This article was originally published Architecture and governance.
New technologies always make headlines, and artificial intelligence (AI) is no exception. Remember when innovations like the telegraph, the radio, and the iPhone first hit the market? People often get excited and get ahead of themselves. Hype about new technology can drive adoption, but it can also lead to disappointment when things don’t go as planned and promised.
Gartner calls this process from hope to disappointment the hype cycle. A catalyst for innovation creates an upward trend that eventually reaches a peak. For AI, the first peak of hopes might have come when IBM’s Deep Blue defeated chess champion Garry Kasparov. But it would be unwise to discuss AI as a broad, single category defined by one game of chess. All applications are at different stages of the hype cycle. Google recently hyped up its own chatbot, but lost a massive $100 billion market cap after the tool made an error during a live demo. ChatGPT is another AI chatbot that has been making headlines as many experts see it as raising hopes.
The question is, how do you cut through the noise to determine if a particular AI solution is approaching the trough of disillusionment or has already overcome it and is delivering real value? is. Whatever your use case, I recommend approaching artificial intelligence with a healthy dose of skepticism. More specifically, to assess whether the particular application you are considering is hype or reality, ask him the following five questions (in the context of the Corporate Legal Department (CLD): will be used to explain).
- Can this be done without AI? Artificial intelligence is sometimes touted as a silver bullet for an organization’s toughest problems. But if departments don’t understand how to properly deploy people, processes, and technology to accomplish specific tasks without AI, adding AI is unlikely to work. Many CLDs were intrigued by the hype about AI for Contract Lifecycle Management (CLM), but in reality they understand CLM well enough to know how it can help them. was not So it’s no surprise that most of the AI offerings for contract lifecycle management have been disappointing.
- Are the goals clearly defined? Another reason the application of AI to contract lifecycle management has been a major disappointment in the legal arena is that the technology promises to boil the oceans. AI is more likely to be effective when working on something very specific and well-defined. Simply put, ambiguity is a big red flag. For example, instead of promising to manage the entire contract lifecycle, look for technologies that offer unobtrusive solutions for one part of the lifecycle. For example, his changes to one section of the contract apply to the entire document. Using AI to enhance the scrutiny of bills and apply external counsel guidelines is also much easier and more clearly defined.
- How much data is available? AI works well not only in well-defined missions, but also in data-rich, well-structured environments. Without centralized, standardized and anonymized data, AI tools will struggle to achieve their intended purpose. For contract-related AI, there is often not enough training set to create a useful tool. On the other hand, in legal invoice review, mature tools have hundreds of billions of dollars in legal invoices to work from and learn from. Additionally, all of that data is in the industry-wide LEDES format. Having a large amount of structured data is a big green light for developing AI products.
- How established is the company selling your product? Buying AI solutions from more mature companies has many advantages. First, it will likely have a larger data repository to train the technology. Additionally, startups often have to make overblown claims to secure funding and drive adoption. Being a newcomer you are essentially in a hopeless position. Survival is not in jeopardy if you are a large company. Still, further due diligence is warranted. Talk to existing users to assess the credibility of the company you are considering. And again, be skeptical of everyone and everything.
- Is it a top priority for your department? Finally, it’s important to ensure that the AI you’re considering is aligned with departmental and, ideally, company-wide goals. Hold an organization-wide brainstorming session to determine your top three to five priorities. If the priority cannot be defined, the technology sales person will define the priority. While we outline the priorities, we also outline the easiest and least tedious ways to resolve each priority. You may not actually need the shiny new technology, and you should start scrutinizing AI solutions only for those use cases that really need such new technology. AI, like anything else, can become obsolete if not properly evaluated and implemented.
There are many applications where AI can live up to the hype. But there are just as many places where it is destined to fall short. The truth is, the hype cycle around AI (or any other technology) is neither good nor bad. It’s simply an unavoidable pattern, much like the sun rises and sets each day. There is no doubt that AI-powered tools are becoming more nuanced and sophisticated as they benefit from the accumulation of data. Especially in the legal context, many departments have overcome their initial hesitation, largely thanks to early adopter success stories. As the demand for AI continues to grow, ask yourself the right questions to break through the noise and find use cases that justify the hype.
