Are we already on our way to being ruled by machines? Governmental artificial intelligence (AI) “advisors” are already here; how long before we (or they) decide they no longer need the human factor to run the show?
Plenty of debates about the role of AI in making governmental decisions are sure to materialize soon in the face of recent technological developments, particularly in response to recent reporting that Romanian Prime Minister Nicolae Ciucă has added an artificial intelligence advisor to his Cabinet.
As Politico reported, some of Ciucă’s ministers were taken aback when the Romanian leader this week presented the Cabinet’s newest member, an AI named “Ion.”
Ion holds the title of “honorary advisor,” and is touted as the first of its kind. Its role within the government is to analyze social media networks in order to inform policymakers “in real time of Romanians’ proposals and wishes,” the prime minister said on Wednesday.
Ion has a mirror-like chrome screen that can display words and even a digital face.
Because Ion will ostensibly allow the Romanian government to have a clear understanding of the moods and thoughts of the people, Ciucă said Romanians should consider it their duty to use the online website and Twitter account to give their opinion on political issues and current events.
“I have the conviction that the use of AI should not be an option but an obligation to make better informed decisions,” said Ciucă.
Romanians can also provide Ion with their opinions at some in-person locations that are soon to be announced.
Kris Shrishak, a technology fellow at the Irish Council for Civil Liberties, spoke with Politico and recommended that government authorities exercise caution in implementing these kinds of technologies.
“This should be explained to the public,” Shrishak said, asserting that this raises questions about how Ion will go about determining what feedback should be considered a priority.
Like anything digital, there is always the possibility that this “government” advisor can be hacked or otherwise manipulated in order to make the government believe that any given issue or policy has more support or opposition than it really does.
It doesn’t even have to be hacking. What if one political faction organizes its adherents to express the same view to Ion in waves, giving the impression that they are the majority, when in reality they’re just the loudest voice in the room?
And what if Ion develops political preferences of its own and, just like a human Cabinet minister, discreetly inserts its own biases into what is supposed to be cold, objective analysis?
We have already seen cases of AI developing political views. In one shocking example from 2016, Microsoft created a “smart” AI Twitter account named TayTweets that was designed to learn from user communication.
The idea sounded innocent enough — except online trolls quickly turned TayTweets into a Nazi within a matter of hours. Microsoft ended up deleting the account.
We now have AI lawyers helping defendants get out of parking tickets and AI writers producing content for top websites.
As The New American has reported, AI is increasingly being applied to military purposes. In one notable example, an AI-powered drone aircraft bested a human-controlled aircraft in a dogfight organized by Chinese military researchers.
The dogfight involved two small, unmanned aircraft. One had an AI pilot on board, while the other was remotely controlled by a human pilot on the ground. The project team’s paper based on the dogfight concluded that the “era of air combat in which artificial intelligence will be the king is already on the horizon.”
While artificial intelligence is still in its infant stage, the leaps it has already made in a relatively short window of time indicate that the rate of progress will only accelerate in coming years. While it’s easy to shrug off concerns at this early phase, society must come up with answers about the hard questions posed by AI before the genie is completely out of the bottle.
Take this latest government-advising AI, for instance. Besides the potential for abuse and corruption detailed above, there is the wider implication of what policy-formulating artificial intelligence means for human government.
Gradually, more and more functions of government are likely to be placed in the hands of artificial intelligence. It’s likely that many will see AI as being more efficient and less partisan than human beings, leading a portion of the public to demand that software replace “corrupt” human politicians.
Eventually, the public may toss out elections and human government altogether and simply have an AI govern the masses.
And even if it doesn’t get to that point, there is the possibility that AI may reign, if not de jure, at least de facto, through the creation of widely used software that analyzes candidates and “tells” people whom to vote for.
