Steve Cuozzo tests AI, finds it gets him all wrong | Latest Tech News
Content about my demise has been vastly exaggerated.
A current Chat GPT post summarizing one of my articles referred to me as the “late” author Steve Cuozzo, before restoring me to life the next day.
If artificial intelligence is so good, how come it’s so silly? In fact, “AI” may better stand for “Appalling Ignorance.”
The Post’s Steve Cuozzo has no style for AI. EMMY PARK
We’re told AI will soon control our governments, minds and our bodies. The pope warned us to beware AI’s dehumanizing, “Tower of Babel” impact. Economists concern it may wipe out jobs and cut back city downtowns to uninhabited Pompeii-like ruins.
But, those who attribute life- and humanity-altering energy to AI ought to first question the bots about themselves to see whether or not the platforms mangle their identities and histories.
The apocalyptic prognosticators not often think about whether or not AI is any good at one of its core missions: to acquire, distill and synthesize info into a type that transcends what extraordinary serps can do.
Sure, ChatGPT can readily make an elephant dance with a monkey or create completely made-up people just like the “actress” on the duvet of the Sunday New York Times Magazine. But it’s all thumbs when it comes to data of real people.
Buzzy Claude is no better.
Dario Amodei is the co-founder and chief government officer of Anthropic. Earlier this week, the California-based company, which is valued at almost $1 trillion, filed with the Securities and Exchange Commission for an initial public offering. I received’t be investing. Bloomberg via Getty Images
I never wrote a cookbook. I never lived in Greenwich Village. I’ve no kids. But Claude — the marquee product of Anthropic PBC — said I had achieved all of those issues when I requested it about myself.
Claude is aware of I’m a longtime author on business real estate. Way to go! But when I requested whether or not I’ve written about Herald Towers on West thirty fourth Street, it discovered “nothing specific.” It even recommended I search the New York Post web site! Isn’t that what AI is for? In fact, I’ve written about Herald Towers at least thrice, most lately in January when I reported TJ Maxx is opening its first Big Apple store there in 10 years.
Earlier this week, California-based Anthropic, which is valued at almost $1 trillion, filed with the Securities and Exchange Commission for an initial public offering. I received’t be investing.
But Anthropic’s Claude failed to discover basic info about Cuozzo.
Of 5 AI platforms I examined, Elon Musk’s Grok did the best job of summing up my life and profession. It even precisely and pretty in contrast some of my restaurant reviews with former New York Times critic Pete Wells’ takes on the same locations — an spectacular show of data, context and understanding.
But issues went down the rabbit gap elsewhere. Although Chat GPT, owned by Open AI, does a fabulous job analyzing Vladimir Nabokov’s “Lolita,” it’s a lazy lunkhead when it comes to little outdated me.
Not only did it say I used to be lifeless, it also couldn’t inform if I used to be educated. When requested about my education, Chat GPT claimed, “Publicly available biographical information about Cuozzo does not show a detailed educational background such as colleges attended or degrees earned … there is little widely published information about his formal education.”
Ahem —my degree from Stony Brook University is definitely discovered on Wikipedia and websites that crib from it.
Perplexity AI boasts it “provides accurate, trusted, and real-time answers to any question.” So, I tossed it a hanging curveball — did Cuozzo ever write a guide?
Claude claimed Cuozzo had written a cookbook about finger meals. He has achieved no such factor.
“I don’t see any evidence that Steve Cuozzo has published a book or memoir,” got here the reply.
Now, that’s perplexing! My 1996 “It’s Alive: How America’s Oldest Newspaper Cheated Death and Why It Matters,” pops up tons of of occasions in any cursory web search and on Amazon.com.
Meanwhile, Gemini AI claimed I “co-authored” a 1990 guide — “Power Partners: How Two-Career Couples Can Play to Win” — in which I had no hand at all,
Gemini, like most AI engines, lifts from Wikipedia the best way high college college students of my day ripped off encyclopedias in their time period papers, which is to say not very effectively.
Cuozzo provides both Dot Cakes and AI a go. Steve Cuozzo/ NY Post
“Power Partners” was written by my spouse Jane Hershey Cuozzo and S. Diane Graham, as Wikipedia accurately says.
Claude gave me the largest belly-laugh when it claimed, “Cuozzo authored a book called ‘The Finger Food Cookbook’ published in 1997.”
I wrote no such guide. A few comparable titles exist by different authors.
Once a howler pops up in AI land, there’s no stamping it out. The finger-foods fiction made its approach into Perplexity, which said, “another source mentions” my authorship of the nonexistent work.
Maybe, AI will someday surpass human intelligence as Musk predicted. Or possibly, as detractors concern, trigger so many job losses, it will cut back metropolis downtowns to empty, Pompeii-like ruins.
But, until then, it wants to get its act together — and be taught how to steal from Wikipedia.
Stay informed with the latest in tech! Our web site is your trusted source for breakthroughs in artificial intelligence, gadget launches, software program updates, cybersecurity, and digital innovation.
For recent insights, skilled coverage, and trending tech updates, go to us recurrently by clicking right here.
