Albania is one of those countries you hardly hear about in the news. Because of this lack of visibility, very little of its cultural exports reaches Africa. This South-East European country recently made the news over the appointment of Deilla, an AI avatar, as minister whose task is to ensure that public tenders are corruption-free. As if people’s fear about artificial intelligence displacing them from their jobs is not valid enough, now we are talking about the inclusion of AI in government. It couldn’t get more misanthropic than that.
Already, AI-slop has metastasised across the internet. YouTube is particularly annoying to watch these days as you constantly get interrupted with AI-generated ads about the signs of procrastination or of a middle-aged woman in military fatigues talking about how women can exercise to look good or something. Corporations like Coca-Cola unashamedly now use AI to create ads. And what about the endless AI-generated videos you chance upon as you doomscroll through astoundingly absurd plotlines of people doing the most insane things? I saw one a few days ago of a man building a jumbo jet entirely out of wood. When you give people a godlike technology, they will definitely misuse it, especially when there is the possibility of monetisation. AI-slop is proof of that.
Yet, in the midst of this AI mania, Albania decided to take it a notch higher with the appointment of Deilla. This has been called Avatar Democracy. The logic is simple: as humans, we are susceptible to moral failings. We are self-interested beings. According to Public Choice theory, we will always exploit every avenue to pursue this self-interest, even in government. We should expect politicians, public servants, and bureaucrats to be more selfish than they are altruistic. Whatever solutions they propose are ultimately geared towards their vested interests. And that’s why new policies are created as solutions to older policies. The more things change, the more things remain the same, as the saying goes. In this case, what remains the same is the corruption status quo. And precisely because of this, society fails, especially in countries with weak institutions.
Taking this fatalistic approach, the appeal of Deilla becomes understandable. Deilla, on the one hand, is an indictment of human nature. But on the other hand, it is a solution to it. If the human condition is irredeemably hopeless, then we need a deus ex machina. We know this messiah can’t be made of flesh, bones, and blood. But we could care less if it is made of codes and bits, like Deilla. If programmed without bias, AI could deliver an efficient system in public tenders by getting rid of red tape, cronyism, and fraud. The entire system, in theory, will become transparent and meritocratic. Gone would be those days you needed to grease people’s palms to get a contract. Truly, there is something misanthropic about outsourcing an important government task to artificial intelligence. But from a consequentialist perspective, if Deilla works, Albania would be better off, and that would be a win for everyone.
In many ways, Nigeria is like Albania. Corruption is a huge part of our existence. Administration after administration, policy after policy, the problem persists. In Nigeria, the world is your oyster as long as you have money and power. The rules do not apply to you anymore. Your proximity to power comes with perks you can parlay into juicy contracts or appointments. And so, we live in a country where the corruption industrial complex has made the system unmeritocratic. Ours is a country of quintessential extractive institutions; a country ruled by cartels. Everything from government procurement to public tenders reeks of this cartelisation. With weak institutions, it’s hard to see how this problem could go away.
What if Nigeria goes the way of Albania in fixing this problem? We complain a lot about how unfair the process is. But what if artificial intelligence could fix that by making it a transparent, meritocratic, and accountable process? With AI, citizens can request information about government procurement at the speed of light, as opposed to the current system, where everything gets lost in a bureaucratic maze. With AI, consideration is not given to religion, ethnicity, state of origin, or other nonsensical criteria. Everyone gets assessed on the merit of their application. Wouldn’t Nigeria be better off?
At some point, a conversation becomes inevitable: when and where do we draw the line? If we adopt artificial intelligence in one area of governance and it becomes a success, the temptation to adopt it in other areas becomes stronger. If the premise is that we want to make governance more efficient and every time we integrate AI it performs miracles, the logical conclusion would be that fewer humans would be needed in governance. And this is where fears of the misanthropic nature of artificial intelligence are justified. Yes, the process becomes efficient, but it comes at the cost of human redundancy. This techno-dystopia undercuts Thomas Hobbes’ theory on the role of government. If in the state of nature life is brutish, lawless, and only the Leviathan (government) can set things in order, well, in Nigeria, we’ve enjoyed very little of that orderliness since independence. If anything, the government has contributed more to disorderliness and lawlessness in society. Resorting to artificial intelligence then casts a sense of irony on the Hobbesian conception of government.
The idea of flawed humans programming AI to be unbiased is in itself ironic. It’s even a bit more complicated given how Nigerians typically view everything through the prism of identity. Hence, open sourcing would play a part in assuaging whatever qualms or fears people may have about the adoption of AI.
The Economist released a report sometime last year on why Africa is largely missing from the AI revolution. While America and China account for a substantial percentage of investments in artificial intelligence, Africa’s contribution has been Lilliputian. Of course, there are structural challenges like poor electricity (data centres consume a lot of energy) and the lack of basic digital infrastructure. Given these constraints, it’s hard to build homegrown AI solutions. A lot of Nigeria’s tech success over the past decade has largely been confined to the fintech and crypto space, so much so that our default understanding of what a startup is is something fintechy in nature. This may be because venture capitalists are biased towards funding fintech ideas. Or because we are hostile to novelty as Nigerians; hence, we like to stick to a winning formula. And maybe it’s a combination of all these factors. The point is that there is a need to diversify innovation in the country. Artificial intelligence is one area demanding our attention and investment.
I’m usually pessimistic when it comes to the AI discourse. My pessimism isn’t necessarily Luddite in nature (though it’s hard to make a case against the erosion of jobs being the logical conclusion of mass AI adoption). My discomfort with AI is largely sociological. Modern tech has changed us in many ways for the better and in many other ways for the worse. But we rarely talk about the downsides. Smartphones and social media are reducing our attention span, reducing our intelligence, making us asocial, amplifying extreme political and religious views, atomising society at a scale we’ve never seen before, and doing so much more. Academics are complaining about how ill-equipped students are these days. Now, if we add AI into the mix, the problem can only get worse. AI turns knowledge into fast food. It’s a cheat code. But if done right, its potential for good is limitless — like if we go the Albanian way by adopting it to make government procurement transparent and meritocratic. But maybe without the avatar.
Guest