-->

KMWorld 2024 Is Nov. 18-21 in Washington, DC. Register now for Super Early Bird Savings!

The trust problem with GenAI

Article Featured Image

The irony here is that GenAI should never have been trusted in the first place. Tech, as such, has no moral or ethical core; it does not know the difference between good and evil, right and wrong. Regardless of how “intelligent” it is marketed to be, it is not intelligent. Instead, it is a machine. If there is a second irony here, I am a strong proponent of AI in the enterprise. Over the years, I have been a fanboy and enthusiastic promoter of using AI for information and knowledge management, but in a practical, not Machiavellian, way. This is because I understand that AI does some things well, and other things not so much.

Using GenAI effectively

If we can’t trust GenAI, and we know for a fact that it is regularly mistaken yet convincingly plausible, how can we use it effectively? Indeed, why should we use it at all? Well, the first thing to note is that it should be used sparingly, and the second is to ensure that human intelligence plays a significant role in its deployment and ongoing usage. Maybe this is implicit in the fact that Microsoft (for example) calls its GenAI system Copilot, not Autopilot; it’s there to work with you, to assist, not replace, you. Additionally, we are seeing tremendous interest and investment in highly specialized large language models (LLMs) that power GenAI. In other words, LLMs that are not generalists and do not need to guess and invent answers. Instead, they are LLMs trained solely on highly accurate and specific information.

Yet, worryingly, currently unanswered and often unasked questions remain: Who specifically will assist the AI? Who will curate, build, and maintain these specialized LLMs? Who will check their output for accuracy and truth? As a reader of this magazine, indeed, this column, you already know the answer is information managers and KM professionals. But who else knows this? “Not enough people” is the most generous answer I can give. Nine major tech conferences, thousands of attendees, billions of dollars of tech on display, and terms like “information management” and “KM professionals” were barely ever used by attendees or technology vendors. And that is a serious problem. Worse, it’ s a disaster waiting to happen.

Trust is central to relationships. We can accept an honest mistake or two, but when we come to distrust someone, it can be nearly impossible to rebuild that trust. Furthermore, it’s a two-lane highway: We do not trust what they say, but we also won’t share with them anything that is important to us. That is just as true for our relationships with technology as with people. One of the key lessons here, I believe, is that we must be dispassionate about technology and see it for what it is—a tool to be used by us and not against us. Similarly, we need to learn that, just like our fellow human travelers, technology has to earn our trust; we should never be in awe of and trust it unquestioningly. 2023 has been the year of ultra-hyping GenAI, and who is paying for this deluge of marketing? Technology vendors that want us to buy it. Again, it’s impressive stuff, but when we shift from selling to buying and ultimately using it, many tough questions need to be asked.

KMWorld Covers
Free
for qualified subscribers
Subscribe Now Current Issue Past Issues
KMWorld on Facebook
KMWorld on X
KMWorld on LinkedIn
KMWorld on YouTube
Best Practices
100 Companies That Matter in Knowledge Management 2020
KMWorld Trendsetting Products of 2019
Knowledge Management Webinars
Site Sponsors