One of the challenges we at Lagaviti face is explaining to lawyers the differences between the various artificial intelligence solutions available to them. What makes this difficult is that, at a distance, these solutions often appear similar: a text box where the user enters a question and receives a written response from the system.

​

But this barely tells half the story, because what truly matters is the work that takes place between the question and the answer. Lawyers should be well acquainted with this from their own professional experience, as they often spend a great deal of time and effort behind the scenes on a document that ultimately amounts to only a few pages.

 

To illustrate this more clearly, it is helpful to think in more visual terms by replacing lawyers and AI solutions with a contractor and the tools available to them for different tasks:

 

✅ If the contractor is planting a tree, a shovel is probably sufficient to dig the hole.

✅ If the contractor is building a deck, it may be useful to use a small excavator to dig the foundation, a circular saw to cut the boards, and so on.

✅ If the contractor is building a house, it is likely best to have a large excavator and drilling equipment, as deep excavation may be required to ensure a solid foundation. A concrete truck will also be needed to pour the walls, and the list goes on.

​

Most of us would be quite alarmed if a contractor we hired to build a house arrived on site with a group of workers whose only tools were shovels and pickaxes. On the contrary, we would expect the contractor to use the most powerful and suitable tools available in order to complete the project efficiently and professionally. At the same time, most of us would also recognize that such tools are more expensive to acquire and operate than less powerful tools, even though the latter may be perfectly adequate for certain tasks.

 

For a long time, lawyers had only a limited selection of tools at their disposal, but this has changed dramatically with the advent of artificial intelligence. Lawyers now have the opportunity to place greater emphasis on their roles as the “architects,” “project managers,” and “reviewers” of legal work, while delegating other, lower-value tasks to AI. Such an approach can lead to significant gains in efficiency and allows lawyers to devote more time to higher-value work and to undertake larger and more complex projects.

 

With the above example in mind, lawyers might ask themselves the following question: Is it better for me to use a “shovel,” a “pickaxe,” and/or a “small excavator” and then do everything manually that these tools cannot handle—or should I opt for a “heavy-duty machine” that can serve as the tool I need at any given time, a kind of all-purpose machine?

​

For me, the answer is obvious.

​

Tómas Eiríksson
13 October 2025