When you get down to basics, information technology is about communication and keeping track of data.
Unsurprisingly, as the ability to communicate and track information increases, the more that there is to communicate and track.
In theory, more information leads to better outcomes. While better information tools can process more information, they also generate more data. Yet, humans can only process a certain amount of data in a given amount of time.
While automation can handle some of this information, it is often processed invisibly in the background.
More often than not, this information is handled properly, but there are always exceptions. There are situations that need to be dealt with on a case by case basis, and those are the types of things that can fall through the cracks.
But the biggest problem of all, is that we build systems on top of systems to solve problems that we have with our systems. There is always a gap.
When organizations just had the interoffice memo, phone calls and in person meetings, email was a great addition to the communication stack.
And that was great for a while. You could asynchronously communicate quickly and have a written record of the communication. Because it was pretty easy and convenient, inboxes got abused and overwhelmed with too many messages. As issues mounted with email, tools like Slack emerged to address that problem. These streaming information tools combined the best of both the synchronous and asynchronous communication tools. However, each new channel added more communication modes to track.
The emerging AI tools are more likely to make the problem worse instead of better. On its own AI is not going to help with communication. AI can generate content and AI can process information, but it does not necessarily improve communication. Communication is often about nuance and ideally confirming mutual and aligned understanding. While IT vendors are rushing to add AI features to their products in any way that they can, it feels, well in a word, completely artificial.
Being able to say that your product has AI capabilities, is something that has to be on the product manager’s checkbox, but is it something that is really beneficial. Certainly AI is useful. It can do all sorts of things, but how many of these things are really worth doing. If I hire an AI to do a job for me and you hire an AI to do a job for you, are either of us better off? Have your AI talk to my AI. Maybe two AIs from two different sources will come up with a better solution. But did either of those jobs need to be done in the first place?
As the barriers to communication have been reduced, the frequency and quantity of communication has increased exponentially. But the quality of communication has not necessarily kept pace. Will AI make this better? Will it increase mutual understanding or will it just be more talking at each other trying to out-clever each other.
AI will allow more information to be collected and processed, but will humans be better informed? Chances are that this will be an additional layer of processing that insulates individuals from data that can feed into their autonomy.