The summer holidays are a perfect time to dive into the pile of books I collect during the year. On the pile of this year was: "Noise: A Flaw in Human Judgment," by Daniel Kahneman. The book is about the often-overlooked concept of "noise" in decision-making.
Interesting, I figured, so I gave it a go and whilst being surrounded by the sounds of nature, the book reminded me of the cacophony that can infiltrate lawyers’ decision-making processes and our judgement, when advising on legal and tax matters.
Here is how it works
Imagine, you are a criminal defense lawyer. On Monday, the CFO of a major corporation walks through your door. He is accused of fraud and turns to you for help. He provides you with the content of his inbox and you read it. On Tuesday, another CFO of another major corporation also walks through the door. He is accused of the exact same fraud, his facts are exactly the same and he also hands you his inbox. On Wednesday, both of them receive a call from you. Your message to CFO 1: you have a great case, I will help you. Your message to CFO 2: sorry, I will not take your case.
Why did you give CFO 1 the green light and CFO 2 not? Maybe it was sunny outside when you went through CFO 1’s inbox and you were in a more optimistic mood. Maybe CFO 1 made you think of Max Verstappen and you just feel like you can win this case. Or maybe you were bored of reading once you got to the paperwork of CFO 2. Or, none of these factors were relevant. What probably influenced your decision, is noise. Noise refers to the variability or inconsistency in judgments and decisions made when presented with identical or nearly identical cases.
In addition to these unrelated and unpredictable factors that influence our decisions, there is more. When we make predictions, such as when assessing the merits of the cases of CFO 1 and CFO 2, we are easily led by what “feels right”. For example, if you wrongly take the case of CFO 1, he might get away with his crime. If you wrongly deny his case, he might lose his freedom. Weighing the consequences of the possible outcomes, you’ll call on your experience and make a decision that “feels right”.
Unfortunately, it seems that we are bad at making accurate predictions. Kahneman described a study where an algorithm was created to produce bail judgements. It appeared that even a super rudimentary formula, outperforms human judges by 24%-42% when it comes to rightly sentencing bail.
Mind the noise
So, whenever you as a lawyer go through a decision-making process, you will be influenced by noise. Noise that negatively affects your decision. This made me think, can AI help lawyers with making better decisions?
I think so. Take the example of the CFOs. Unlike me, AI would, if being tasked with extracting relevant information from the inboxes, objectively extract the relevant facts. It won’t be influenced by whether CFO 1 resembles Max Verstappen, nor will it get bored. It will remove many of the unrelated and unpredictable factors that cloud your judgement.
Some use cases where I see AI reducing noise are:
- Document review: where inconsistencies in reviewing contracts, agreements, or evidence can lead to errors. AI driven document structuring and summarising tools, can ensure uniform and accurate document analysis, reducing noise;
- Standardised legal processes: AI can facilitate the establishment and enforcement of standardised legal procedures within law firms. These processes can help ensure that similar cases are handled consistently, reducing the risk of noise in decision-making;
- Fact finding in litigation: similar to the CFO example, AI driven tools that extract fact patterns from documents can help ensure that the facts are objectively assessed and no facts are forgotten, making sure you see the full picture;
- Client Management and Communication: Noise can also emerge from inconsistent client communication and management. AI driven solutions that are open to clients can contribute to client expectations being met consistently.
As I close the pages of "Noise: A Flaw in Human Judgment," my attention drifts from the text, back to the noise that is surrounding me.
