Microsoft CEO Satya Nadella Announces End of App Era, New Direction for Software Development

I think the concept of business apps may disappear in the age of intelligent agents.

Microsoft CEO Satya Nadella says apps as we know them are disappearing and being replaced by intelligent agents.

It was a big statement, and he actually set a flag for the future direction of software development.

This means no more apps, no more SaaS (Software as a Service), and probably a significant reduction in the number of developers needed.

The nature of SaaS, pointing out that they are just a layer of user interface and business logic on top of the database.

Regarding SaaS applications or business applications, let me talk about our own Dynamics product.

The approach we've taken is that I think the concept of business apps may disappear in the age of the intelligent agent.

Because if you think about it, they are essentially CRUD databases with a bunch of business logic.

What he's really saying is that SaaS (Software as a Service) is essentially just a thin layer of user interface on top of a database.

When he says CRUD, he means create, read, update, and delete, which are the basic units of interaction between the interface and the database.

So create is to create new records, read is to extract information from the database, update is to modify existing records, delete is to delete.

What he means is that there will be the underlying database, which is the basic data you need, and then there will be intelligent agents that interact directly with the database.

Explain how AI Agent replaces traditional business logic and interacts directly with the database to accomplish tasks.

If you've been following this channel, you know I've been talking about this for a long time.

I really believe the entire app stack is going away.

When you have AI interacting directly with the core underlying data in the database, you don't really need it anymore.

Again, what does this mean for the SaaS industry as a whole?

What does it mean for application developers?

I'm not entirely sure, but I have a feeling it's going to look a lot different than it did today.

I've been in the SaaS industry my entire career, I've built multiple SaaS companies and worked for SaaS companies.

If I'm honest, I probably wouldn't start another SaaS company right now, and I probably wouldn't invest in any SaaS companies.

The situation is too uncertain. That's not to say that there won't be software companies in the future, just that it will look very different.

Business logic will be transferred to the AI layer, and the AI Agent will become the core of multi-database operations.

All business logic is transferred to these intelligent agents, which perform multi-database CRUD operations.

So they won't distinguish what the backend is, they will update multiple databases and all the logic will be in the AI layer.

When he talks about business logic, he's basically talking about actually doing something with the core data in the database.

Whether it's updating a CRM record or sending an email based on different criteria in the database, it's easy to use.

Or any piece of software that sits on top of a database.

All of this business logic will be compressed into the Agent, which will be able to do anything, you just tell it what you want.

How AI Agents can accomplish complex tasks without the need for traditional hard-coded business logic.

Give me a chart of the top 5 clients by revenue.

Okay, now write e-mails to all five of these clients telling them I want to do business with them in 2025.

All of this will be done automatically, whereas before it required hard-coded business logic on top of the database.

Now that this is no longer necessary, you simply tell the agent what you want and it will either know how to accomplish it, or it will know how to do it, or it will know how to do it.

Either will write the tools for you to do it.

When I say tool, I just mean that it writes its own code to interact with databases, send email, etc.

Importance of database selection

He made it clear that the back end wasn't important.

Agents don't specialize in a particular type of database, they really don't care.

So it's really about building a database that works most efficiently with the agent, is cheap, and is best suited to the particular use case.

As an engineer or data provider, you really just need to figure out what the best database to use is, and then Agent can ignore that compatibility and just use it, without caring about the specifics.

So it will be interesting to see which database the company chooses to use.

Nadella responds to the need for tools like Excel, looks to the future of AI-powered data analytics

This means that the business logic layer can be orchestrated by AI and AI agents.

In other words, it should be very seamless from Copilot to the agent to my business applications.

Now, by the same token, you could even say, why do I need Excel?

(This sharp question came from one of the hosts... so bold... haha)

Interestingly, one of the most exciting things for me is Excel with Python, like GitHub with Copilot.

That's basically what we did.

When you have an Excel like that, which by the way would be fun for you guys, you should open up Excel, open up Copilot, and start playing around with it.

Because it's no longer just, you know, it's like having a data analyst.

So it's no longer just about understanding the numerical type of data you have, it's about making plans for you.

It will be like the GitHub Copilot workspace that creates the plan and then executes the plan.

It's like a data analyst using Excel as an R-column visualization tool for analysis.

Microsoft Copilot By using Excel as a tool, human users are freed up to use proprietary tools and focus on interacting with Copilot.

Microsoft 365 transformation, Copilot will become the core organizational layer, each application becomes a special Agent

Why do you even need Excel, he may have considered this question a lot, because Excel is actually what?

It's a thin user interface layer on top of the database, but of course you can add a lot of business logic and analytics.

But the question is, if AI can write Python code to analyze data, why do you need Excel?

If I need to analyze a lot of data, I just say what I need and I just describe the end product I need.

I need a chart, I need a graph, I need a 3D visualization, whatever it is, you just describe the end result, you have the core data, the raw data, and then the agent writes the Python code that actually extracts the right data and puts it into the right format that you're asking for.

So it has tools, so Copilot is using Excel as a tool that has all the maneuvering space as it can be generated, and it has the Python interpreter, which actually does.

By the way, when he says Python interpreter, that just means the ability to execute Python code.

In fact, it's a great way to reconceptualize Excel, at some point you can say, hey, I'm going to generate all of Excel, and that's true, after all, there's a code interpreter, so you can generate anything.

So yes, I think there will be upheaval, but the way we're dealing with M365 at least is.

The first step was to build Copilot as an organizational layer user interface for AI, acquiring all the agents, including our own Agent, the

You could say that Excel is a proxy for my Copilot, and Word is a proxy for that.

They're a kind of specialized canvas, like I'm making a legal document, and they're a kind of a canvas for me to work on.

Let me put it into a page and then go to Word, then let Copilot process it, go to Excel and then let Copilot process it.

Impact of this change on the software industry as a whole

This is a fundamental change in the way Microsoft applications work, and indeed in any software-as-a-service business.

It's a complete reorganization of the entire industry, and for someone like me who's been in SaaS for a long time, it's so interesting to see it all happening.

Companies can either adopt it and evolve with the times, or they will die out, there is no middle way.

Obviously, Microsoft's Satya is thinking ahead and starting to make long-term plans, just like when Microsoft transitioned to cloud computing.

He has invested heavily in OpenAI, with an eye toward AI PCs, an AI Agent Store, and a host of other ambitious plans.

And Microsoft occupies the huge traffic entrance of the Windows desktop system, and the curtain of AI-enabled strategy for enterprises is slowly opened.

There is a huge amount of cake in this.

Original video address: https://www.youtube.com/watch?v=9NtsnzRFJ_o