From a live tweet of the proceedings around the lawyer caught using ChatGPT:

"I thought ChatGPT was a search engine".

It is NOT a search engine. Nor, by the way are the version of it included in Bing or Google's Bard.

Language model-driven chatbots are not suitable for information access.

>>

Honestly, this lawyer is actually lucky, bc he is working within a system that was able to catch his misstep.

When MSFT, GOOG & others present their chatbots as a replacement for search, they are setting people up for similar fails, typically in much less regulated spaces.

>>

Say it with me as many times as it takes to make the lesson stick: Chatbots are not a suitable replacement for search engines.

https://dair-community.social/@emilymbender/109570353001872254

Emily M. Bender (she/her) (@[email protected])

Chatbots are not a good replacement for search engines https://iai.tv/articles/all-knowing-machines-are-a-fantasy-auid-2334

Distributed AI Research Community

@emilymbender

Doesn't this demonstrate just how poorly people understand what's going on behind search engines? And so the only thing they have left to judge output on is "how much does this sound like a person?". Big tech is definitely exploiting this, but how are we giving them anything else to go by?

@jztusk

For my part, I am putting enormous effort into educating the public about this, including with this thread and the one that it links to.

@emilymbender

👍 That's excellent! The tech areas I follow aren't very good about doing that. Glad that you are.

@jztusk @emilymbender Bear in mind though, this wasn't a case of even using a search engine. That lawyer didn't use Bing chat, you.com or any other search engine that at least claims to provide results that exist somewhere.

The lawyer used ChatGPT. It's a make believe engine. And it's been widely advertised as such.

If that lawyer can't even bother to use an actual legal database, is the issue the tech platforms?

@emilymbender

Even search engines are not search engines. At best they are selective web crawlers, that offer a biased search of their local database of outdated webpages. Then show sponsored links like “Buy supporting cases at amazon”, then popular pages that contain some of the search terms and sometimes the correct search result on page three.

@Life_is @emilymbender
*SIGH* but webcrawlers are cheap once they’re running, and maintaining a database that will answer semantic queries about locations on the web is probably very expensive (I doubt anyone knows for sure just what the cost would be if we had one). And we all know what a profit-driven tech corporation will choose to do given those alternatives.
@emilymbender I can't boost this enough
@emilymbender I don't know how phind.com works or defines it's product... I just recognized it's the first one I tested so far answering to my question somehow it's to hard to answer it with the available information instead of printing out something convincing but wrong.
https://www.phind.com/search?cache=aa9381f4-3343-4c8b-a951-f86c3eed8b1a
What do you think about it?

@emilymbender it seems the AI techbros have concluded that PEBKAC is something they can solve and profit from because Problem Exists Between Keyboard and Chair is a universally applicable analysis of monetizable need.

Unfortunately, eliminating the Human between Keyboard and Chair assumes the entirety of the Human is worthless

@emilymbender I'm willing to concede that there _may be_ a _theoretical_ chat bot/natural language dialogue system that could work better than a search engine for _some_ people on _some_ domains, because we all don't understand or interact with information in the same way, and I'm an inveterate optimist.

But yeah, the current crop, as a catch-all tool for _everything_? Not even close.