Attorney in Trouble for ChatGPT Generated Fake Citations

As the excitement surrounding AI has reached a fever pitch in recent months, many people are concerned that ChatGPT could eventually force them out of work. That dread may come true sooner than anticipated for one New York attorney, but for a different reason than you may suspect. The New York Times reported that lawyer Steven Schwartz asked OpenAI's chatbot for help with writing a legal brief, and the result was catastrophic.?

The opposing attorney drew attention to the hypothetical situations that ChatGPT had referenced in its legal brief. Six of the submitted cases confirmed by a US district judge "appear to be bogus judicial decisions with bogus quotes and bogus internal citations".

On a lighter note, I personally think the AI did this on purpose, for fun.?

In an affidavit filed on Thursday, attorney Schwartz wrote “he was unaware of the possibility that (ChatGPT’s) content could be false." He even shared screenshots showing that he had asked ChatGPT if the cases it cited were real. The program responded that they were, claiming the decisions could be found in " legal databases," including Westlaw and LexisNexis.

This form of AI could change lives, including those of people who are unaware of ChatGPT or any other similar AI capabilities. We are just beginning to anticipate the implications that these AI applications will have.

Watching the untamed beauty of the wild horses speed through the lake is an amazing sight. Thematically, the way they move is like watching poetry in motion.?

What about the broken horse??

A broken horse is typically thought of as one that has been domesticated by humans. A "wild spirit" that has been restrained and transformed.

I do not intend to horseplay, but any comparison between AI and a real thing like a horse is not intended here - I am merely using this as a metaphorical construct; today's AI is not sentient. You aren't using the "wild spirit" version; rather, you are using the iterated (domesticated) version. The old saying goes that history was borne on the back of a horse.?

Get to it

The typical approach for using ChatGPT or other such generative AI is to have an interactive chat. Doing so is undoubtedly astounding and occasionally startling as the conversations sparked by AI seem to flow naturally.

As most of us know that the ChatGPT we use now is actually a carefully modified version of the ChatGPT that was first developed. It is true that ChatGPT underwent a number of iterations before its creator decided that the AI program is ready to be made available to the general public.

How does generative AI do the generation aspects?

Take for example AI-powered image generators. Several pieces of online art were pre-scanned using computer algorithms in a text-to-art generative AI manner, and the elements of the scanned art were computationally analyzed for the components involved. Think of an online image with a horse in it. Consider another, independent image with a chair in it. Another image includes an umbrella. A sort of mathematical network is subsequently developed when these components are computationally identified, sometimes without human guidance and other times with it.?

The AI program uses the mathematical network to discover and put together those parts when you ask it to make artwork of a horse on a chair with an umbrella. It is possible that the final work of art will not turn out as you intended. The horse might not be that attractive. Even though the horse was barely covered by the umbrella, you wished the umbrella were bigger. You wanted the horse to be sitting instead, the image of the horse is currently standing on the chair.?

The wonderful thing about these artificial intelligence programs is that you can repeatedly make your request and, if you would like, provide more details. As a result, you modify your request and say that you would like a lovely horse sitting on a chair holding a large umbrella. The newly created image might be more like what you were looking for.?

One significant drawback is that essays generated (you provide a text prompt and receive an essay in return) by generative-based AI apps may contain a variety of fabrications, such as facts that are obviously wrong, and inaccurately depicted, like what happened to the attorney. These created elements are frequently described as a type of AI hallucination, which regrettably is gaining popularity, nonetheless.?

AI chatbots are still work in progress.

The makers of ChatGPT, OpenAI, Google, and Microsoft have stated that the AI-powered chatbots are still in development and that the responses they give may not be accurate. These businesses have also emphasized that users should not accept these chatbots' responses at face value.?

Assuming AI will continuously train itself like a human is the greatest mistake you can make. If there is no constant human input (content), AI becomes obsolete. AI is a tool of the future, not the future itself.?

Attorney Steven Schwartz said, "he greatly regrets" using ChatGPT "and will never do so in the future without absolute verification of its authenticity."

On the other side of the world in India, an accused used ChatGPT to cheat in Telangana State Public Service Commission (TSPSC) exams. AI helped get answers in real time. ?