After the US election, many people feel we’re on the verge of an apocalypse. What might not be obvious is that it’s an artificial intelligence apocalypse, like the ones in The Terminator and The Matrix. The fact is the machines have taken over, enslaved us, and may now destroy us.
No, seriously. Let me explain.
Here’s what an AI apocalypse looks like in broad brushstrokes:
1. We carefully program AI with a well-defined goal for the betterment of the human race, give it power, and set it loose.
2. It develops unexpected emergent behaviours, and begins to have impacts unforeseen by its creators.
3. The well-defined goal is achieved. Unfortunately, the unintended consequence is the downfall of humanity.
Let’s compare this to current circumstances.
The Nieman Foundation for Journalism at Harvard reports that nearly half of all Americans get their news from Facebook. The social media site delivers news via its algorithms, which decide what stories you, as an individual user, see, and in what order. You might ask what’s wrong with that, when Facebook’s mission statement is to “give people the power to share and make the world more open and connected”.
‘We’re on the verge of an AI apocalypse, like the one in The Matrix.’ Photograph: Cine Text/Sportsphoto Ltd /Allstar
But the fact is that rather than serving this goal, Facebook’s AIs are servicing a far older and more well-established social goal, which was designed for the betterment of mankind. This is the goal of maximising value in pursuit of economic self-interest.
Neoliberal economics is based on the idea that the pursuit of individual, economic self-interest leads to the creation of greater value, which benefits society and results in the emergent betterment of mankind.
The individual self-interest being pursued in this case is that of Facebook, a legal person who is programmed, within our economic and legal system, with a single mandated goal: the delivery of maximum value to its shareholders. Facebook’s AIs are the technological limbs of that person, and they must ultimately reach out into the world to carry out that goal.
To that end, an election cycle is a goldrush, where these value-maximising AIs can reach out to mine your eyeballs. The more you check your feed, the more ad-based value the AIs can generate for the shareholders.
What no one foresaw was that in order to keep us clicking and sharing, headlines had to work harder to grab our attention, shouting ever more sensational and titillating news stories, millisecond-by-millisecond. The resulting lightspeed feedback loop coupled our fears and hopes for the future to the simple, relentless value system of the AIs and their corporation. It was only a matter of time before humans became collaborators, or as Arron Banks, friend of Nigel Farage and financier of the Leave.eu campaign, stated in the recently published account of its Brexit campaign: “The more outrageous we are, the more attention we’ll get. The more attention we get, the more outrageous we’ll be.”
If the outrageous headlines can get you to take action, and click through, then there are even more AIs out there ready to feed their own revenue loops, with highly shareable, click-baiting stories and memes. They even exploit simple human foibles, with finger slips that activate ad links, and malware that infects our devices to keep other AIs informed about us, so we can be further targeted with material tuned to our particular trigger points. Each step in this massive computational network manipulates you for a click-through, generating a micropayment-driven AI ad placement feeding frenzy.
Of course, it’s not just Facebook involved as an end supplier in this new economic boom. Note that beneath each CNN article, every one of them sold with a tempting headline, there usually appears copious paid content. This is part of an algorithmically driven revenue generation strategy called “smart ad placement” by Turner Broadcasting, the subsidiary of Time-Warner that owns CNN. Turner’s site boasts of how optimising algorithms place ads “in the style of CNN editorial content, with the ability to be placed directly in editorial streams of content and alongside relevant CNN editorial videos and articles”. Even a “pure information” site like Nate Silver’s FiveThirtyEight has been accused of warping its statistical algorithms to make the US election seem closer than it might have done, and its graphs therefore more tempting for obsessive checking. Silver has defended himself against these accusations, tweeting: “The reason we adjust polls for the national trend is because **that’s what works best emperically** [sic]. It’s not a subjective assumption.” But there is no doubt that his site is now a for-profit poll predictor, owned by ESPN.
Being able to discern what is true and what is fake is not a goal that AIs can achieve easily. As Zuckerberg himself has said, the problem is “complex, both technically and philosophically”. This is because the “truth” is a complex human concept that is very hard to programme. Witness that the website Politifact, dedicated to fact-checking US politics, has to place “truth” into six categories (only half of which are definitive) and provide a detailed, carefully written analysis to explain the nuances of the verity of each statement it examines. Human truths and falsehoods are not simple 0s or 1s.
Furthermore, the war on fake news misses the point. Regardless of whether AIs present “true” or “false” news, they will still be working towards their primary directive: the maximisation of value for the corporation.
The unforeseen outcomes of neoliberal values coupled to highly efficient AIs are ultimately emergent effects on our social fabric. AIs “personalise” newsfeeds for you, but this personalisation isn’t really for your benefit, it’s to place you in a tighter demographic category, making you a more targeted value proposition to sell to advertisers. The effect of this is that we have all been herded into digital echo chambers. And, with our emotions engaged by clickbaiting, profit-generating, dubious headlines, we’ve aided the process through enraged blocking and unfriending of anyone whose opinions differ from our own.
The truth is that Americans are no more simply Republican and Democrat than news is simply true and false, and our segregation into crisp, sale-able demographic packages serves only to reduce a complex human picture made up of difficult social and political issues. But the vast, unsympathetic network of AIs is not concerned with this complexity, it is working towards its programmed goal: to maximise value. And it is doing a great job. Through the election cycle Facebook’s stock rose 924% faster than the US stock market index, the S&P 500.
The unintended consequence of this massive AI success may now be our own downfall: an apocalypse, heralded not by trumpets, but by Trump/Pence. Leaders who are more than willing to collaborate with our new robot overlords, communicating with us through them via a series of ever more inflammatory tweets. A world where we are more closed as individuals, and where we are less connected as communities, the ironic opposite of the world envisioned in Facebook’s mission statement.
I’d write more, but right now I’ve got to go and check my Facebook feed, CNN, and any other news site I can find to see what outrageous things our president-elect may have done in the past few minutes. Or perhaps find something to share that might help impede his progress. Perhaps a funny Joe Biden meme.
Contact author
After the US election, many people feel we’re on the verge of an apocalypse. What might not be obvious is that it’s an artificial intelligence apocalypse, like the ones in The Terminator and The Matrix. The fact is the machines have taken over, enslaved us, and may now destroy us.
No, seriously. Let me explain.
Here’s what an AI apocalypse looks like in broad brushstrokes:
1. We carefully program AI with a well-defined goal for the betterment of the human race, give it power, and set it loose.
2. It develops unexpected emergent behaviours, and begins to have impacts unforeseen by its creators.
3. The well-defined goal is achieved. Unfortunately, the unintended consequence is the downfall of humanity.
Let’s compare this to current circumstances.
The Nieman Foundation for Journalism at Harvard reports that nearly half of all Americans get their news from Facebook. The social media site delivers news via its algorithms, which decide what stories you, as an individual user, see, and in what order. You might ask what’s wrong with that, when Facebook’s mission statement is to “give people the power to share and make the world more open and connected”.
But the fact is that rather than serving this goal, Facebook’s AIs are servicing a far older and more well-established social goal, which was designed for the betterment of mankind. This is the goal of maximising value in pursuit of economic self-interest.
Neoliberal economics is based on the idea that the pursuit of individual, economic self-interest leads to the creation of greater value, which benefits society and results in the emergent betterment of mankind.
The individual self-interest being pursued in this case is that of Facebook, a legal person who is programmed, within our economic and legal system, with a single mandated goal: the delivery of maximum value to its shareholders. Facebook’s AIs are the technological limbs of that person, and they must ultimately reach out into the world to carry out that goal.
To that end, an election cycle is a goldrush, where these value-maximising AIs can reach out to mine your eyeballs. The more you check your feed, the more ad-based value the AIs can generate for the shareholders.
What no one foresaw was that in order to keep us clicking and sharing, headlines had to work harder to grab our attention, shouting ever more sensational and titillating news stories, millisecond-by-millisecond. The resulting lightspeed feedback loop coupled our fears and hopes for the future to the simple, relentless value system of the AIs and their corporation. It was only a matter of time before humans became collaborators, or as Arron Banks, friend of Nigel Farage and financier of the Leave.eu campaign, stated in the recently published account of its Brexit campaign: “The more outrageous we are, the more attention we’ll get. The more attention we get, the more outrageous we’ll be.”
If the outrageous headlines can get you to take action, and click through, then there are even more AIs out there ready to feed their own revenue loops, with highly shareable, click-baiting stories and memes. They even exploit simple human foibles, with finger slips that activate ad links, and malware that infects our devices to keep other AIs informed about us, so we can be further targeted with material tuned to our particular trigger points. Each step in this massive computational network manipulates you for a click-through, generating a micropayment-driven AI ad placement feeding frenzy.
Of course, it’s not just Facebook involved as an end supplier in this new economic boom. Note that beneath each CNN article, every one of them sold with a tempting headline, there usually appears copious paid content. This is part of an algorithmically driven revenue generation strategy called “smart ad placement” by Turner Broadcasting, the subsidiary of Time-Warner that owns CNN. Turner’s site boasts of how optimising algorithms place ads “in the style of CNN editorial content, with the ability to be placed directly in editorial streams of content and alongside relevant CNN editorial videos and articles”. Even a “pure information” site like Nate Silver’s FiveThirtyEight has been accused of warping its statistical algorithms to make the US election seem closer than it might have done, and its graphs therefore more tempting for obsessive checking. Silver has defended himself against these accusations, tweeting: “The reason we adjust polls for the national trend is because **that’s what works best emperically** [sic]. It’s not a subjective assumption.” But there is no doubt that his site is now a for-profit poll predictor, owned by ESPN.
Some are concerned that the problem that has developed is so-called “fake news”. While Wikipedia defines “fake news” as a synonym for news satire, in the post-election shock it has now become an umbrella term for the uncontrolled explosion of misleading information dispersed on social media during the election. Mark Zuckerberg has pledged to make Facebook impede “fake” news. Perhaps he’ll use real human editors, but in the recent past, his company has eliminated real people in this role, in favour of algorithms, to ensure less bias in their presentation of news.
Being able to discern what is true and what is fake is not a goal that AIs can achieve easily. As Zuckerberg himself has said, the problem is “complex, both technically and philosophically”. This is because the “truth” is a complex human concept that is very hard to programme. Witness that the website Politifact, dedicated to fact-checking US politics, has to place “truth” into six categories (only half of which are definitive) and provide a detailed, carefully written analysis to explain the nuances of the verity of each statement it examines. Human truths and falsehoods are not simple 0s or 1s.
Furthermore, the war on fake news misses the point. Regardless of whether AIs present “true” or “false” news, they will still be working towards their primary directive: the maximisation of value for the corporation.
The unforeseen outcomes of neoliberal values coupled to highly efficient AIs are ultimately emergent effects on our social fabric. AIs “personalise” newsfeeds for you, but this personalisation isn’t really for your benefit, it’s to place you in a tighter demographic category, making you a more targeted value proposition to sell to advertisers. The effect of this is that we have all been herded into digital echo chambers. And, with our emotions engaged by clickbaiting, profit-generating, dubious headlines, we’ve aided the process through enraged blocking and unfriending of anyone whose opinions differ from our own.
The truth is that Americans are no more simply Republican and Democrat than news is simply true and false, and our segregation into crisp, sale-able demographic packages serves only to reduce a complex human picture made up of difficult social and political issues. But the vast, unsympathetic network of AIs is not concerned with this complexity, it is working towards its programmed goal: to maximise value. And it is doing a great job. Through the election cycle Facebook’s stock rose 924% faster than the US stock market index, the S&P 500.
The unintended consequence of this massive AI success may now be our own downfall: an apocalypse, heralded not by trumpets, but by Trump/Pence. Leaders who are more than willing to collaborate with our new robot overlords, communicating with us through them via a series of ever more inflammatory tweets. A world where we are more closed as individuals, and where we are less connected as communities, the ironic opposite of the world envisioned in Facebook’s mission statement.
I’d write more, but right now I’ve got to go and check my Facebook feed, CNN, and any other news site I can find to see what outrageous things our president-elect may have done in the past few minutes. Or perhaps find something to share that might help impede his progress. Perhaps a funny Joe Biden meme.
Somewhere an AI would smile. If it had a face.