All Articles
Alexi Announcements

Two Principles for Crafting an Alexi Memo: Deference and Concision

We are constantly thinking about legal research, and, in particular, identifying guiding principles for crafting an Alexi legal research memo.

First, it is important to understand that an “Alexi memo” is different than most other kinds of legal research memos. There are a lot of similarities, for sure, but the differences are less obvious.

The purpose of this post is to identify two of the key differences between an Alexi memo and more traditional legal research memos, and discuss an important principle for crafting an Alexi memo that emerges from each.

1. The Non-Iterative Constraint

Principle: Significant deference is given to the questioner.

Traditional legal research almost always involves some iteration – a back-and-forth between the lawyer with the question and the lawyer drafting the memo. This iteration can consist of a step-by-step adjustment to the issue statement, the relevant facts, and depth of research required, among others. Clearly, Alexi in its current state does not handle this iteration very well, and in fact has been intentionally eliminated. The specific reasons for this are beyond the scope of this note; the implications, however, are not.

The first important implication of an Alexi memo being non-iterative is that deference is always given to the questioner. That is, we must assume by default that the questioner is asking the right question. We can do this for two reasons. First, the questioner also recognizes this non-iterative constraint, and as such will usually spend a little more time arriving at a precise issue statement. Second, our users are lawyers, with general knowledge regarding the topic, and are usually correct in the way an issue statement is phrased.

In some cases, when it appears the question should have been phrased differently, there may in fact be an implied assumption or ignorance inherent to the question. The best answers, though, address this in a narrative that is sympathetic to the implicit misunderstanding. The right way to think about it is this: questions that appear to be improperly phrased are only ostensibly so; the ‘misunderstanding’ is merely an implied component of the question. Consequently, significant deference is still given to the questioner.

2. The Scope Constraint

Principle: the best answers are highly concise but sufficiently thorough.

Alexi memos are almost always 3-5 pages. Sometimes, but rarely, a memo may be 2 pages or even 6. This is the scope we have set intentionally, and our users understand this. The reasons for this scope are in part practical and in part substantive.

The practical reason is that it is difficult to provide a service like Alexi if there is significant variation in scope. The substantive reason is that by adhering to this constraint, the best answers will be very concise and much more useful. A very concise answer that includes only the most relevant information is exactly what we are striving for. 

Over the next several months we will be providing additional posts that set out more detailed principles of what we strive for in crafting an Alexi memo.

Leaders in generative AI for Law

Latest articles

Create your account today and get started for free!

AI has changed how lawyers do research. Try it for yourself with 5 free questions.

completed memos