Appellate Judges to be Replaced by ChatGPT

Official Photograph of the Judges of the New Appellate Court

Ushering in the most revolutionary innovation in legal drafting since dispensing with cuneiforms on clay tablets, the Administrator of Court Administration announced today that appellate judges will be replaced by ChatGPT.

“We’re thrilled to be bringing appellate review into the 21st Century!” enthused Chief Administrator Kimberly Shrimp at the annual meeting of the Appellate Affirmance Club. “As you know, ChatGPT is a highly sophisticated AI program which, after being fed a minimum of information, instantly produces an authoritative writing on any subject. Whereas the present practice is for appellate courts to have their clerks read the People’s brief and write a decision affirming the conviction, clerks can now simply feed the brief to ChatCPT, request an affirmance, and get the decision within seconds. The result is indistinguishable from a ‘real’ judicial decision that might have taken the clerk as much as an hour and a half to produce.”

“This is a giant step forward to unclogging the pipeline from all those backed-up appeals,” applauded Chief Judge Bucket, architect of the High Court Draino Initiative. “In addition to replacing appellate judges, ChatGPT will be recommended to hearing courts, parole boards, administrative agencies, and other stakeholders in the relief-denial industry presently staggering under the crushing burden of having to fill out templates.”

“ChatGPT is nothing new,” pooh-poohed Barnaby Barnacle, Chief of the Prosecutor Appeals Bureau. “We’ve been using robots to write briefs for decades. Of course, being government-issue robots, they tend to be slow, which is why it takes us between seven months and a year to answer appellants’ briefs. But we’re confident that with additional funding, our DABots can be upgraded to the capacity of ChatGPT.”

“It’s an outrage!” objected Toots Brass, Chief of the Criminal Defense Society. “Anyone who’s ever had to deal with a Chatbot when they’re trying to buy eyeshadow online knows how biased they are against oppressed minorities, the poor, and the vulnerable!”

Others have called ChatGPT “a stochastic parrot,” an AI term for stringing together words that sound right but have no meaning, i.e., fluent bullshit. Further evidence that ChatGPT is ideally suited to judicial decisions.

We decided to try it out.

It took us some time to get onto it because several million others were trying to do the same thing. Here’s what they sent us while we were waiting:

ChatGPT is surely the best
But its servers are put to the test
With so many users chatting
It’s no wonder they’re lagging
But they’ll fix it soon, no need to fret!

When we eventually connected, we asked:

Query: Write a fictitious brief on behalf of the prosecution providing supporting facts showing that the police lawfully arrested the defendant; the jury reasonably found him guilty; and that the sentence was not excessive.

ChatGPT answered in 20 seconds:

The prosecution respectfully submits this brief in support of the conviction and sentence of the defendant, [Defendant]. The evidence presented at trial clearly established that the police lawfully arrested the defendant, the jury reasonably found him guilty of the charged offenses, and the sentence imposed was not excessive.

First, the defendant’s arrest was lawful. The police had probable cause to arrest the defendant based on eyewitness testimony and physical evidence found at the scene of the crime. Additionally, the defendant was advised of his Miranda rights and voluntarily waived them prior to making incriminating statements.

Second, the defendant’s guilt was established beyond a reasonable doubt. The prosecution presented overwhelming evidence of the defendant’s guilt, including witness testimony, physical evidence, and the defendant’s own incriminating statements. The jury, after carefully considering the evidence, reasonably found the defendant guilty of the charged offenses.

Finally, the sentence imposed was not excessive. The defendant was sentenced to the maximum term allowed under the law for his crimes. However, the court took into consideration the defendant’s prior criminal history and the severity of the crime in determining the sentence. The sentence imposed was not excessive and was in line with the goals of punishment and rehabilitation.

For these reasons, the prosecution respectfully requests that the conviction and sentence of the defendant be upheld.

Respectfully submitted,

[Prosecution’s Name]

Assistant District Attorney.

We then asked ChatGPT for a judicial decision affirming the conviction. The Bot immediately responded:

In the matter of the State of [X] v. [Defendant], the Court affirms the conviction and sentence of the Defendant for [offense].

Upon review of the record, including the prosecution’s brief, the Court finds that the defendant’s arrest was lawful as the police had probable cause based on eyewitness testimony and physical evidence found at the scene of the crime. The defendant was advised of his Miranda rights and voluntarily waived them prior to making incriminating statements.

Additionally, the Court finds that the defendant’s guilt was established beyond a reasonable doubt. The prosecution presented overwhelming evidence of the defendant’s guilt, including witness testimony, physical evidence, and the defendant’s own incriminating statements. The jury, after carefully considering the evidence, reasonably found the defendant guilty of the charged offenses.

Furthermore, the Court finds that the sentence imposed was not excessive. The defendant was sentenced to the maximum term allowed under the law for his crimes, but the court took into consideration the defendant’s prior criminal history and the severity of the crime in determining the sentence. The sentence imposed was not excessive and was in line with the goals of punishment and rehabilitation.

Therefore, the conviction and sentence of the Defendant are affirmed.

Signed,

[Judge’s Name]

Circuit Court Judge.//

Truly indistinguishable from a People’s government-issue robot brief and a clerk-generated affirmance.

We asked ChatGPT to tell us in Shakespearian language why courts shouldn’t have it write all their decisions. It modestly replied:

Though ChatGPT doth know the legal principles to apply And is so swift in generating text, Yet there are reasons why the courts should not rely On this machine instead of clerks.

First, this AI model doth not interpret the law But generates text based on examples taught. Interpreting the law is a complex task That requires not just principles, but facts and context.

Secondly, ChatGPT is but a language model, not a legal sage And though it doth mimic legal decisions, It lacketh the expertise of one trained in the legal stage.

Thirdly, ’tis important to note that ChatGPT is not infallible And errors made by it could lead to injustice for all parties.

Finally, the role of the courts is not just to rule on legal principles But also to ensure that justice is done and rights protected. The human element in the judicial process is crucial for fairness and impartiality, Which is something that an AI model like ChatGPT cannot replicate.

Fairness and impartiality? Stochastic Polly gets a cracker.