In the debate over the value of legaltech startups, a common criticism is that they offer little more than a “thin wrapper” around a large language model (LLM). This criticism is reminiscent of the distinction between making a simple sweet and creating a Hershey bar, raising intriguing questions about what value is truly embedded within a wrapper.
When consumers purchase a Hershey bar, they might think they’re paying purely for the chocolate. However, the wrapper itself carries significant value. It assures a consistent experience, delivering quality no matter where it’s purchased. Additionally, it acts as a conveyor of essential information, such as nutritional and allergen details, thus ensuring compliance with various regulations.
In the legaltech world, much like a well-designed candy wrapper, a robust AI wrapper incorporates brand integrity, reliability, and trust. Legal professionals rely on this layer for assurance on issues like security and privacy. The wrapper supports integration with existing systems and provides assistance for navigating legal-specific workflows.
Industry leaders like Lexis and Westlaw exemplify strong commercial-grade wrappers in the legal sector. These service providers not only deliver databases filled with essential legal information but also foster trust through brand reputation and comprehensive support, training, and user community interaction.
For legaltech companies, especially startups, the challenge becomes creating a wrapper that resonates as credible and indispensable. Many investors focus on the potential for productization, customer relationships, and workflow adoption. An AI solution may hinge on the underlying model’s functionalities, but the trust built through an effective wrapper is invaluable.
As the tech landscape continues to evolve with open-source solutions taking root, the comparison to Hershey’s iconic offering underscores a vital consideration: while anyone can attempt to replicate functionality, crafting a consistent, trusted experience wrapped in user understanding is a more formidable—and necessary—task.