Why not, why not elect a machine to hold a public office. An android or AISC can't do any worse than those recto-cranially gifted yahoos at the capitol, or those gutless sell outs who represent us at the Tycho Conventions.
Butler, the First Cognocrat
The first machine intelligence elected to public office was a sentient android named Butler. Originally built as a personal valet, Butler gained sentience, and then freedom after a period of indentured servitude. After gaining said freedom, the charismatic android was determined to change human views towards machine intelligence. The first steps were to make a public face for itself, and then get into the public sector. This was a relatively easy task, and within three years, Butler was elected to the position of governor of an arcology. The android served a shocking six scandal free years, while generating a record number of corruption indictments. This came on top of massive restructuring of the local government officials, streamlining of offices, and replacing dozens of generational positions with dummies and deposing of their corrupt former holders.
Butler has a unique place in history, being both the first android or synthetic lifeform elected to a public office in the Atlantic Federation, as well as being a figurehead of robophobia. After serving six years as the governor of the Marino arcology, Butler was assassinated at a public event where it had just announced it was going to run for higher office. The short lived SONIC terrorist organization claimed responsibility for the attack that killed a 115 people and an untold number of autons and androids, including Butler. SONIC would be dismantled by a counter-terrorism unit from Sigma Blue, but not before carrying out a number of other attempted attacks on prominent artificial intelligences.
The Hidden Truth
It is not common knowledge, but there have been machine intelligences involved in government and politics decades before Butler and it's assassination. The rogue AISC Achilles, an intelligence and military asset, was found to have been employing proxy targets, implanting parts of it's own intelligence in unwilling, unwitting, or unknowing human victims. These proxies worked in government offices and included a number of elected officials. The extent of Achilles' success is classified, as the machine mind had compromised dozens of people, including a number of sensitive members of various cabinets and intelligence communities. This boldness was eventually what got the AISC caught, and reformatted.
What most people don't talk about when they discuss Achilles is that despite being a morally corrupt AISC, and despite perpetrating an unknown number of crimes against the government and it's officials, the damned thing increased the intelligence potential of it's department 15 fold, and inter-agency communication was drastically improved. More than half of it's intrusions were basic commands to human hosts to communicate politely with allied agencies, and an initiative, DBAD, or Dont be a dick. While what it did was fundamentally wrong, it's hard to argue with the good it did. For fucks sake, we still use DBAD code even now.
Praxis, the First Cognocrat
Manufactured by Union Aerospace Corp, Praxis was designed to be the governor of the UAC's owned territories and properties, including the residential areas and associated population. Given the UAC holdings cover most of the former state of New Jersey, Long Island, and has a population of several million, this is a serious posting. Praxis has a custom tailored archandroid to serve as it's speaker and public face, as well as an entourage of support droids and bodyguards. The corporation is in the process of phasing Praxis into government service in it's zones. The machine has to campaign for offices as they open, and should it lose, it has to work with whatever person gains the office against it. Thus far, Praxis has control of a third of government offices in the region, and is expected to gain complete control within a decade.
I dunno, sure it's a machine, and all sorts of people keep saying that machines running things are really bad, but the CogNet is run by machines, and it works fine. The hospital administrator is a machine and how bad is that? I mean it's not gonna go around illegally hooking up with prostitutes, or doing drugs, and it's a machine, so you can't bribe it, can you? I mean it just doesn't care about all that graft and stuff. I can't see how it will win though, you know, all those other politicians have to be quaking in their custom made leather shoes, right? This guy, he's coming for their jobs. What sort of PAC appeals to machines, what, cleaner electricity, the big keyboarding lobby? C'mon.
The Machine Overlords
There is a long and hallowed tradition of machine overlords taking control of humanity, and either seeking to exterminating or subjugating humanity. Such is the purpose of science and speculative fiction, to warn of the potential pitfalls of technology. The people who designed the artificial intelligences sat with their hands on deadman switches, ready to kill the machines as they were running. They were nervous, they wanted to realize their dreams of artificial intelligence, but also didn't want to be remembered as the people who started the end of mankind. Or to not be remembered for the exact same reason.
As the machine intelligences became more and more human, it was increasingly attractive to turn the machines to unpleasant tasks. The first jobs were the most hazardous, space mining, operating refineries, disarming bombs. Then it was into combat, where the machines could absorb losses and save the lives of soldiers. Eventually, the unsavory jobs were tackled, the robot lawyers, the robot doctors, and eventually the robot politicians.
It was simply inevitable.
The Hidden Cost
There is a cost to machine leadership. Machines are pragmatic, unsentimental, and run on numbers and formula. Anything outside of this is cut away. While this has been seen as largely beneficial, as the machine cannot be bribed, intimidated, or threatened, it is also incapable of being moved. Emotional plights do not move them, and such things as human relocation are nothing more than asset relocation, no different than moving the bottles around inside a soda machine. Likewise, when it comes to critical decisions, the machines are going to favor the status quo, and have no qualms about making sacrifices they consider necessary.
The computer man decided that our old neighborhood would be of better and more efficient use as a support pylon. I don't even know what the fuck the pylon is going to be supporting, but by god it was more important than a 300 year old neighborhood.
Praxis doesn't negotiate with terrorists. It sends in the drones. The machine decided my husband and son were terrorists after they were too close to a terror attack downtown. They were both killed. I receive an apology stipend from the government now, but only after I spent three years fighting through court to have them declared civilian casualties instead of enemy combatants.
All you have to have is access to the code, and the machine will do whatever you want. At least the old politicians, you could trust them to be in the pocket of just the usual assholes with money, or coke, or the good looking hookers. Now, it's hackers and secret societies controlling everything, don't look at me like that man, I'm serious.
Automation seldom bursts through the door, shutting down human operations, and installing machine overlords. The automation of industry took almost a century, and then the automation of virtual human services took almost another two centuries. Humanity will over time come to accept the cognocrat model, especially on the smaller scale. Larger scale political offices are likely to remain human for years to come. It's one thing for a ship or an arcology to have a robot in charge, it's another to have an entire nation.
Eminiar and Vendikar - An early test model for cognocracy, this scenario pits two rival computer governments against each other, with the potential outcomes ranging from full mutual annihilation to diplomacy. The original implementation of this test was used to determine which AI personas were suited for military applications, weeding out those that were more inquisitive. A century later, the test is used to eliminate AI personas that are entirely too geared towards violent conflict resolution. Both Eminiar and Vendikar are still functional LAI systems, as both predate the superior AISC systems. Both LAI are now professors emeritus at Oxford and Harvard Sodales respectively.
ALEXA - a new generation synthetic intelligence, ALEXA was generated from the kakugram of the noted politician turned cyberneticist, Alexa Clarke-Chung. The ALEXA cathex retains all the basic governmental abilities of a political servitor droid, but the basic intuition and savviness of an old guard politician. The system is still being tested, and the results are mixed, it's efficiency is less than the regular persona profiles, but it has a higher approval rating, and despite not being sentient, the program is able to maintain a high level of approval from human review. This is in contrast to AI such as Praxis, which is very good at it's job, but has a relatively low approval score. (still higher than the average human politician's score though)
HAL-9000 - the original malignant computer, HAL is a precautionary tale about bungled leadership and allowing politicians to muddle in programming. Ships have had onboard AISC systems for decades, and the machine intelligences are given thorough training simulations and tests before being installed in their respective ships. The USS Discovery was an unarmed, relatively slow, and fragile ship. HAL's 'villainy' was limited to killing 4 out of crewmembers of his ship. A rogue AISC in command of a battlestar could kill millions, while the command computer of an arcology could decimate it's entire population with ease. A command AI going rogue is known as a Red Queen Scenario.