AI takes over SaaS, but changing the way we work will bring real disruption!

(This article is written by me with no AI help except at one place where ChatGPT is credited. The art is generated using ChatGPT with additional editing by me!)

In these last few days of an extraordinary year from a tech perspective, Microsoft announced two very important things: one, it renamed the Microsoft 365 App to Microsoft 365 Copilot, and two, Satya Nadella spoke about the potential end of the SaaS era taken over by AI and AI agents. You can refer to these two developments here:

These developments had me and many others thinking: is Microsoft, a firm that has defined how business software and, in many ways, businesses work for many decades, really going to disrupt itself? I am a bit late in analyzing what we heard at Microsoft Ignite, but the keynote clearly highlighted that Microsoft is banking upon our current ways of working. So it's surprising coming from Satya when he calls SaaS just performing CRUD on top of a database. At Ignite, we heard about multiple uses of the Copilot via Pages, Word, PowerPoint, Excel, etc., and today, they all have become AI Agents! Microsoft is banking upon the phrase "UI for AI" for Copilot. But will this just be limited to Microsoft, or can it extend to other SaaS firms as well?

SaaS (and Microsoft has perhaps the biggest chunk in it) applications have been a center point for business operations for many years. Software has been a way to solve complex business problems by understanding client requirements and building solutions to address them. The notion that developing AI would require less SaaS (and therefore a smaller number of developers) may sound theoretically plausible. Still, practically, developers will be needed, and their skills and roles will evolve. You can not go to a client today and say, "You don't need this SaaS app anymore; I can give you an AI agent instead that can solve all your problems. But hey, by the way, I can only guarantee 80% accuracy". Which client would like to invest in such a value proposition? AI still hallucinates and is still not there where it needs to be. So, there will be a need for developers in their current ways of working for quite some time. But it will change, and developers will further up their skills to address problems and things that AI can't do.

At this point, it is also important to reiterate what I have already written about. It will be important to assign a reliability score with every AI output: Something like, 1 for completely AI generated output, 2 for mostly AI and little human efforts, 3 for part human and part AI, 4 for mostly human with little AI help and 5, completely human. Such an assignment should be incorporated while developing modern AI-powered applications and the developers can add value here to introduce business and industry vertical expertise to handle exceptions that may be at a risk for level 1 or 2 reliability score. So the roles of developers will change and will require more business and industry vertical skills than ever before.

From a SaaS application's perspective, "UI for AI" is a very intriguing concept. Every SaaS firm, big or small, has already developed AI capability or agent within its interface. Do the users of such applications need a dashboard where AI is an option, or would they rather talk to just one interface, the AI, through voice or chat and let it execute every other action? Seems possible. We should remember that the generation entering the workforce in the coming years will be more comfortable with a single and simple UI to do multiple things rather than navigating through multiple modules or dashboards. Knowing a particular SaaS application or how it works may not be needed, only the way to interact with its AI-enabled UI.

However, the most important aspect of contemplating these possibilities is considering how we can change our ways of working. Are we looking at AI for mere efficiency improvement or real innovation? I mean, if I have to use tech as powerful as AI to create a 10-page document or 30-slide presentation quickly and send it to someone who would use their own AI agent to summarize it anyway, then it is neither improved efficiency nor innovation. The real potential of AI is not in improving our presentations, creating or summarizing documents, creating charts, or generating reports. It is about focusing on what we do once we have all those spreadsheets, presentation decks, etc. Unfortunately, too much focus is on efficiency improvements in what we do today rather than envisioning what we can do tomorrow when AI will do all the mundane tasks.

Maybe we can have HR teams that no longer create policies or answer queries but focus on the emotional connection of employees and show more human-like empathy? The sales teams can focus on more human-connect and building customer communities rather than focusing on sales-related document creation. We can have finance teams that focus on underexplored and unconventional financial opportunities. Or the future enterprise may not even need these distinct business functions at all!! According to ChatGPT, Humans excel over AI in critical thinking, creativity, emotional intelligence and ethics. (I am sure there are many other areas!). Our future ways of working should revolve around aspects beyond AI's scope rather than those that improve our efficiency using AI.

These are exciting times indeed, and there are ample opportunities for both AI tech and humans to evolve. This may be the end of SaaS in the traditional sense, but business applications will be needed, perhaps with only a powerful UI. The real disruption will happen when all these changes actually transform the way we work today.

Thanks for reading. Please let me know your thoughts on this!

-M