When the original Vals Legal AI Report (VLAIR) was released, legal pundits, thought leaders, and bloggers almost universally declared that AI was now ‘better than lawyers’.  We have been working on the following article for the past six months because we did not accept the ‘better than humans’ claim at face value.

Brandolini’s Law states that it takes a lot more time and effort to debunk hyperbolic claims than it does to make them in the first place.  So, part of the reason it has taken so long is that we have day jobs making our own products and services better, but part of the reason is that careful evaluation takes time and effort. It demands wrestling with nuance:  appraising, weighing, and valuing both the parts and the whole.  The angle that we share here doesn’t make for quick, attention-grabbing, click-bait social media shares. Instead, it should matter to anyone who wants to think (and act!) more deeply on the technology we are all collectively building and deploying.

Coincidentally, right as we were getting ready to publish this article, Part Two of the VLAIR study was released.  And almost as if on cue, the LinkedIn posts proclaiming that GenAI is ‘better than humans’ started flooding in.  We may or may not have a second thought piece six months from now, but we wanted to include this whitepaper to encourage the industry to think in a more careful, rational, reasoned, measured, and thorough way rather than jumping to conclusions.