Fifty-five years ago, the world met Hal. Not me, but Hal 9000 in the film 2001: A Space Odyssey. It was, perhaps, the first peek most of us had at artificial intelligence (AI) and computers that could think and make decisions and communicate what they wanted to tell us and persuade us to do what they wanted us to do.
Well, bad actors leveraging powerful computing technology are busy communicating (and curating) what they want us to learn in a way that will make each of us receptive to internalizing their messages. And yes, of course, let’s give a nod to all the wonderful things AI can and will provide, and let’s acknowledge that the age of the supercomputer intersecting with our lives is upon us.
However, let’s not lose sight of the dangers these learning machines represent. Machine learning is that part of AI that gives supercomputers the ability to learn from massive volumes of data without explicitly being programmed. They can and do learn how to construct a message to which you, specifically, are apt to be receptive, even if the news is deceitful and untruthful. We can all celebrate the great evolving potential of AI, but we would be fools to be dismissive of the potential for harm.
We are being bombarded by Bots, short for robots which are simply messages from supercomputer artificial intelligence programs that, in effect, have created non-human agents (Bots) for human puppeteers who may be here in America or anywhere else in this interconnected world.
Bots have one purpose. They are designed to simulate human activity, that is, to fool the receivers of the messages into believing that a human being sent it to them. Be concerned. Be very concerned.
And now we have generative AI that can understand and communicate with us in a way that is exceedingly close to conversing as we humans talk.
It is known that AI (compliments of the Kremlin) was hyper-active in the 2016 U.S. presidential election in the form of Russian Bots. Specifically, more than 36,000 Russian-Bot accounts tweeted to American home computers “news or commentary” about the U.S. election between Sept. 1 and Nov. 15, 2016. These were online accounts masquerading as humans, but they were actually messages contrived by Russian-developed artificial intelligence.
These phony online “persons” created and sent an estimated 288 million impressions of Russian Bot tweets. And they were just getting started. In the first three months of 2019 alone, Facebook took down an estimated 2.2 billion utterly fake accounts, supercomputer-generated Bots pretending they were people like you and me communicating to their Facebook universe of friends and family to be read, “liked,” and often “shared.”
This new and insidious interference in our political life and national elections is an outrage. But even more sobering is the reality that Bot speech, that is non-human, carefully crafted messaging designed to appear as though it comes to us from human beings, from fellow citizens like you and me, may very well be speech protected by the First Amendment.
The First Amendment protects speech (written or spoken) and not so much the speaker of speech. It protects the message more than the messenger. It protects the listener of speech (our right to hear or read what a messenger is saying or writing) just as much as the “speaker” or writer of speech.
There is a consensus among many lawyers that “speech” by non-human, computer-generated robots is as protected as speech produced by you and me.
John Frank Weaver, an attorney specializing in artificial intelligence, contends that speech created and distributed by Bots should have all the constitutional protections as actual speech by real humans. Nothing in the First Amendment stipulates that written or spoken speech is limited to human beings. (Who knew?— an American founder might ask).
Weaver posits that our right to free speech should include speech contrived by AI, robots, and Twitterbots. And Weaver is not alone. These champions of robot-generated free speech suggest that it is up to us, approximately 330 million Americans (or at least those of us old enough to vote), to differentiate between AI Bots and human-generated content. Be concerned. Be very concerned.
Some believe the Supreme Court’s Citizens United decision which conferred on corporations the same free-speech protections humans have, also grants the same protections to artificially conceived and generated messages from supercomputers. Of course, human beings do write and speak for corporations.
On the other hand, troll farms generate what I’ll call Bot-beings. They “pretend” to be human. They can and do develop messages on behalf of governments or other special interests, politicians and wannabe politicians, or others who may simply want to disrupt civil society. They then send or share these machine-generated messages as though thinking people had shared them.
They aren’t really engaging in free speech. They are manufacturing speech as well as the purveyors of speech, often to achieve political objectives, and do not necessarily represent the thinking, concern, or passion of any human being. Be concerned. Be very concerned.
California has passed a Bot law that makes it unlawful to use a Bot to communicate or interact with another person online in California with the intent to mislead that person about its artificial identity, thus deceiving that person about the content of the communication. The Bot must be identified as a Bot. Good for California!
The idea that AI speech is just babble generated by humans is disingenuous. Pew Research warned that two-thirds of all the headlines tweeted on Twitter were then shared by Bots. When artificial intelligence can produce content tailored to specific people based on what it has learned about their likes and dislikes, and what comforts them and discomforts them, we are, indeed, in a strange new world. Welcome to the future. Be concerned. Be very concerned.
All comments regarding these essays, whether they express agreement, disagreement, or an alternate view, are appreciated and welcome. Comments that do not pertain to the subject of the essay or which are ad hominem references to other commenters are not acceptable and will be deleted.
Invite friends, family, and colleagues to receive “Of Thee I Sing 1776” online commentaries. Simply copy, paste, and email them this link— www.oftheeising1776.substack.com/subscribe –and they can begin receiving these weekly essays every Sunday morning.