Archon Fung and Lawrence Lessig write in Governing how artificial intelligence could manipulate voters in 2024 and produce a President with no genuine democratic mandate:
Imagine that soon, political technologists develop a machine called Clogger – a political campaign in a black box. Clogger relentlessly pursues just one objective: to maximize the chances that its candidate – the campaign that buys the services of Clogger Inc. – prevails in an election.
…First, the messages that Clogger sends may or may not be political in content. The machine’s only goal is to maximize vote share, and it would likely devise strategies for achieving this goal that no human campaigner would have thought of.
One possibility is sending likely opponent voters information about nonpolitical passions that they have in sports or entertainment to bury the political messaging they receive. Another possibility is sending off-putting messages – for example incontinence advertisements – timed to coincide with opponents’ messaging. And another is manipulating voters’ social media friend groups to give the sense that their social circles support its candidate.
Second, Clogger has no regard for truth. Indeed, it has no way of knowing what is true or false. Language model “hallucinations” are not a problem for this machine because its objective is to change your vote, not to provide accurate information.
Third, because it is a black box type of artificial intelligence, people would have no way to know what strategies it uses.
…Political scientists and pundits would have much to say about why one or the other AI prevailed, but likely no one would really know. The president will have been elected not because his or her policy proposals or political ideas persuaded more Americans, but because he or she had the more effective AI. The content that won the day would have come from an AI focused solely on victory, with no political ideas of its own, rather than from candidates or parties.
In this very important sense, a machine would have won the election rather than a person. The election would no longer be democratic, even though all of the ordinary activities of democracy – the speeches, the ads, the messages, the voting and the counting of votes – will have occurred [Archon Fung and Lawrence Lessig, “How AI Could Take Over Elections and Undermine Democracy,” Governing, 2023.06.11].
The key to avoiding this ugly scenario of automated mind manipulation may be to deny First Amendment protection to such automated speech. Free speech revolves around the idea that individuals use their free will to compose and express their own ideas. A candidate or PAC boss or blogger who uses artificial intelligence to generate text to manipulate an election is not expressing personally formulated thoughts. The person hitting the Go button on an AI manipulation machine has no idea what the machine will say. That is not free speech; that is noise without will, without intent, without control.
Fung and Lessig note that my denial of First Amendment rights to textbots is far from settled philosophy or law. But if we are to prevent such a grave reduction of democracy to machine-language tricks, we must find a way to distinguish genuine civic discourse from the algorithmic manipulation of human will.
Hi Cory!
Are you stating that the human mind has free will in some manner that computers do not?
That human decisions are not wholly the naturalistic result of prior naturalistic causes?
That humans have supernatural consciousnesses which inject causes into an otherwise clockwork universe?
Last I remember discussing this with you, you thought our brains were hardwired by evolution. No real choice. No ought or should statements. We can’t control what we do any more than a carrot or a computer can.
If we have genuine free will, please tell me how that fits within your atheology. In chapter 3 of Miracles it was argued that genuine free will requires the supernatural…
Kind regards,
David
David..I don’t reject your notion of a Super Natural. I’m puzzled by what came before the Big Bang. I’m comfortable with the notion that the Super Natural force put the Universe in motion, said “It is Good”, and moved on to other things.
Cambridge Analytica’s BREXIT campaign, even the 2016 and 2020 trumpian campaigns showed that the voters, too many voters, are subject to easy manipulation.
Those campaigns also showed that the horse-race theatrics and ‘gotta be first’ of mass media is incapable of serving the public square as a viable Fourth Estate.
Those campaigns also showed that congress is incapable and unwilling to regulate campaigns.
We must be concerned.
From The P. Aitch Production Group (first one’s free)
– One strategy to maximize vote share without any limits that hasn’t been utilized by human campaigners is to use artificial intelligence and machine learning algorithms to analyze big data and voter behavior patterns. By studying and analyzing data that is available on social media, voter demographics, and past voting records, AI can enable campaigns to provide targeted messages to the right set of voters at the ideal moment.
Moreover, tapping into emerging technologies like virtual assistants, chatbots, and online platforms that encourage voter participation can also help campaigns engage with a broader audience. AI-powered chatbots and virtual assistants can offer conversational responses to voters’ questions and concerns, provide information on candidates, and assist with voter registration and early voting.
Furthermore, AI-driven sentiment analysis and predictive analytics can help gauge the success of the campaign, perceived popularity, and potential impact on voting behavior. By leveraging these insights, campaigns can fine-tune messaging, allocate resources more efficiently, and improve their chances of reaching and persuading the electorate.
Overall, by leveraging AI and emerging technologies, campaigns can create hyper-personalized and timely messages targeting specific demographics, boost voter participation, and ultimately win (or retain) majority share in a particular race.
Cory writes: “Free speech revolves around the idea that individuals use their free will to compose and express their own ideas.” And therein lies the rub. Free speech is also a recognition given to corporations; and a Supreme Court that is willing to imbue corporations with individual right status certainly will extend that to all forms of speech that entity can produce. Money, chatbots: smoke ’em if you got ’em.
I also reject the label “AI” for these predictive algorithm language producers. Nothing is being created — creation is the test of intelligence.
I don’t think we need meddling by anything that cannot take the stand to be cross-examined and asked why or how they chose to do something. Since even the writers of these programs have no clue once they push GO, these out of control functions gotta get the hammer.
The original AI is the church. Then, came the corporation. Both had a goal and few perimeters and took over the world.
AI does in fact create.
That’s why it’s named artificial “intelligence”.
– Each person’s AI starts out with the same algorithms but every time you give your “insert the name you’ve given the little rascal here” a task, it learns from what you ask it to do, and the next time will use your previous commands, inflections, and annotations to help create the next task you give it.
– In time your l’il rascal becomes like a “mini me”. It’s pretty cool but when you live in a state that’s last in innovation skills all you’ll get in SD is negative feedback and scary ass fear, from your majority of citizenry.
– *Don’t Worry SD Dems. No algorithm, even one designed by the greatest geniuses on the planet, can lie and be more deceitful, deceptive, and misdirecting than Ian Fury, Kristi Noem, Pat Powers, Lee Schoenbeck et al.
– The lesson of AI is simple. It’s the same lesson y’all are already following. Don’t believe anything you read and only half of what you see.
– It’s the group that believes whatever BS they agree with who vote for Don “The Con” Trump, believe what he says, and repeat it to other simple minds.
All Mammal, I am with you on your church/corporation idea. However, I would argue BOTH were an artificial construct to hide the man behind the curtain really pushing the agenda/saying the words — not artificial intelligence any more than Oz’s projection was.
Cory writes: Second, Clogger has no regard for truth. Sounds like present day politicians.
P. Aich, here is why I disagree on if AI (as it is now) creates: all these chatbots are is a predictive algorithm to form sentences. The bot does not begin with an idea — much less an original idea (which is the gist of how I would define “create.”) — instead to charts a pathway of choosing the most popular next word, word-after-word, until it gets to the predetermined length. It’s syntax is nothing more than a statistical analysis of the available records on the given topic. It is plagiarism; it is theft of other’s thoughts and creations.
Adam Conniver has been insightful on this issue and has several excellent YouTube dissertations on this topic.
David, I reserve the right to defend our prerogatives over the destructive impacts of our technology.
Were I to adopt a religious perspective, I might note that we have no rights before our angry Creator. No one took Yahweh to court for wiping almost everyone out in the Flood. No one can take us to court for banning our robots from manipulating public perception.
I won’t stand in the way of supernaturalists who want to contend that our magical free will distinguishes us from machines and allows us to outlaw automated election interference. I welcome such assistance in protecting democracy.
I’ve said in my previous conversations (and will maintain now) that I cannot explain free will outside of mechanistic processes and that decisions resulting from mechanistic processes aren’t really free will. Yet even if I am correct in that understanding, mechanistic processes still produce within me (and, it appears, in all the other gibbering hominids around me) the sense that I have free will, that I like free will, and that I don’t want widgets created, wound up, and turned loose without control by my fellow hominids to infringe on my free will. Who am I to argue with mechanistic processes that drive me to resist manipulation by mechanistic processes?
O – I disagree.
– Once the data has been collected, AI uses various algorithms to analyze it and identify patterns, connections, and relationships that humans are not able to see. From this analysis, AI can generate new ideas and insights that can be used to develop new products, improve existing ones, or solve complex problems.
For example, an AI system may analyze customer reviews of a particular product and identify common complaints or suggestions for improvement. It can then use this information to generate new ideas for product features or design changes that can address these issues and improve customer satisfaction.
Overall, AI has the ability to generate new ideas by leveraging its ability to analyze and interpret large amounts of data, identify patterns and insights, and develop innovative solutions based on those insights.
Hi Cory!
We certainly agree in not wanting to cede our government or our vote to automatons. We also agree in having a strong subjective sensation of free will. The question is, where does this sensation come from?
What is the most real thing we know? Where do we start our philosophy? From what do we derive the rest of reality?
To me, the beginning is the sensation of free will. There are many things that I could do or think or say, and from that, I choose one and do it. Since matter is not free, there must therefore be something else in me that chooses. The ancients may have called it the soul, but since that’s a pretty nebulous term, I prefer to call it ooda. Something observes, orients, decides, and acts… and matter alone isn’t capable.
If ooda is real, then so is morality, because morality is about aiming ooda toward individual and collective human flourishing. If I want to slake my thirst, I shouldn’t reach for a bag of sand. If I want to be invited to play games with my friends, I shouldn’t cheat.
I think morality is the point of your blog. Correct me if I’m wrong, but I believe that the reason you type a thousand or two words to the public each day is because you think limits on abortion and petitioning stifle human flourishing, and you think that when elected officials lie, spin, and dodge hard questions, it’s bad for the electors. You wish to guide our ooda accordingly.
However, ideological possession is an omnipresent danger, and the same way that Einstein was duped into thinking that the universe was static, other intelligent people can be duped into denying the most-real thing. That’s where a reality check comes in. Matter is not the most-real thing. Ooda is. You can’t get ooda from matter, but if you start with ooda, you can know and understand matter.
To me, the explanation that free will is the result of mechanistic processes, is like saying that Achilles can’t catch the tortoise. Yes, I understand that there is a seemingly valid line of reasoning to back it up, but the conclusion is obviously wrong.
Kind regards,
David
Free will gives we humans the ability to act and make choices freely, without being determined by external forces or predetermined rules.
On the other hand, artificial intelligence is created and programmed by humans to perform specific tasks based on pre-set rules. While AI can learn from experience and improve its performance, it does not have the capacity for free will or independent decisions.
One might argue that free will makes humans superior to artificial intelligence because it gives us the ability to make decisions that are not pre-programmed or limited by a set of predetermined rules. However, it is important to note that AI can provide significant benefits and advancements to society through its ability to process vast amounts of data and perform complex tasks efficiently and accurately.
Artificial intelligence is only as deviant as the deviants who command it.