top of page
  • Robert Farago

AI Is Committing Suicide

Now what?

AI chatbots are the bomb. You want information in an easily digestible form, instantly? You’ve got it, no Bards held! The most obvious downside: chatbots are prone to “hallucinations.” In other words…

They get shit wrong. Sometimes badly wrong. Unless you know, you don’t know. The AI chatbots present “misinformation” with the same coherent credibility as they spit out accurate information.

I think it’s safe to say correcting AI’s simple factual errors isn’t an insurmountable task. The problem certainly isn’t frequent, obvious or serious enough to slow down its adoption. But wait! There’s more!

As we’ve seen with the non-debate over COVID-19 vaccines and the pre-election suppression of the Hunter Biden laptop story, the word “factual” is a slippery little bastard.

The AI industry was quick to respond to that challenge.

In case you’re laboring under the impression that fossil fuels are more environmentally-friendly than renewable resources, or don’t like affirmative action, rest assured that AI’s progress is Progressive. The guardians’ “guardrails” make it so.

To paraphrase an earlier digital creation, AI chatbots don’t mean to be wrong. They’re just programmed that way.

The information presented doesn’t come from a single reliable source. It’s an average based on data scraped from hundreds of millions of online sources: blogs, forums, news sites, wikipedia, my Substack, everything.

Lest we forget, AI chatbots are built on the back of billions of hours of human endeavor. Endeavor that AI chatbots are busy supplanting. As in destroying.

Why search for a website for information when you can just ask AI? Ba-bam! There it is. The intel may not be accurate, but it probably is.

Anyway, who cares? As long as it’s politically correct, chocks away! Think of all the time you save. Think of the death toll.

As I reported earlier this month, Google shared a preview of their new Search Generative Experience (above). It’s the tech giant’s attempt to split the proverbial baby.

What are the odds the user will dig any deeper than the AI “snapshot”? If we’re talking straight info retrieval, you can round it down to zero. Don’t take my word for it. Wall Street sees Google’s self-hoisting AI petard problem.

What was it die-hard Democrat Steven King said? Suffer Google’s little children. No wait. It was “Tomorrow a bad thing will happen.”

If no one goes to websites for information, information-based websites won’t get enough traffic to make enough money to stay in business. Less authoritative websites, less data for AI to scrape.

At the same time, cash-strapped/profit-driven websites are turning to AI to goose their profits. Germany’s Bild is replacing hundreds of editors with AI. Also “investigating the potential” of AI-generated copy: BuzzFeed, the Daily Mirror, Daily Express, and Men's Journal.

Which AI will be scraping. So, at some point, AI chatbots will be generating answers to prompts based on previous answers to previous prompts. It’s a digital ouroboros. The information it summons will be less reliable. More biased? That too.

The web as we know it – a viable economic eco-system for millions of content creators – is doomed. What does Web 3.0 look like?

For one thing, how-to websites are fucked. A chatbot asked for stair cleaning advice may quote/plagiarize Martha Stewart, but the answer eliminates the need to go to the ex-con’s website. Websites depending on informational blogs for Google rankings are SOL.

Product and service providers who relied on Google SEO – i.e., all of them – will have to pay a LOT of money to get listed on AI chatbots (the dictionary definition of exclusive). Did I forget to mention that AI recommendations will be for sale?

Vendors and service providers who can’t afford to pay-to-play – i.e., the vast majority – will have to redirect their money to less expensive ways to attract new customers online. My money’s on social media, but that’s an open AI question.

Large “destination” news websites already living in paywall paradise (e.g., The New York Times) are safe from AI exclusion. Smaller news or opinion websites who depend on Google listings to generate traffic to pay their people are paddle-less up the excrement tributary.

AI chatbots are magnificent. Setting aside accuracy issues, they make it possible to know more about more in less time with no money down (for the moment). And yet, ironically, the more successful AI is, the worse it’s going to get.

Good news or bad news? One thing’s for sure: it’s gonna be different.

0 views0 comments


bottom of page