What is the alternative to Bad AI?
I’ve just read a LinkedIn post by a lawyer where he had noted that another lawyer had been in trouble as he had used AI in important legal work. Apparently, the lawyer using AI had inadvertently, due to AI ‘hallucinations’, cited cases which never existed and, of course, this had proven to be problematic.
As such, the moral of the story in the post was, don’t use AI. The assumption was that AI is the problem. I’d like to propose another alternative which suggests avoiding AI’s pitfalls. Or to put it more eloquently, using AI well, accurately, morally, and responsibly.
Many people are quick to assume that poor use of AI makes AI an inappropriate tool and throw out the baby with the bathwater, saying that the AI is the problem.
What if the alternative to bad AI or AI “misuse” was better practice and use of AI? What do you think?
We’ve developed our own suggested guidelines around how we think AI should be used. Please see here: AI Standards
Here’s a brief summary of some principles of AI use to achieve better practices:
Training: Get your staff to learn how to use AI tools by attending PD focused on AI.
Talk to peers, establish best practices
Seek regular review of the current uses of AI
Audit, check, and validate current AI practices in your organisation.