The "Power Struggle" at the Desk: AI Is Quietly Stealing Your Mouse
Hello everyone, I'm your financial journalist. Recently, there's been a particularly lively and somewhat intense phenomenon in the tech world: four internet giants—Tencent, Alibaba, ByteDance, and Baidu—have almost simultaneously launched their own AI-powered office products.
Tencent has WorkBuddy, Alibaba has Qianwen Office, ByteDance has DouBao Work, and Baidu has Baidu Dazǐ. At first glance, you might think, "Oh, just another battle among the big companies—whom will win, and whom will lose?"
But if you only focus on things like "whose PPTs look better" or "whose AI models are smarter," you're missing the bigger picture. As an observer, what I see is a much deeper and more brutal "battle for control of the workplace entry points."
Today, I'll break down the logic behind this in simple terms, covering five key aspects, to show you how this seemingly calm storm will change the way we work over the next decade.
---
1. Core Summary: The Shift from "Buying Software" to "Buying Results"
In one sentence: In the past, we used software to perform tasks; in the future, we'll give instructions, and AI will handle the software for us.
Previously, if you wanted to create a report, you had to open Excel to calculate data, then Word to write the text, and finally PowerPoint to format it. You were using the tools themselves. Now, you just need to tell the AI, "Analyze last month's sales data, identify the three products with the fastest growth, and create a PPT for tomorrow's report." You're buying the results.
The essence of this competition isn't about which AI product is the best, but about their combined efforts (although they also compete with each other) to challenge the "office infrastructure" that has been built over the past few decades by Microsoft Office, WPS, and countless SaaS solutions (such as CRM and project management tools). What they're after is the "entry point" to the workplace—the first thing you see when you turn on your computer every day.
---
2. Explanation 1: The Entry Point Has Changed, and Your First Action Has Been Taken Over
**Title: You No Longer Open Excel; You Open Your "Brain"
To understand this change, we need to define what an "entry point" is. For the past twenty to thirty years, the entry point to the workplace was the software itself:
- Write a proposal? Open Word.
- Calculate numbers? Open Excel.
- Create a report? Open PowerPoint.
- Manage customers? Open a CRM system.
Each task corresponded to a separate software window, and you had to switch between them like a skilled multitasker.
With the advent of AI agents, the logic has reversed. You don't need to know which software to use for data analysis or charting; you just need a unified interface (like WorkBuddy or DouBao Work) and input your requirements.
Previous process: User -> Open Excel -> Enter formulas -> Get data -> Copy -> Open PowerPoint -> Paste -> Format.
New process: User -> Tell the AI to create a PPT -> AI automatically finds data, calculates, generates charts, and formats it -> User -> Receives the file.
The key point is: The first thing users open is no longer Excel or WPS, but the AI.
This is why Tencent, Alibaba, ByteDance, and Baidu are so eager to get involved. If users get used to asking the AI for help first, the AI becomes the new "operating system" at the entry level. Office and WPS haven't disappeared, but they've shifted from being the main players to being mere "executors" for the AI.
It's like how you used to hail a taxi: you had to find the car, negotiate the fare, and drive it yourself. Now, you just open a platform like Didi, and it handles everything for you. You don't care whose car it is; you just care if it gets you there.
---
3. Explanation 2: Office and WPS Aren't Gone, But Their Status Has Declined
Title: From "Main Players" to "Tools," the Embarrassing Situation for Traditional Giants
Many people worry that AI will make Word and Excel obsolete. My prediction is that they won't disappear in the short term, but their status will significantly decline.
Why? Because Word, Excel, and PowerPoint have decades of development behind them: standard file formats, complex editing logic, enterprise procurement systems, and user habits. Excel, in particular, has so many advanced features that AI can't yet fully replace.
However, their role has changed:
- In the past: Office was the entry point, and AI was just a feature within Office (like the "smart assistant" in Word).
- In the future: AI is the entry point, and Office becomes one of the capabilities that AI uses.
Imagine this scenario: You tell the AI, "Create a report for next quarter based on yesterday's meeting notes and the latest business data." What does the AI do behind the scenes? It might use Excel to process data, Word to generate text, and PowerPoint to format it. You never see the Excel or Word interfaces; you just see the AI working.
For Microsoft and WPS, this means they can't rely solely on selling software licenses or simple AI add-ons. They need to become essential components of the AI ecosystem.
Microsoft has already realized this and integrated its Copilot AI assistant directly into Word and Excel, even allowing it to perform multiple tasks independently. WPS is doing the same, embedding AI deeply into its documents, spreadsheets, and PDFs.
Conclusion: Traditional office software won't disappear, but they'll become "underlying services" rather than the primary tools. Users won't interact directly with them; they'll interact with the AI.
---
4. Explanation 3: SaaS Software Is in the Most Dangerous Position, Becoming "APIs"
Title: Your CRM and Project Management Tools May Become Invisible Data Sources
If Office and WPS are just losing their prominence, then SaaS software (Software as a Service, such as CRM systems, project management tools, knowledge bases, and marketing automation platforms) is in even more trouble.
The SaaS industry thrived because each business task could be turned into a separate software:
- Manage customers? Buy Salesforce or FenXiangXiaoKe.
- Manage projects? Buy Jira or FeiShu.
- Manage knowledge? Buy Confluence or YuQue.
As a result, companies used to have many software systems, with employees switching between them daily to enter data, check status, and write notes.
With AI agents, this logic is changing. If AI can directly handle these tasks, why bother with the software itself?
For example:
- Previously: Salespeople logged into CRM daily to check customer lists, manually update records, and generate reports.
- In the future: They tell the AI, "Find the customers I followed today but haven't responded, sort them by interest level, and generate a follow-up plan." The AI handles all the work behind the scenes.
Is the CRM still there? Yes. The database? Yes. The workflow? Yes. But the user interface has changed. SaaS software is becoming an API, a data source, or a tool that AI can use.
It's like how you don't care which database (MySQL or Oracle) a website uses. In the future, employees won't care whether the AI uses Salesforce, DingTalk, or another platform; they'll only care if the task is completed.
This is a huge challenge for SaaS providers. If they can't become standardized tools that AI can easily use, they'll be marginalized, just like functional phones were marginalized by smartphones.
---
5. Explanation 4: What Are the Giants Really After? The Trillion-Dollar Value of the "Super Entry Point"
Title: Whoever Controls the First Task of the Day Controls the Data Gold Mine
Why are Tencent, Alibaba, ByteDance, and Baidu all rushing into this? Because AI-powered office could be one of the few areas where large models have a real chance to become a "super entry point."
Why is the workplace such a strategic area?
1. High frequency: It's used for 8 hours a day.
2. Complexity: Work involves a series of tasks (data analysis, document writing, emailing, meetings, follow-ups).
3. Fit: AI agents are great at understanding goals, breaking down tasks, and using tools to complete them.
If an AI becomes the entry point to your work, it gains a lot more than just a chat partner. It connects to your files, communications, business systems, and knowledge bases.
This is why companies are investing heavily in connectors, MCPs (Model Context Protocols), skill packs, enterprise knowledge bases, and computer operations.
What's really valuable isn't just "writing a document for you"; it's "whom you open first when you go to work."
In the past, this position belonged to Windows, Office, WPS, and browsers. Now, the giants want to add another layer on top of the desktop: the AI agent.
If users get used to asking the AI for help first, this AI will become a powerful tool. It can sell software and control the flow of corporate data, becoming the central hub of enterprise information.
---
6. Future Outlook: Three Key Indicators to Watch
Title: Don't Just Focus on PPT Quality; Look at These Three Things
If you want to predict which AI product will be successful or how the industry will change, don't just judge by the quality of PPTs or the size of their models. Focus on these three indicators:
1. Connectivity: Which AI can connect to the most office software and business systems? Can it directly use your local files and systems?
- The more connections, the more tasks it can handle, and the more loyal users it will be.
2. Task Completion: Which AI can complete complex tasks from start to finish? Many AI can generate content but not execute it. For example, it can write an email draft but not send it or insert charts into a PPT.
- Completing tasks from generation to execution is the dividing line between good and great AI.
3. User Habits: Which AI can make it easy for users to ask for help? This requires a low learning curve and high success rates. If the AI often fails or is slow, users will return to traditional software.
Final Prediction:
In the future, the software we're familiar with today (Word, Excel, WPS, SaaS) may not disappear, but the primary tool users will change. The value chain of the office industry will rearrange:
- First layer (entry point): AI agents (Tencent, Alibaba, ByteDance, Baidu, Microsoft Copilot, etc.).
- Second layer (capabilities): Traditional office software (Office, WPS) and SaaS systems (CRM, ERP), becoming tools for the AI.
- Third layer (users): You just give instructions and enjoy the results.
This battle has just begun.