The PMO and Artificial Intelligence, AI, weekly post #07

This is starting out to grow into a nice set of posts wouldn’t you say? If you want to find all of them, you can go to the category section and read up on other articles in the serries. Seven, well, eight weeks into the articles Microsoft is still working tirelessly to get Co-Pilot solutions into every application of its portfolio.

A project manager, chasing his next deadline, in the style of Vincent van Gogh.
A project manager, chasing his next deadline, in the style of Vincent van Gogh.

Will Artificial Intelligence fix work?

A very important and interesting study was presented recently. You can view the whole document here and if you are still curious about why AI is “such a big deal”, this might shed some light on that matter.

With close to 60% of our work time stuck in communication, and the remaining 40% in creative work. We might be curious to know what will happen when communcation is greatly improved thanks to Generative AI assistants.

We have all heard the arguments stemming from fear that AI will cost us our jobs. It’s interesting to read in the article that 70% of the respondents would offload as much work as possible to the AI to lessen their workload, and I can imagine that this is mostly mundane workload that is offloaded to AI.

Again, the study is very interesting to read, so please do take some time to dive into this easy to read and interesting article.

Co-Pilots vs Agents

In Artificial Intelligence articles I start to read more and more about two different approaches to AI. One approach is the route Microsoft seems to be taking: The route of an assistant AI, where the end user retains control over the end product and autonomy remains with the user. This is what Microsoft dubbed the co-pilot, which I think is a nice and appropriate name.

The other approach is one that leaves autonomy in the hands of the AI. These programs are called agents, yes, like Agent Smith in the Matrix. These AI’s will be given a prompt (AI speak for command) and will go out and try to fulfill the assignment to the best of their abilities. Some of these agents will have the capability to enforce other agent AI’s, just like ChaosGPT did when trying to bring down humanity.

Chaos GPT and it’s plans for world domination. Spoiler: it’s not a success.

As for my take on the matter. I would like to keep some form of autonomy, and therefore I’m rooting for the Co-Pilot camp of AI’s.

If Digital Dept is really a thing as stated in the previously mentioned article. Then it would be great to have an option to decrease that dept and work more efficiently and with an increase in focus time.

Artificial Intelligence and plug-ins

Just imagine a Large Language Model (LLM) like ChatGPT, it’s trained on so much data but in a very broad sense. To give the AI more depth it would need to be trained on specific data. Maybe the stock markets, maybe a certain sport, maybe project management?

This is where ChatGPT plus users have an edge, with the plus version of ChatGPT you get access to plug-ins. How I understand them is that these additions to the general model have a specific use case. More information can be found here.

Final notes

For our little piece of the pie, project management and all it’s associated managements together are currently still much to diverse and complex for an AI to fully control autonomously.

Let’s rejoice in all the additional focus time the AI’s will provide us with, so that we can be our best selves and deliver great projects!

Two interesting sources I use to get up to date with the AI changes (apart from Microsoft’s blogs):

Last thing: do you like this weekly blog so far? Is there anything I should change? Reach out through my newsletter. Or connect on LinkedIn 😁.