Thursday Thoughts: Claude Code Is Lotus 1-2-3, and Copilot Is Playing the Excel Game

The world of AI-assisted software development is evolving at a breakneck pace. In 2026, we're not just seeing autocomplete on steroids; we're witnessing a fundamental shift in how code is written. Two names dominate the conversation: Claude Code from Anthropic and GitHub Copilot from Microsoft. But to truly understand their current roles, we need a historical lens. Claude Code is the Lotus 1-2-3 of this era—a pioneering, powerful tool that reshaped the landscape. Meanwhile, Copilot is playing the Excel game—a ubiquitous, integrated standard that most developers actually use. This Thursday, let's unpack this analogy, explore the concept of "vibe coding," and figure out what it means for your workflow.

The Rise of Vibe Coding

"Vibe coding" is a term that emerged in 2025 to describe a new, intuitive way of programming. Instead of writing code line by line, you describe what you want in natural language—your "vibe"—and let an AI model generate the implementation. Think of it as pair programming with a patient, hyper-fast partner who never sleeps and never gets annoyed by vague requests.

This approach has exploded in popularity. Tools like Claude Code and GitHub Copilot are at the forefront, each taking a different philosophical stance on how to integrate AI into the developer's workflow. The core promise is the same: dramatically lower the barrier to entry for software creation, and boost productivity for seasoned engineers. But the execution differs in fundamental ways, which brings us to the spreadsheet analogy.

A Tale of Two Spreadsheets: Lotus 1-2-3 and Excel

To understand the current dynamic, we need to rewind to the early 1980s. Before spreadsheets, financial modeling was done on paper or with complex mainframe programs. Then came Lotus 1-2-3 in 1983. It was a revelation—the first killer app for the IBM PC. It combined spreadsheet, database, and charting functions into one fast, powerful package. Businesses adopted it en masse; it was the industry standard for years.

But Lotus 1-2-3 had its quirks. Its command-line interface was powerful yet unforgiving. To use it well, you needed to learn complex keystrokes and slash commands. It ran best on a PC with a specific setup. Upgrading required relearning. It was the domain of spreadsheet specialists and power users.

Then came Excel. Microsoft didn't invent the spreadsheet—they copied and refined. Excel was built for a broader audience. It integrated seamlessly with Windows, the emerging GUI standard. It had a menu-driven interface that was easier to learn. It handled macros and later VBA, but also made simple tasks accessible to everyone. Critically, Microsoft bundled Excel with Office, making it the default choice for businesses. Over time, Excel won not because it was universally better, but because it was good enough and everywhere.

The rest is history. Lotus 1-2-3 faded into obscurity, while Excel became a verb. This is the lens through which we should view Claude Code and Copilot.

Claude Code: The Lotus 1-2-3 of AI Programming

Claude Code, launched by Anthropic, is a terminal-based AI agent. You type natural-language instructions directly into your command line, and it can read, write, and edit files across your entire project. It has access to your file system, can run terminal commands, and even use tools like web search or API calls. It's an autonomous coding companion, not just a suggestion engine.

What makes Claude Code the Lotus 1-2-3 of this era?

• Pioneering power: It demonstrated the possibility of a fully agentic coding experience. You can say, "Explore this codebase, understand the bug in the payment module, and fix it," and it will do that. This shifts the developer from writing code to directing an AI.
• Steep learning curve: As of 2026, Claude Code is primarily used by experienced developers. It's a command-line tool, which is intimidating to newcomers. The context window, token management, and prompting strategies require practice. It feels like the old-school spreadsheet power users—mastering slash commands vs. modern point-and-click.
• Resource intensive: It's a heavy-duty model, often requiring a subscription to Claude Pro or API access. Running it locally or via the cloud consumes significant compute.
• Niche appeal: While incredibly powerful, it's not yet the default for most programmers. It's used by early adopters, AI enthusiasts, and those building complex, context-heavy applications.

For an expert developer, Claude Code is intoxicating. It feels like driving a Formula 1 car—fast, powerful, but requires skill. However, that complexity limits its mass adoption.

GitHub Copilot: The Excel of AI Pair Programming

GitHub Copilot, developed by GitHub (a Microsoft subsidiary), takes a very different approach. It's integrated directly into IDEs like Visual Studio Code, JetBrains, and others. It works as a real-time code completion and chat assistant. Instead of delegating entire tasks, you collaborate with it line by line.

Here's how Copilot is the Excel of this analogy:

• Seamless integration: Copilot lives where developers already work—inside their editor. No separate terminal, no new workflow. You just install the extension and it's there.
• Approachable to all: A junior developer on day one can use Copilot effectively. It suggests entire functions based on your comments or code context. The learning curve is minimal.
• Ubiquitous adoption: Since its launch in 2021, Copilot has become the standard AI pair programmer for millions of developers. It's bundled into GitHub, Visual Studio, and even enterprise via GitHub Copilot Business and Enterprise plans.
• Good enough for most: For many tasks—boilerplate code, tests, simple algorithms, and even whole components—Copilot's suggestions are excellent. It may not solve complex architectural problems autonomously, but it handles the grunt work.

Excel didn't outshine Lotus 1-2-3 in raw capability in the early days; it just outmaneuvered it on distribution and usability. Copilot is doing the same—by being the default choice in the most popular development environment, it captures the majority of the market.

Side-by-Side Comparison

What This Means for Developers

Why does this analogy matter for you in 2026? Because it affects how you choose your toolkit and how you invest your learning time.

If you're a senior developer or a tech lead looking to automate complex refactoring, write entire tests, or explore a legacy codebase, Claude Code can be a force multiplier. Its agentic nature allows you to delegate meaningful chunks of work, freeing you to focus on architecture and business logic. However, adopting it means investing time in mastering human-AI interaction—prompt design, scoping tasks, and verifying outputs.

On the other hand, if you're teaching new developers, or working in a large team where consistency and low barrier to entry matter, Copilot is your Excel. It's already integrated into GitHub, code review tools, and CI/CD pipelines. It makes every developer instantly more productive, with minimal training. Indeed, for organizations that build custom learning workflows, the integration of such AI tools into internal processes is becoming critical. ASI Biont supports connecting to GitHub and other development platforms via API—learn more at asibiont.com/courses.

The real insight is that both tools have their place. Lotus 1-2-3 didn't lose because it was weak; it lost because the market wanted something easier and better integrated. Similarly, Claude Code could remain a powerful niche tool, while Copilot (or a similar IDE-integrated assistant) takes the mainstream. But the same kind of disruption could happen again: just as Google Sheets and web-based tools are nibbling at Excel's edges, we're seeing new AI-native editors, like Cursor or Windsurf, that incorporate agentic features while maintaining a friendly UI.

The Road Ahead

So, what should you bet your career on? Let's extrapolate current trends to 2026 and beyond.

• The desktop vs. web, revisited? Claude Code is terminal-based; Copilot is IDE-based. New agents like OpenAI's Codex (as of 2026, integrated into GitHub) are also going terminal-first. This suggests a trend toward more autonomous agents that can interact with your whole system.
• Hybridization. Expect Copilot to absorb agentic features—indeed, GitHub Copilot is already rolling out agent mode. And Claude Code may get more IDE integrations or cloud environments. The lines will blur.
• Democratization. Just as Excel democratized spreadsheets, these tools will continue to democratize coding. We'll see more non-programmers building software with natural language—the "vibe coding" movement will be a core skill.
• Market consolidation. In the spreadsheet wars, the winner wasn't necessarily the best technology but the one with the best ecosystem. Today, Microsoft's GitHub has a huge ecosystem advantage. Anthropic is partnering with tech giants to stay relevant, but distribution remains key.

Conclusion

The analogy is stark, and it's not a prediction that Claude Code will die. It's a warning about what happens when visionary tools fail to become accessible. Lotus 1-2-3 was legendary, but Excel became the standard because it was universal and easy. Similarly, Claude Code is pushing the boundaries of what AI can do in software engineering, but Copilot is winning the hearts and minds of the average developer.

As a developer in 2026, your best strategy is to be ambidextrous. Learn to use Copilot for your day-to-day tasks—it will make you efficient. And explore Claude Code for complex, high-leverage projects—it will make you powerful. The future belongs to those who can ride both waves: the Lotus innovation and the Excel ubiquity.

So, on this Thursday, take a moment to appreciate the history repeating itself. And ask yourself: are you building for the power users, or for the world? The answer might determine whether you end up like Lotus or like Excel.