Boiled Frogs: AI Slop, Phishing, Deep-Fakes, & Spam, Spam, Spam

Authored by Charles Hugh Smith via OfTwoMinds blog,

We’re frogs in a pot that’s being heated so gradually that we no longer notice the sewage is extinguishing the utility of the Web.

Let’s take “Show me the incentives, and I’ll show you the outcome” and direct it on the Internet, the digital realm that is now central to modern life The incentives are making money from attention, i.e. clicks, engagement, etc. by any means available, and a burgeoning universe of cons, deception, extortion and fraud.

And with those incentives, the outcome is an ever-expanding river of toxic sewage, a river of AI slop, deep-fakes, phishing, clickbait and spam, spam, spam on every device, every platform, every screen.

This is not inconsequential. A physician-correspondent recently reported that he was researching a cardiovascular condition online and realized the article he was reviewing was AI slop, a conglomeration of inaccurate diagrams and plausible-sounding nonsense slapped together to get whatever meager income would be generated by a modest number of views.

That’s the incentive the Big Tech platforms set up: since it’s a low-odds gamble that any post will go viral on a large enough scale to make serious income, the incentive is to post 1,000 AI slop posts which each collect 1,000 views. In other words, make it up on volume.

Since the Internet is global, people in low-income nations have an incentive to generate AI slop to earn what is a pittance in developed nations. The barrier to entry is low–anyone can produce veritable mountains of AI slop with free tools and low-cost bandwidth–and the gains, however modest, are welcome if paid work is scarce.

The same “make it up on volume” approach incentivizes churning out millions of phishing and spam SMS, emails and posts on every platform under the sun. If there’s only one sucker per 10,000 entreaties, then send out 10 million.

Since views and engagement generate income, the more outrageous the clickbait, the better. And of course, the greater the volume of clickbait, the greater the income stream flowing to platforms hosting the clickbait.

AI tools incentivize creating deep-fakes of celebrities’ voices and personas which can then be deployed to con older Internet users who are often credulous enough to believe that yes, Owen Wilson is talking to me, see, it’s him.

Every legitimate institution is now a tripwire for phishing and spam. Your USPS package can’t be delivered, here’s your Social Security Statement, and so on.

AI Search is broken, too. I couldn’t find the original PropOrNot “fake-news about fake-news” list from 2016, and AI search concluded it was not available. Then a correspondent sent me a post on Zero Hedge which prominently displayed the entire original PropOrNot list. (oftwominds.com was on the list, thank you very much.)

Washington Post Names Drudge, Zero Hedge, & Ron Paul As Anti-Clinton “Sophisticated Russian Propaganda Tools” (November 25, 2016)

The burden of shadow work required to delete, unsubscribe and purge our lives of all this sewage is growing heavier by the day. This calls to mind the boiled-frog analogy: we’re frogs in a pot that’s being heated so gradually that we no longer notice the sewage is extinguishing the utility of the Web.

And the reason is–drum roll–that’s how everyone makes money: views, engagement, scams, cons, fraud and above all, sheer volume. And who makes money from volume? The Big Tech platforms. So what if it’s misleading AI slop, deep-fake scams or clickbait, the more “engagement” we get, the more money we make.

So where’s the incentive to staunch the flood of sewage? There isn’t one. The incentive is to shrug and let the user sort it out by burning their own time.

If we want a different outcome, we have to change the incentives.

*  *  *

Check out my new book Ultra-Processed Life and my updated Books and Films.

Become a $3/month patron of my work via patreon.com

Subscribe to my Substack for free