One more newsletter issue for the Perspectives Plus 💎 subscribers before 2024 is done. 🍾 

This isn’t a “my top X highlights” type of a post. In the end, the years of our lives are not about achievements. They are about experiences. So, today I’m going to first reflect on how I experienced the past year of living with Microsoft products and technologies. Then, I’ll share a few details about the personal experiences in my professional life and how they shaped me into a different man than I was one year ago.

Microsoft’s 2024

The least surprising element of 2024 was how everything revolved around AI. The script for this was written well in advance, given how the CEO declared Microsoft to be the Copilot Company at Ignite 2023. The only thing that remained to be seen was how the market would accept this all-in strategy.

Internally, the plans for how to execute such a strategy weren’t all that clear. The number of tweaks in product names and features around Copilot were classic Microsoft. I lost count of how many “simplifications” there were during the year, as I was kept busy making meme pics of the latest announcements. Ending 2024 with a Microsoft 365 icon update and a bunch of name tweaks I can’t recall without opening the M365 Message Center post is just a perfect way to celebrate the year of “something something Copilot”.

The reason it’s easy for Microsoft’s product marketing to get away with it all is that everyone else is scrambling to find their AI story, too. Whether you looked at Google, Amazon, Salesforce, Apple - no one in big tech seemed to have their shit together in 2024. Apart from Marc Benioff’s marketing stunts that took a swipe at Copilot, we saw hardly anyone high up dare to draw comparisons between their AI products and the offering from their competitors.

What did the customers say then? To simplify it, I saw two types of behavior patterns:

1) “We’re serious about transforming our business with AI, so we bought M365 Copilot licenses for everyone”. The top-down management pressure to do something (anything) with AI leading to embracing tools first and then figuring out how to apply them in business context.

2) “Our strictly limited PoC’s showed that the technology isn’t mature enough to meet our high bar for output quality and we are therefore putting everything on hold.” Approaching AI as a readymade solution that should replace existing practices. Thus, setting it up for failure when the nondeterministic nature of LLM based tools differ from all the prior technologies deployed.

Rather than addressing the gaps between technical capabilities and expectations (which we can blame their marketing departments for), the vendors seemed to collectively agree in 2024 that the best way out from this looming credibility crisis is to shift the goal posts further out. As a result, everyone set their targets on building AI agents instead of AI assistants. Or mostly, to talk about agents without really deploying any of them in the field of operations.

What this meant is 2024 gave us far too few tangible and measurable results from AI. Even though the products are out there, many in GA, and you have to pay real license fees for it - almost all of the AI talk is still about the future. What will eventually be possible. Not what already is happening today.

I find this sad and frustrating, to be honest. We all know that Microsoft is about so much more than just this recent layer of GenAI frosting on top. I’m quite confident that these layers provide a solid platform to eventually support agentic AI solutions in the business domain. Maybe Meta, Google or someone else will grab the consumer AI market. Still, under Nadella’s lead, Microsoft is positioned to take AI applications into mainstream for business customers - once they become reality.

Subscribe to Plus to read the rest.