Artificial Intelligence - Hype or Hope?

 

In my work life earlier this year, we did a pilot program using Microsoft's OpenAI implementation to create a chatbot for internal usage. The rollout went well, although the adoption was not what we had hoped for. Was it how we presented it to users, or was it we were so caught up in the tech "wow" factor that we missed something? Maybe we were trying to solve a problem that didn't exist for anyone else?

Since then, it seems like AI continues to get more and more press. Is it the next big thing, or overblown hype? Every year or so, some new tech is launched and the IT world stands agog at how THIS will be a game changer. Remember how Docker containers were going to be how all DevOps were created and it would replace the existing server farms that a lot of companies used? Or how Big Data with Hadoop would revolutionize how analytics were developed? 

As a follow up - I experimented with Microsoft's Copilot technology across the three tiers they offer, CoPilot Pro (for home users), Copilot for Microsoft 365, and Copilot free. I won't delve into the distinctions between the options here, but if you are interested, here is a link providing more information.

I worked with Copilot on the Office suite with the Pro and Microsoft 365 tiers on two different PCs, for my testing. Using Word, I was able to generate a one page document using a simple prompt that provided the factual information I requested. The supplied information was accurate, and I was able to use additional prompts for more detail.  A follow-up test to generate a 3000 work piece of fiction was successful, if a bit bland. 

Excel tests were along the same lines, but when I showed an Excel power user at work some of the features, she was largely unimpressed. Being familiar with the application, she could generate the insights herself quicker that typing in the Copilot prompts.

Power Point was an interesting test. I first asked Copilot to create a presentation listing the top reasons to move to Florida. Within a minute, I had an eight slide file, covering many of the reasons people cite for making the move. As a second test, I asked for a presentation discussing the top reasons for relocating a corporate office from Chicago to Dallas. Copilot generated pros and cons, and an additional section on how to mitigate the cons. 

Finally, I had Copilot use a Word document that I had created with an outline for a presentation and create a Power Point file. The finished product was very well done, highlighting my main ideas, and moving a lot of the content to the Notes section of the slides. Appropriate graphics and images were also added by Copilot.

For Outlook, I asked Copilot to draft an email for a staff lunch and learn. The results were functional, but again, could have been handled manually with almost the same effort involved. Where it shined was taking a long email chain, and distilling it down into the main areas of discussion through its summarize feature.

So at the end of the day, the question to be answered is whether AI is all hype or a something else? Working with the tech for the short duration that I have, I am intrigued. I think there are is some useful aspects to AI in the state its in now. Its not going to take over the world in this form, but as a useful tool for some it could he very handy. 

and while the tech is interesting, I feel like its not ready for prime time.