One Solution Fits All or Single and Specific?

Recently, I caught myself asking the same question of everyone I met professionally, whether at conferences or while initiating conversations. Equally, I have been asked the same question by almost everyone: How are you using AI in your work? And the answers have been informative.

In this post, I want to explore this question, focusing on the applications, solutions and systems used for regular day-to-day operations in organizations, rather than the underlying LLMs powering these applications.

Before AI, there were already different solutions

Before ChatGPT made its debut (sort of), there were already different solutions designed for specific requirements. And that made sense too. In-house legal teams have different requirements from law firms. CLM and spend management systems for in-house teams serve different purposes from contract drafting, reviewing and billing systems used by law firms. Interestingly, matter management across different practices within a law firm also has different requirements (I am biased towards IP management systems). To a certain extent, management teams making technology investment decisions understood these differences.

What changed with AI?

Some of the most recognized AI solutions in the legal industry include Harvey, Kira and Relativity AI. While they were developed with different purposes and strengths, somehow, somewhere, they started being compared with one another. The same thinking has found its way into the question, How are you using AI in your work? — including when asked by yours truly.

With AI penetrating almost every application, and the growing expectation that it can replace certain activities performed by junior associates, the perception seems to be that AI can do almost everything, disregarding the differences in requirements that still exist. One of AI's strengths is its ability to adapt to different tasks. This has increasingly shifted the conversation from selecting solutions for specific organizational requirements to evaluating AI platforms based on the range of capabilities they offer. But organizations still have different requirements across functions, practices and levels of work. The ability of one platform to perform multiple activities does not necessarily make it equally suitable for all of them.

What my experience tells me

To answer the question, I use different applications in my work. There is no single solution that fits everything, even if an AI application claims otherwise. While one may argue that influencers are pushing different applications, there is merit in using different tools for different purposes.

For research and brainstorming, I use ChatGPT, Perplexity and Claude depending on the nature of the task and quality of results. For work-related documents, Copilot is more convenient within M365, while extensive document reviews may require specialized applications. My choice depends on the task, expected output, performance (review and creation), and cost. 

Take patent drawings, for example. I prefer MS Visio or PowerPoint because patent drawings must meet specific technical and legal requirements. While generative AI can create images providing creative flexibility, and some patent drafting tools offer drawing capabilities, producing compliant drawings requires technical control that general-purpose AI applications may not consistently provide.

In my experience of using Legora, which supports legal research, contract review, drafting, tabular analysis and workflow automation, I have found its tabular analysis of large volumes of documents particularly useful.

But this raises a broader question. When an organization already has specialized systems supporting established workflows, should the expanding capabilities of an AI platform necessarily lead to replacing those systems? Or should AI complement the existing technology environment where that makes better operational sense?

From individual preferences to organizational decisions

An individual choosing different applications for different activities and an organization investing in enterprise technology are two different questions. Organizations already use different systems for different functions. With AI platforms expanding their capabilities and integrations, they now have another option: bringing multiple activities into a common environment without necessarily replacing every underlying system. This offers potential advantages in accessibility, integration and management of technology. But it also raises questions about overlapping capabilities, existing investments, workflow suitability and costs.

The decision, therefore, should not be driven merely by how many activities an AI platform can perform. Organizations need to assess whether consolidating capabilities improves their operations, whether specialized systems continue to serve particular requirements better, and whether the additional investment delivers sufficient value.

One solution fits all or single and specific?

While the question is binary, the answer need not be.

An integrated platform may be suitable for some requirements, while specialized solutions may remain necessary for others. The decision should begin with understanding the organization's operations, not the expanding list of capabilities offered by AI applications.

The question is not how much AI can do, but how much of what it can do an organization actually needs.