Gen AI’s ultimate potential? Hive mind teamwork
LLMs will enable a shared consciousness without those limits:
Knowledge sharing. Unlike a static database, LLMs function more like a brain. They get smarter as more information is fed to them. They can also use that information to improve themselves.
LLMs will enable a shared consciousness without those limits:
Knowledge sharing. Unlike a static database, LLMs function more like a brain. They get smarter as more information is fed to them. They can also use that information to improve themselves. Not only can anyone access exactly the information they need anytime and from anywhere, but the LLM will also be able to provide plain-language clarification and suggest additional resources in nanoseconds.
Scalability. LLMs can accommodate growing teams without limiting anyone’s ability to participate, seek assistance, or provide input.
Updates. LLMs can mine the queries and clarifications they receive from users to absorb and distribute the most current information.
Toward the ‘hive mind’
How might this notion of LLM-enabled shared organizational consciousness look in practice?
Imagine it’s 2028. You’re reviewing a message before it’s broadcast to the organization. A bot appears on your screen and says, “Carol just suggested the VPs each get this message directly from you. That seems wise, given the timeline. Do you want me to write those up and personalize them?” “Yes,” you say.
The messages appear, and your bot announces that Nick is requesting a video call. “Would you like me to take that call for you?” the bot asks. “Please,” you sigh in relief.
While you review the messages, your bot takes the call and returns with another suggestion. “Nick has fresh numbers on the Henderson deal. Would you like me to incorporate them into the VP messages and alert Carol?” “Yes.”
There was a time when this scenario would have seemed far-fetched. Not anymore.
Like JSOC did years ago, I believe most organizations will soon find themselves up against “hive minds” — competitors using LLMs to think and act as one at incredible speeds.
Fortunately, we were able to catch up to our adversaries, but I think organizations today have only two options: Get in front of this power curve now or become irrelevant. LLMs are improving too fast for anything in between.
Take action
Here are three actions for organizational leaders to consider as they navigate the potential of LLMs.
1. Test the limits of LLMs: It’s hard to imagine the benefits (or threats) of LLMs if you don’t know what they can do. Find ways to incorporate them into your work and that of your teams to get as many perspectives as you can. Build prototypes to explore opportunities around messaging, calendar management, coordination of tasks, and more. (I know a CFO and team who build one chatbot every week.)
2. Embed transparency everywhere: The technology won’t be enough. We learned that the hard way. In fact, the key was transparency, which we would not have been able to maintain without relentless discipline. As powerful as they are, LLMs will still be at the mercy of your teams’ willingness to share their information. If that sounds like a tall order for your organization, a cultural change is likely needed, and it must be led from the top.
3. Develop LLM guidelines: LLMs and the AI that powers them are not perfect. Concerns such as data bias, privacy, the accuracy of output, and what to do about them continue to be debated. As governments zero in on how to regulate this technology, set boundaries on how your organization’s data can be utilized. Know how to decommission projects and delete data if necessary. Establish protocols for ethical use and monitoring, and be prepared to update them as regulations evolve.
As LLMs continue to improve, organizational leaders face a critical decision: Embrace this transformative technology and stay ahead of the curve, or risk becoming irrelevant. There is no reason to think there will be a third option.
Chris Fussell’s colleague Will Smith also contributed to this article.
About Author
