Video of a lawyer being busted mid-hearing for using ChatGPT to write briefs

quote:

---

17-year legal journey epitomizes a dedication to excellence. His proficiency as a plaintiffs' litigator is reflected in the meticulous handling of over a thousand cases across diverse legal realms. Holding memberships in multiple state bars and licensed in the US Supreme Court and 27 federal courts, Tristan's commitment to precision and integrity defines him as a leader in the legal arena. Clients can trust in his unwavering pursuit of excellence for unparalleled representation.

---

His bio.

Man. That hurts.

Stupid people assume that generative AI is to the level that it can consistently do the exact same quality of work that humans can do. It’s just a glorified search engine that can synthesize and communicate the things it finds through the materials it’s been trained on. It has zero incentive (because it’s a lifeless machine) to ensure its accuracy, so making shite up through misinterpretations or misguided findings is always a possibility.

What’s especially concerning to me is the research that’s come out about the impact on the brain when using AI too much. I imagine the lawyer in this video has literally lost his ability to read, understand, and synthesize legal materials because he’s used AI too much. I bet he was once a decent lawyer and now can’t bring himself to do the long work because he’s so used to letting AI do his thinking for him.

That’s not just conjecture. Emerging researching is showing the deleterious effects AI is having on us.

quote:

---

One caught, thousands not. Welcome to the future.

---

It’s one of the first things we look for when we get a brief. The question is how you handle it when you find hallucinations.

quote:

---

I’d bet he had an associate do the work and did not check it.

---

Naw. That guy is a solo firm. No way he has associates. Maybe a "paralegal" who was formerly a stripper that he represented in the past.

quote:

---

How do we know you didn’t just use AI?

---

You're an inanimate object

Also sounds like he blew the deadline for RFAs.

These are all indices of an overloaded lawyer.

quote:

---

Stupid people assume that generative AI is to the level that it can consistently do the exact same quality of work that humans can do. It’s just a glorified search engine

---

This is true. Nowhere close to ready for prime time.

I asked for a ready to drink protein shake with no artificial sweetener and no stevia.

I looked up the drinks it recommended and they had artificial sweetener. Oops.

I also asked it to do a 16-team snake draft of players in the MLB Hall of Fame. It picked multiple players not in the Hall.

It’s a phenomenon called hallucination. At its core, these programs create sentences by choosing a word that’s the most likely to follow after some other word.

If it doesn’t know the answer to a question, it sometimes gives answers that seem plausible, but are just random BS.

quote:

---

Are you an actual, real person?

---

quote:

---

How is it finding cases that aren't real? Is it pulling from fictional novels or movies?

---

The concerning part isn't that it finds or "hallucinates" information. The most concerning part is that it often doubles down on that fictitious information when challenged or questioned about it. It has created scientific research out of whole cloth, often attributed to real researchers or experts in a field, and given what was supposedly links to journal publication. Other times it just creates a fictional scientist to go along with the bullshite research. When challenged, it will double down and defend its information.