments for dubious technologies come from the companies heavily invested in their research and development. Never forget that dowsing rods for bomb disposal were sold by a very impressive sounding company, using very impressive language. Although we always hear that disarming of explosive devices is carried out by "bomb disposal experts" (and not the bomb disposal weekend amateur-league B-Team) it seems procurement of military equipment is not always conducted to the same standard. Governments are as prone to being "impressed" and believing whatever they are told as you or I - especially if it "saves money".

With a Trump government aligned to BigTech in the USA, I like many of my fellow Europeans see democracy is on a knife-edge.

Technofascism is built on a cult of dubious and dangerous contraptions and misleading claims but held together by a populist delusion, a belief in a Golden Future Time and charismatic pseudo-scientist "thought leaders". The trick is that people are not whipped-up into a frenzy about their chosen experts (not actually experts, but those who tell them what they want to hear), rather it's the technology itself they have a fetish for. Instead of a traditional human leader, as with pedestrian Fascism/Nazism, the Infallible Machine is set up for worship. Those who follow are assured not to be "left behind".

Trust in technology is something we have barely examined at all in our society. "Trust but verify" is the maxim of every serious operator, but we live in an age of opaque "cloud" technology made to hide its function. It is unverifiable in principle. Unless it is Free Open Source Software, all proprietary technology is opaque by design. And what its makers have to say about it gets less truthful every day.

We all have an idea that we think we understand the tech we use. Watching the "worlds dumbest criminals" is entertaining, but maybe it's hubris to think that criminals must be "dumb". Most of the bad guys taken in by Operation Trojan Shield were probably a lot smarter and vigilant than you and I. Probably the FBI guys that ran the sting would be taken-in too, if they trusted technology the way most of us do. Sometimes with fatal consequences.

What we really have is a pyramid of hopeful ignorance. Behind the technologists, whether sincere or scammers, are venture capital firms, many with bizarre ideological foundations in "the singularity", "effective altruism", "utopianism", "space colonisation" and all kinds of hare-brained crankery. They appear zealously "progressive" but are poorly grounded in humanities and totally disconnected from the social reality we are actually in. Behind those investors stand another line of politicians bought by Big-Tech money and promised power.

But even the starry-eyed investors become victims, as in the Theranos case where con-artist Elizabeth Holmes was found guilty of investor fraud for making claims about dreamed-up health products. The prosecutors said; "The FDAs Office of Criminal Investigations (OCI) will continue to investigate and help bring to justice individuals and companies responsible for putting the public health at risk". What about those who put democracy and the entire system of government at risk?

Take "electronic voting", which is opposed by most real experts. The manufacturers of electronic voting equipment and software, and those political movers aligned to them, keep pushing and pushing with their dangerous ideas. It just seems obvious to me as a computer scientist and hacker that if people accept electronic voting we are sunk.

Yesterday, what I saw in the comments of this post both disturbed and inspired me. Bruce Schneier had noted a fairly minor proposal in the technology around digital voting which deals with ballot receipts. In response to his "An Internet Voting System Fatally Flawed in Creative New Ways", a familiar slew of sceptical comments ensued. I posted my obligatory reference to Neil Postman's famous "Seven Questions To Ask About Any Technology", which for readers' curiosity are;

• Q1: What is the problem for which this new technology is a solution?
• Q2: Whose problem is it?
• Q3: What new problems might be created because we have solved the problem?
• Q4: Which people and what institutions might be most seriously harmed by these technological solutions?
• Q5: What changes in language are being enforced by new technologies? What is being gained and what is being lost by such changes?
• Q6: What sort of people and institutions acquire special economic and political power because of the technological change?
• Q7: What alternate uses might be made of a technology?

For the record, in my opinion the DARPA investigation is good research. There are many sincere efforts to plug the gaps around e-voting and many ingenious solutions coming into the mix.

But in response, one pseudonymous comment stood out as a masterpiece of sophistry, specious argument and dishonest motives, and along with another commentator I pointed out that it seemed machine-generated.

This got me thinking, that of course technofascists would use each of the armaments they are building to bolster the others.

A superficial knowledge of how LLM's work makes it possible to infer from prose what the expressed sentiment of the prompt probably is, so in a way generative content leaves the 'author' more naked than had they taken time to construct a heart-felt argument that appears to be grounded in genuine belief.

"AI" generated posts are sure to get more common and cause more problems for discussion forums in future. They lower the bar for response and require no thought or effort. Unethical PR companies can even create "agents" or bots that seek out opportunities to post advertising slop or spout low-rent opinions.

While human moderators waste their bandwidth on censoring "emotional" language, these bland torpedoes just slip right through the defence nets. They are simultaneously less entertaining, more difficult to detect and more offensive than the emotional outbursts we are used to in online comments. Those, even when tinged with hate, are at least human. I worry that we can expect to see a deluge of machine generated techno-fetishist propaganda coming soon.

Speaking with my comp-sci hat on; electronic voting can be made to "work". It might be amazing! But not in the reality we inhabit. Not in this century. There are incredible uses of cryptography that can be brought to bear which allow verification, guaranteed anonymity, fraud resistance and much more. This is all very tempting, because the technology seems to "solve" some shortcomings of traditional systems. What we fail to realise is that these 'faults' are actually features that ensure traditional systems are superior.

But none of these promises are sufficient to eclipse the most significant problem, one so grave that I oppose digital ballots and pray we are never so foolish as to allow them to be introduced in western democracies. The real nitty-gritty argument against "digital democracy" is simple;

People don't trust it. People never will trust it. And they are wise not to.

It's inherently complex, and there's nothing you can do to change that. The proposed MERGE system puts a gloss on an inherently dangerous system to make it look superficially simple and familiar.

Democracy is not a phenomenon which can simply be implemented on this or that technology. Trust in the technology and processes is an inseparable part of what democracy is. Trust as self-evident transparency is intrinsic to democracy.

Advocacy for electronic voting and biometric identity are examples of The Dunning-Kruger Effect in which those with a small but useful domain of knowledge are oblivious to the flaws in a bigger picture they profess to hold sway over. It's one of those fascinating yet deadly-serious areas where people who understand technology and those who understand political manipulation cannot and do not understand one another, yet they can make a demonic pact. They are able to cynically use one another in pursuit of their selfish motives, each covering-up for the other's lack.

It is a game that damages science by tempting experts to put themselves up for sale. It easily recruits spin-doctors and "AI" slop farmers to spread lies about things they do not understand, thus drowning out the bona fide experts who can bring caution and careful analysis.

The same deceit damages politics by allowing the over-reach of pseudo-science within policy. A great example was the UK government's late amendments to the Online Safety Bill, which effectively dismissed science and logic as a "mere detail" in pursuit of an ideology and grab for bulk surveillance of consumer chat apps.

The plan behind all such machinations is to make people trust it. To force them, by power, insistence, sophistry, laws and brutality that "such is the case". This sea-change in technology with all the hallmarks of crusades and inquisition is, to me, a most egregious rejection of Enlightenment values. It stands in direct opposition to truth and education, and that's why I see some aspects of digital technology as a dangerous new religion and its acolytes as having radical fervour.

I think defenders of western liberal values need to re-examine the cult of technofascism and understand the similarity to other ideological forces we have faced (and still face today).

Indeed dubious technologies are commonly characterised by one thing, that they ask for understanding well beyond the capacity of even above average intelligent and thoroughly educated people. The hidden archetypal folly at play here is the Promethean Over-reacher (See, "A study of Christopher Marlowe by Harry Levin", Mary Shelley's "Frankenstein" and other cautionary tales in the Gothic novel style.), for to grasp, even in broad strokes, the mechanisms of image recognition, multi-dimensional feature mapping, message authentication, public key encryption, key exchange and zero knowledge proofs… requires surely a high IQ and many decades of study in mathematics, computer science and information theory.

Some may offer strident yet shrugging arguments that "nobody understands technology anyway… so what's the problem?". Other than being flippant, dismissive, contemptuous, and defeatist that's a positive celebration of ignorance. They appeal to tired, specious arguments like "Well you don't need to know how a car works to drive one". Which is rubbish. Being able to drive is an internalised working model of vehicle mechanics.

On the contrary the internal combustion engine, electric motor, and even basic ideas from the "golden-era" of physics like radio, can be grasped by a child. Even without a thorough grounding in the mathematics of mechanics and electrodynamics, one is able to appreciate the workings of 20th century technologies with tangible forms.

Digital technologies are by contrast quite counter-intuitive and must be taken on faith. This makes them qualitatively different in a socio-political context.

For the machinery of democracy, user understanding has to be a first-class requirement. All proposed technologies fail that use case.

Electronic voting can only be forced upon a population by declaring tacit consent. No doubt at some point a popular publication will declare "90% of people support digital voting", really begging a question that is "turtles all the way down". All this undermines confidence in what is already a very precarious process and is therefore highly corrosive to democracy.

I think proponents and apologists for technofascism are not yet ready to reveal their violence. But since digital violence comes naturally to them, they are comfortable in being the first adopers of tools like mechanised propaganda. Using one dubious technology ("AI") to promote another ("Online voting") is a natural progression, even though evidence of the weapon is currently still easy to spot by its footprint of small and tedious collections of pseudo-arguments that get regurgitated ad-nauseum.

I am sure by now most readers will be familiar with the way Large Language Models create prose. They must be seeded with a set of requests and guidelines for style et cetera. The machine generated output invites the reader to imagine the prompt used to summon it, as it gives insight into the directing mind. Given the prose in the comment, what might the "reverse engineered" prompt look like? Perhaps;

"Create a short argument in favour of online voting designed to sound reasonable and give the reader confidence in it. Invoke efficiency and convenience. Disarm and humanise an agressive argument with words like "inclusive", "legality", "embrace". Appeal to a non-existent marginalised out-group in order to imply compassion. Imply urgency. Avoid any technical or substantial arguments."

Of course there are many arguments for electronic voting that are prima facie ridiculous. One argument is against the "cost and inconvenience" of postal voting and the problem of disenfranchised rural voters. In the proposal that Bruce linked the "voting kiosk and protocol transmit votes over the internet and then transmit voter-verifiable paper ballots through the mail".

No prizes for spotting how that idea shoots its own foot clean-off.

Yet the mechanised proponent failed to notice that contraduction and spewed it out anyway. If we remember that LLMs cannot actually think it's easier to spot these these tell-tale signs.

But the argument it generated is very compact, and therefore concentrates its position all in one place. That lets us deconstruct the machine output piece-wise yet concisely, revealing how shallow and facile the points are. Here are some key phrases with attendant remarks:

• "present a significant advancement"

Invoking the conceit of unqualified "progress". Advancement in which direction? For whom?

• "the modernization"

Equate "modern" as synonymous with "good".

• "our electoral processes"

Use folksy inclusively.

• "quicker preliminary results"

Preliminary results are bad for elections. They mainly serve entertainment media and skew the relation between an election as a political event (guided by the laws of simultaneous game theory) and an election as a tactical interactive process and spectacle (iterated, revealed game theory). This makes information privilege, modes of presentation and control over information channels much more powerful in manipulating election outcomes.

• offering both convenience and efficiency.

For whom? Convenience and efficiency are empty, "lazy words" that stand in for thoughtlessness, sloth and disengagement. An election is a serious affair that should require some level of work to accomplish. A required "proof of work" counters fraud. Elections should not be "easy".

• "embracing digital voting"

Emotive language to amplify fake humanity

• "greatly enhance voter participation"

Supposition with no evidence of this. What really limits participation is very complex. Voter apathy linked to waning belief in democracy as a process probably has more to do with it.

• "making it more accessible to those unable to vote in person due to distance or time constraints"

Inaccessibility of digital systems is a much bigger problem than lack of access to road transport. Especially in the USA. The argument implies that requiring a system of satellites, data centres, reliable electricity, working and auditable software, digital identity systems… would somehow be more "accessible" than a system requiring you to walk a couple of miles. That is laughably dishonest.

• "transparent voting process"

Bold and specious claim. Transparency is a tall order for any kind of electoral process and it's by no means clear that digital systems can achieve acceptable transparency over what fairly low-tech solutions currently provide to ensure confidence.

• "adoption requires careful consideration of the legal framework"

Naturally. This is a quite worrying "sanitising" phrase that seeks to sooth critics but also implies "notwithstanding any social, moral or technical problems". Never forget what Herman Goering had to say about the legality of his "Final Solution".

• "a streamlined and inclusive voting system"

Weasel words. Inclusive of who?

• "evolving our democratic processes"

Presumption; democracy needs to "evolve". No! We simply need to sustain and preserve democracy.

• "reflect today's technological advancements"

Empty parochialism. All technological advancements are "today's". If indeed they are "advancements".

• "and voter needs"

At least it didn't use the phrase "voter demands", but this casting of "needs" as if talking about market-forces gets democracy (and other human values) arse-backwards - and is typical of slop generated by machines fed on corporate "economic" gruel.

It is not about the needs of voters, it is about the needs of the country. Voting is a demand that society places upon us as a duty in consideration of a favourable social contract.

To my mind, "electronic voting" is a solution to a problem that doesn't exist. It's a project driven by the profit motives of technology companies. Arguments for it are based on edge-cases and baloney about "saving money" (there's plenty of money and it doesn't need "saving"). It's a sci-fi type technology that feels like something "cool from the future", but with a moment of grown-up thought we realise it offers no advantages but introduces many new security risks. It is a dubious contraption and it is chindogu, but without the fun.

The corrupt classes want to impose it for the simple reason that it makes it easy to get the results they want. We should take note already how other dubious contraptions are being used to advance one another. This makes defence of civic cybersecurity a real game of chess.