As many people have pointed out now, Google’s search engine — the “killer app” that made Google one of the world’s largest, wealthiest companies — has become completely useless.
I was taught how to do boolean searches in high school, by a librarian at what was then known as Penn State McKeesport, in order to use the computerized catalogs there. So when search engines came along, I felt like a wizard. I’ve always prided myself on my Google-fu and being able to dig out weird stuff using Google search.
Not any more. Even when I’m using specific queries and functions to try to force Google to find something on my own websites, the search engine sometimes comes up with “no results.”
Oh, but these days, if Google doesn’t find results, it’s happy to make them up.
I recently wrote about someone and afterwards found out I had their title wrong. I apologized and wrote a correction, but wondered how I had made such a lame-brain mistake.
I entered the person’s name into Google Search, and Google’s “suggested result” was the incorrect title. So I hadn’t simply hallucinated it; Google’s artificial intelligence — more like artificial lack-of-intelligence — had imagined a title for the person whom I was writing about. I had then used that incorrect information when I wrote my story.
(Yes, the ultimate mistake was mine. If I had gone directly to the person’s website, I would have found her correct title there.)
As for web search, I’ve recently started paying for a service called Kagi, an alternative search engine which incorporates results from many other search engines. It doesn’t have as many “suggested links” and “shopping” links as Google Search and it also doesn’t have as much A.I. generated crap. It’s costing me about $10 a month, which is not cheap if you’re on a fixed income, but it’s become a necessity in my work.
Because, to put it bluntly, Google Search is completely broken; I’m astonished how bad the results are now. It’s taken giant steps backwards. Don’t trust it.
Alas, Google’s parent company, Alphabet, owns a big chunk of the web now, including the largest video streaming provider, YouTube.
At the top of this page, Fran Blanche, a musician and electronics technician (she’s the genius behind Frantone effects pedals) in Philadelphia, discusses how YouTube is now forcing creators to accept A.I.-generated summaries of their videos.
A few of the summaries are OK, Fran reports. Others are wildly off-base and contradictory.
YouTube’s A.I., she reports, is also suggesting replies that creators can semi-automatically post as responses when someone comments on one of their videos.
Why? Who asked for this? Who wants this? Certainly not the consumer of Fran’s videos.
I know I don’t want to read a response that was bullshitted by artificial intelligence. What value does it add?
If a creator doesn’t care enough to write a response to me, and instead tasks an A.I. bot with writing the response, why should I care enough to read it?
Over the weekend, for my own amusement, I asked ChatGPT to tell me about myself. The response starts out OK, but then goes horribly, hilariously wrong:
Jason Togyer is a journalist, author, and communications professional known for his work in various media and nonprofit sectors. He is perhaps most recognized for his book, “For the Glory of Their Game: Stories of Life in the Pittsburgh Steelers Football Nation”, which explores the deep connection between the Pittsburgh Steelers and their fans. The book highlights the cultural and social impact of the team on the Pittsburgh community.
I have never written anything about the Steelers. Ever. (If you listen to me on Saturdays, you know how dim-witted and out-of-touch I am when it comes to sports.) Actually, a book about “the deep connection between the Pittsburgh Steelers and their fans” sounds like it might be interesting; maybe someone should write one.
But unfortunately, no such book by that title exists, so far as I can tell.
ChatGPT made up a book no one has ever written and attributed it to me, and says it’s my best-known work. All-righty then.
ChatGPT continues:
Togyer is also associated with Carnegie Mellon University as an editor for the university’s alumni magazine, “Carnegie Mellon Today.” He has written extensively on topics such as technology, engineering, arts, and education.
I did attend Carnegie Mellon University (I am obviously on CMU’s radio station, WRCT). I also worked for CMU — past-tense — from 2008 to 2016 as editor of the computer science magazine and assistant director of communications.
But CMU’s alumni magazine is not called “Carnegie Mellon Today.” I never worked for CMU’s alumni magazine. And “Carnegie Mellon Today” hasn’t been published since 2017. (It was a newsletter.) I never worked there, either.
I then asked ChatGPT, “What can you tell me about Jay Thurber, a DJ at WRCT in Pittsburgh?”
Jay Thurber is the on-air persona of Jason Togyer, a Pittsburgh-based journalist and media personality. He hosts “Radio 9 with Jay Thurber,” a popular oldies show broadcast on WRCT-FM (88.3) in Pittsburgh and simulcast on Tube City Online Radio in McKeesport. The program airs on Saturdays from 12 to 3 p.m. and features a mix of vintage hits from the late 1950s, 1960s, and early 1970s, along with comedy segments, public service announcements, and occasional interviews. The show combines humor, nostalgia, and unique commentary, making it a favorite among local listeners.
Thurber’s broadcasting style emphasizes community connection, and he frequently incorporates local history and culture into his programming. The show originates from McKeesport, with a focus on preserving the charm of classic radio while embracing modern platforms to reach a broader audience.
For more details, you can explore his website or listen to his broadcasts through WRCT or Tube City Online Radio.
Somewhat surprisingly, this is probably 90 percent correct, with the caveat that I do not consider my show “popular” (I love you, my listeners, but any given show on KDKA or WDVE has 100 to 1,000 times the audience that I do) and I don’t consider myself a “media personality” (some would say I have to get a personality first).
The maddening thing about A.I. that’s currently being forced on us by big technology companies is that it sometimes gets things right. I’ve been on the radio in Pittsburgh since 2000, and that summary of my show is probably better than anything that’s ever been written about my broadcasting work.
But too often, A.I. gets things horribly, horribly wrong. It tells people to make sandwiches using ant poison as an ingredient. It tells people to put glue on their pizza.
Imagine if your company made toasters, and you developed a new kind of toaster. Fifty percent of the time, it makes toast that is perfect and delicious. Twenty percent of the time, the toast is either a little bit too dark or a little bit too cold.
And 30 percent of the time, the toaster catches fire and injures the person using it.
You probably wouldn’t dare try to sell that toaster. You’d be afraid your company would be sued out of existence.
But big technology companies like Google (Alphabet) and Facebook (Meta) do this kind of crap all of the time. They put out products that malfunction in horrifying ways maybe 30 percent of the time, and then they blame the users. They take no responsibility for their own actions or for the horrible consequences. (Tesla is perhaps the poster-child in this category.)
A large percentage of Americans believe complete made-up horseshit about vaccines, elections, immigrants, LGBTQ people, teachers, the Moon landing, the climate, electric cars, cancer cures, and scores of other topics. In some cases — as with people trying out dangerous cancer cures such as “black salve,” or not giving their children life-saving vaccines to protect against polio or measles or COVID — this horseshit is literally killing people.
Yet Google (via YouTube) and Facebook have turned themselves into giant, highly profitable horseshit delivery mechanisms. Do they not care?
In fact, I’ve been told by people who work for Google and Facebook that they do care. It’s just that they believe “the marketplace of ideas” needs to be free. Some of them have even quoted Thomas Jefferson: “Errors of opinion may be tolerated where reason is left free to combat it.”
The best ideas, they tell me, will win in the marketplace of ideas.
Well, that’s fine, except that Google and Facebook have sucked up the advertising money that once fueled actual news gathering and journalism.
At the same time they are enabling misinformation and deliberate lies, they are disemboweling truthful information. The errors of opinion run rampant and get wealthy fleecing the gullible. Reason is out in the alley, begging for scraps.
And Thomas Jefferson lived in an era when the printed pamphlet was the fastest way to distribute opinions. You needed a human printer to print those pamphlets. Ben Franklin and other printers exercised editorial control over what came off of their presses. Google and Facebook don’t.
Yes, the Internet has made wonderful things possible. That’s the 50 percent of the time when we’re getting perfect toast.
But the rest of the time — when big Internet companies are spreading complete misinformation, and enabling Nazis and white supremacists and fascists and authoritarians to take power around the world — I have to question whether the occasional slice of warm bread is worth the price we’re currently paying.
Edited to Add: Technology journalist Ed Zitron writes extremely lucidly about complicated topics such as A.I., and why these technologies don’t live up to their promise as implemented by giant conglomerates. Check out his latest entry at his blog, Where’s Your Ed At?
Amen, a-freaking-men.