Lawyer cites fake cases invented by ChatGPT, judge is not amused

The new "dictated, not read"

The problem comes from people who call LLMs AIs. Then people who don't know how they work assume it is intelligent when it is not. I'm pretty sure that OpenAI is at fault in this by not informing users of the truth.

Quite so. I had a visceral aversion to ML being rebranded AI. ML only very recently became anything resembling AI with ChatGPT. (Admittedly it’s so good as a coding assistant that it does feel like talking to Data from Star Trek though.)

it gave me a longer, but yet again very positive sounding summary;

> The article is about the use of large language models (LLMs) in the legal profession. The author argues that LLMs have the potential to revolutionize the way lawyers practice law, by automating many of the tasks that are currently done manually. For example, LLMs could be used to generate legal documents, research case law, and identify relevant precedents. The author also argues that LLMs could be used to improve the quality of legal services, by providing lawyers with access to a vast amount of information and expertise. However, the author also acknowledges that there are some challenges that need to be addressed before LLMs can be widely adopted in the legal profession. These challenges include the need for LLMs to be trained on large amounts of data, the need for LLMs to be able to understand the nuances of legal language, and the need for LLMs to be able to generate accurate and reliable results. Despite these challenges, the author believes that LLMs have the potential to make a significant impact on the legal profession, and that they are likely to become increasingly important in the years to come.

Here are some of the benefits of using LLMs in the legal profession:

LLMs can automate many of the tasks that are currently done manually, freeing up lawyers to focus on more complex work. LLMs can access and process vast amounts of information, which can help lawyers to make better decisions. LLMs can identify patterns and trends that would be difficult for humans to see, which can help lawyers to spot potential problems. LLMs can generate creative solutions to legal problems, which can help lawyers to find new and innovative ways to represent their clients. Here are some of the challenges that need to be addressed before LLMs can be widely adopted in the legal profession:

LLMs need to be trained on large amounts of data, which can be expensive and time-consuming. LLMs need to be able to understand the nuances of legal language, which can be difficult for computers to do. LLMs need to be able to generate accurate and reliable results, which can be a challenge for complex legal problems. Overall, the author believes that LLMs have the potential to make a significant impact on the legal profession, and that they are likely to become increasingly important in the years to come.

Yes and no. Court decisions do generally follow a structure, but the decisions and the reasons for a final determination, may not always be clear. Judges also may throw in hypotheticals which whilst informative are not determinative. Once it gets to appellate courts and with different judges which may agree with the same outcome but on different grounds, it can get really hard to distinguish what is the test and the factors for a certain determination. It can also happen that dissenting opinions can be very persuasive, and subsequently adopted.

The task effectively also requires a case and paragraph impact meter which do exist in some law databases to one extend or another, effectively weighing how subsequent rulings consider, weight, and follow past cases, caveats, exceptions, and outright considering past rulings as bad law.

Then you have the issue of changing laws and the impact these may have on past cases as they may change the test and requirements needed to be considered and even much new case law needed to be developed to interpret the new legislation. So the model would need to have a historical knowledge of the law and how it was applied.

You would also need to feed it relevant surrounding information that may aid in interpreting said law. In the US, clearly the founding father's opinions and beliefs appear to play a significant part on the currently more originalist interpretative school of thought.

In the UK/Australia for example readings in parliament and even the underpinning reports that prompted the change in legislation may be considered where there is ambiguity in order to interpret legislation. Australian legislation nowadays also tends to incorporate an objective of the legislation and a section that says that where ambiguity exists to interpret it in a way that would further the objectives of the legislation.

So, it's really not a trivial problem.