Briefing Knowledge Leaders 2024

Briefing Knowledge Leaders 2024

This year’s Briefing Knowledge Leaders conference was, again, a very engaging and informative day, leaving delegates with plenty to think about.

The keynote speaker (Beth King) on “Creating a Knowledge First Culture”, reminded us that culture is not a thing that can be imposed, but it is an ever-evolving system of complex actions, beliefs and behaviours. We were encouraged to embrace complexity rather than trying to simplify things in order to bring people along, and not to focus on changing what people do, but to address what they think, feel and need. We were given a very vivid illustration of how lawyers’ default mode is competitive and therefore that they may be inclined to see their personal knowledge as equating to power or personal advantage.

We were reminded of the value of storytelling and providing opportunities for team members to share experiences. The time spent on this is more than repaid by the productivity gains which can be reaped from individuals being able to learn from each other’s experiences. Furthermore, storytelling helps to build trust, which in turn means individuals are more likely to admit mistakes and ask for help from their colleagues. There was a comment that the fact that the word “network” includes the word “work” felt like an important reminder when busy knowledge leaders are struggling to prioritise building time into the diary for relationship building chats.

There was inevitably much discussion of AI and the best approach for firms to adopt, particularly when legal work depends so much on quality, veracity, and customer trust. Investing in the hard work to get the firm’s data into good order using controlled vocabularies and knowledge graphs will minimise hallucinations. This can seem like a huge task, but we were encouraged that even taking some steps in the right direction will help.

Many firms are experimenting with AI and realising that the learning curve to be able to use AI tools is steep. Law firms don’t always have an experimental mindset! Knowledge teams often have an advantage here: for example, the skills required to perform complex searches are similar to those required to construct effective prompts. It was suggested that implementing AI had parallels in the introduction of PCs to legal practice when, back then, not everyone could type – it was hard to justify it in terms of concrete ROI, but it became necessary.

However, with new AI products coming on stream in large numbers and very quickly we were reminded of the importance of understanding what lies behind each product. There may be several layers beyond the product developer; the underlying LLM, the computing resources to train and run the model, the data used to train it, hosting, etc.

The instances of AI getting it very wrong and leading people into embarrassing situations, both inside and outside the law, illustrate the importance of users understanding what the tools do and their limitations. Rather than deferring to the answers given to us by AI as if they were magic, we need to treat them as potentially incorrect – like the output from a novice colleague. Finding use cases which don’t depend on 100% accuracy (such as an internal meeting transcript or a first draft of a document may be important to allow experiments with the technology to bear fruit. Ultimately, nobody has “cracked” AI so it was reassuring for participants to feel they aren’t out of the game if they haven’t rolled out tools across the board for client-facing work yet. Taking a measured approach has a lot to be said for it.

If you would like further information about 3Kites Knowledge Management services or any of the other 3Kites’ services, please contact us using the following details: