During a recent panel, I was presented with the following proposition: "AI generates unreliable information." As a co-founder of Uncover, which utilizes AI to automate legal work, this proposition is not unfamiliar to me. What do I think? Yes, AI can indeed generate unreliable information. But let's be honest: doesn't this also apply to humans? So why the fear?

Let's first examine how AI can generate falsehoods. AI systems, such as machine learning algorithms, are based on data. This data is used to identify patterns and correlations and make predictions based on them. If the data on which the algorithm is trained are incomplete or biased, the outcomes may be distorted. This can indeed lead to the generation of unreliable information and incorrect conclusions.

But if we look at the other side of the coin, the human factor, we see exactly the same thing. Humans also generate falsehoods. Consider situations where a lawyer fails to ask the right questions, so the data on which they base their advice is incomplete, and consequently, incorrect conclusions are drawn, or a judge is biased. What constitutes "reliable information" is also up for debate. Even about "facts," people can disagree. There are people who, based on their interpretation of facts, argue that global warming is caused by human activity, while others, based on their interpretation of the same or other facts, argue just as strongly that global warming is not caused by humans.

Unlike AI systems, humans are also subject to fatigue, emotional stress, personal beliefs that can cloud their judgment, and they need sleep. It's interesting to note here that the data on which AI systems are trained predominantly consist of data generated by humans. So, are the inaccuracies generated by AI actually generated by AI, or are they a reflection of humans?

Setting that question aside, all of this emphasizes that both AI and humans are imperfect. Both have their shortcomings and can generate unreliable information.

I think to myself: isn't that obvious? AI tools aren't magical. They're only as good as the data they're based on and how they're used. So, lawyers must remain critical of how they employ AI and the results AI systems produce, always verifying and validating them before relying on them. Lawyers must do this even when reading a case note or reviewing a junior colleague's first draft. What am I saying? People must always think critically.

The fact remains that AI systems can contain infinitely more data than the human brain and can learn from mistakes with every thumbs up and thumbs down. History shows that this isn't the case for humans. So the chance of a human making a mistake is greater than the chance of an AI system making a mistake.

So yes, like humans, AI can generate unreliable information, but by harnessing the power of human expertise and technological advancement, we can create a legal practice that is more knowledgeable, accessible, and effective!

One last observation: the negative framing of the proposition about AI generating inaccuracies during the panel is no coincidence. It reflects a deep-seated tendency among legal professionals to be guided by fear of change rather than embracing the opportunities offered by new technologies. This brings us to a crucial question: are we, as legal professionals, willing to adapt to the rapidly changing world around us, or do we continue to cling to old approaches and paradigms out of fear of the unknown? As always, new technology will prevail, even in the legal profession and among the clients of the legal profession. Hesitators will then become tomorrow's laggards. Once again, I say: embrace AI. Let's not drown in fear of AI, but welcome it as a powerful ally in the fight for justice.

‍