Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
After all, he claims that Pearl is significantly less likely to provide false information than many other AI search engines — which he believes are likely to face “a tidal wave” of lawsuits based on the bad answers they provide. “Other players are creating amazing technology. I call them Ferraris or Lamborghinis,” says Kurtzig. “We’re building a Volvo—safety first.”
This pitch about the superiority of Perl, of course, made me even more eager to try it. Kurtzig seemed so confident that Pearl would still enjoy Section 230 protection. I asked the AI if it agreed.
Pearl said it likely qualifies as an “interactive computer service” under Section 230, meaning it would be protected from being considered a publisher, just as Kurtzig suspected. But, AI said, “Muta’s situation is unique because it uses AI to generate content.” After all it was not a definitive answer for me.
When I asked to speak directly with a lawyer, it sent me back to JustAnswer, where it asked me to provide the answer it wanted to verify. I said I’d have to go back and copy the answer, because it was several paragraphs long, but when I went back to the Perl website, the conversation was gone and reset to a new chat.
When I tried again, this time opening the Perl browser on the desktop, I got the same uncertain answer. I decided to trigger a human-data experiment; After a few minutes, I got the TrustScore™—a measly 3!
Pearl recommended that I seek the opinion of a real expert, porting me to its membership page. I was given a log-in so I didn’t have to pay while I tested the tool. It then connected me to one of its “Legal Eagle” experts.
Unfortunately, the lawyer’s answer was no clearer than the AI. He noted that there was an ongoing legal debate about how Section 230 would apply to AI search engines and other AI tools, but when I asked him to provide specific reasoning, he gave an odd reply that “most shell companies or associations use to file does.”
When I asked for an example of one such shell company – quite confused about what this had to do with a public debate about Section 230 – the “legal eagle” asked if I’d like to put him together a package. More confused I said, yes. I received a pop-up window indicating that my specialist wanted to charge me an additional $165 for data mining.
I declined, disappointed.
I then asked Pearl about Wired’s history. The AI response was serviceable, though essentially the same thing you’d find on Wikipedia. When I asked for its TrustScore™ I was again met with a 3, suggesting it wasn’t a very good answer. I chose the option to connect with another human expert. This time, perhaps because it was a question about the media and not a straightforward legal or medical matter, the expert took some time to appear – well over 20 minutes. When he did, the expert (it’s never been established what gave him his media credibility, though his profile indicates he’s been working with JustAnswer since 2010) gave me a remarkably similar answer to AI. Since I was doing a free test, it didn’t matter, but I’d be annoyed if I paid a subscription fee to get the same mediocre answer from both a human and an AI.
For my last stab at using the service, I went for a straightforward question: how to refinish kitchen floors. This time, things went more smoothly. The AI gave an adequate answer, similar to a transcript of a very basic YouTube tutorial. When I asked human experts to assign a TrustScore™, they gave it a 5. It certainly seemed accurate enough. But—as someone who really wants to DIY the old pine boards in my kitchen—I think when I actually go looking for guidance, I’ll rely on other online communities of human voices that don’t charge $28 a month: YouTube and Reddit.
If you’ve finished testing Perl or another new AI search product and had a memorable experience, please let me know how it went in the comments below the article. You can also contact me by email kate_knibbs@wired.com. Thanks for reading, and stay warm!
