- Supernormal stimuli;
- Replacing strong link community relationships with weak link affinity relationships;
- Training people on complicated rather than complex environments; and
- The asymmetry of Human / AI relationships
Each category is too involved to describe here with any detail so I urge you to read the whole article. However, here is an excerpt from category (4):
Imagine if your spouse, your therapist and your priest all entered into a conspiracy with a team of world class con men to control and shape your behaviour. Sound a bit unsettling? Well consider what the Facebook algorithms alone know about you. Every conversation you have — even those that you type out but don’t send — are perceived by the Facebook AI, and then analyzed by technology designed by thousands of researchers schooled in the very cutting edge of psychology and cognitive neuroscience.
Every conversation you have — and every conversation the other 1.4 billion people on the platform have. In one second, the Facebook AI learns more about how people communicate and how they make choices as a result of their communication than an average person will learn in fifty years.This is somewhat of an exaggeration at this point, but won't be in a decade or two. AI is becoming more capable than humans in many areas and, to the extent that AI is "intelligent" at all, it's a completely alien intelligence. While it sometimes seems almost human, there's nothing human about it. For example, when you're having an oh-so-pleasant conversation with Siri or Google's Assistant or one of the many automated phone systems you regularly encounter, the entity with which you're conversing has nothing in common with a human. Nothing at all. It's not even programmed by a human in any sort of conventional sense. It's just an alien entity trained to do various tasks like set your alarm clock when you ask it to do so.
When the trainers train it to do benign and helpful things it will be (mostly) benign and helpful. However, if the trainers, such as those working for social media companies, train it to manipulate social media users via emotional and other human weaknesses in order to increase profits and power, it will do that too - without having any understanding whatsoever of what it's doing. It's not immoral, just amoral. In fact, the trainers themselves may not really know what they're doing. They may just be trying to sell more advertising or whatever.
The problem is that we are changing our behaviors in ways that are simply bad for social cohesion and civilization itself. We'll see how that works out. Even those of us who are a little older we'll get to see how this one works out because I believe much of the impact will be experienced within the next 20 years.
Related: Nvidia DGX-2 is 2 petaflop AI supercomputer for $399,000
5 comments:
Pretty interesting, Bret.
And there is also this:
---
what if the source of this polarization has little do with where people actually fall on the issues, or what people actually believe in? What if people are simply polarized by political labels like "liberal" and "conservative" and what they imagine their opponents to be like more than they are by disagreements over issues like taxes, abortion, and immigration?
That news wouldn't surprise anybody who's spent time battling it out in a news outlet's comment section, and it's the firm conclusion of new research by Lilliana Mason, a professor at the University of Maryland.
Her paper, "Ideologues Without Issues: the Polarizing Consequences of Ideological Identities," published in late March by Public Opinion Quarterly, uses 2016 data from Survey Sampling International and American National Election Studies to study how and why Americans are politically polarized.
[...]
"There's been a debate within political science for a long time about whether or not the American public is polarized," Mason said in an interview with The Intercept. "I'm sort of making this argument that as you have multiple social identities that line up together, people hate their out groups more regardless of their policy positions."
She noted, for instance, that Americans who identify most strongly as conservative, whether they hold more left-leaning or right-leaning positions on major issues, dislike liberals more than people who more weakly identify as conservatives but may hold very right-leaning issue positions.
----
I was already under the impression such was true, because soccer.
It always looked odd to me how the fans of a team would hate the star scoring player of the opposing team one year, just to love him the next year when he changed shirts (and salary) to play for their team.
Kind of reminds me of Donald Trump, a life long Democrat, being the darling of the Right these days.
Clovis, Isn't each of your examples just tribalism? For example, "the conservative tribe" or "soccer tribe x"?
Bret,
Yes, with a caveat: there is barely any 'tribe' - in the sense of a particular subculture those people share - but mostly superficial labels they broadly identify with.
Clovis,
You're right. I'm using a kinda lame definition of tribe, where tribe = "in group" or maybe "community" and why folks consider themselves part of a "tribe" is fairly arbitrary.
BUT! The thing to consider is that once people decide they're part of a "tribe" many of the things associated with being tribal (in the more standard definition) seem to apply to these "in groups" and "communities" as well.
So not tribe by genetic or proximal relationships, but tribe by voluntary association.
Bret,
It is the height of irony: computer networks connecting roughly 4 billion people made us all more tribal, yet even those tribes are virtual too.
Maybe our brains are really ingrained with this tribal thing, and actually the physical limitations of the world were actors refraining it somehow. The more virtual we get, the more prone to mayhem we'll be.
Damn it, the all connected and united virtual world envisaged in the beginnings of the Internet is just another failed utopia. We can't get enough of them.
Post a Comment