Last month, Greg (Section’s CEO) and I sat down to explore ChatGPT's Canvas feature, which OpenAI released in October. If you missed it, I’m not surprised – its launch didn't generate the usual hype. But Canvas offers interesting insights into OpenAI's strategic direction — particularly their enterprise ambitions and vision for how we'll use AI in our work.

Here’s how Canvas works, why it’s exciting, and what it means for the AI tool landscape.

How Canvas works

Similar to Claude’s Artifacts feature, Canvas replaces the standard ChatGPT chat window with a document editor when you're working on a writing or coding task.

Here's the key feature that sets it apart: Canvas appears automatically when ChatGPT thinks you need it. If you write "Help me draft a blog post" or "Review this code", Canvas opens. If you ask for a quick fact, it stays in chat mode.

Canvas is a familiar-looking document editor, with some AI-powered extras:

When you’re writing you can:

• Highlight text in the editor and ask for specific feedback
• Use shortcuts to adjust tone or length without writing more prompts
• Track your changes like in a Google Doc

When you’re coding, ChatGPT can offer comments on your code, change code from one language to another, or even check for bugs.

The chat interface remains open on the left, so you can chat with GPT to give it higher-level directions while working in the document.

The power play: From chat to workspace

What makes Canvas significant isn't the feature itself — it's the strategy behind it. During our testing, Greg pointed out something crucial: this isn't meant to be a writing tool like Lex or a coding tool like GitHub Copilot.

It's OpenAI's play to become the default workspace where everyone does their AI-assisted work.

Think about it this way: right now, people bounce between ChatGPT and their document editor, constantly copying and pasting. Each time they do, there's a chance they'll try Jasper, Copy.ai, or Lex instead.

Canvas eliminates that risk. By creating a space where you can write, edit, and refine without leaving ChatGPT, OpenAI isn't just improving user experience – they're building a moat around their user base.

Canvas fits perfectly into OpenAI's enterprise playbook:

• Target high-value tasks: They're starting with the two most common AI use cases: writing and coding. These aren't random choices — they're the tasks enterprises are already paying multiple vendors for.
  
  
• Build "good enough" features: Canvas doesn't try to beat specialized tools feature-by-feature. Instead, it focuses on the 80% of features most users need daily.
  
  
• Simplify enterprise adoption: For Heads of AI, OpenAI offers an elegant solution: roll out ChatGPT, get 80% of what you need, have only one solution, and save money. The remaining 20% needs to justify the complexity and cost of additional tools.

The bottom line

If you're building enterprise AI tools right now, Canvas should make you nervous.

OpenAI isn't just improving ChatGPT – they're taking what works well in specialized AI tools, building that directly into ChatGPT, and making it just good enough that paying for another tool becomes hard to justify.

The smart activation of Canvas eliminates the “blank page problem” by appearing the moment the conversation shifts to creation. The integration of simple but strategic features makes using separate specialized tools seem like overkill, and the price point makes paying for other tools seem inefficient (Lex is $18/month, GitHub Copilot is $10/month, ChatGPT is $20/month and offers both capabilities and more).

What This Means For You:

• If you're a Head of AI: This is good news. Fewer vendors, lower costs, simpler compliance.
  
  
• If you're a B2B AI tool vendor: Time to rethink your strategy. Being marginally better than ChatGPT at one thing may not be enough anymore.
  
  
• If you're a business user: Watch this space. The AI tool landscape is about to get simpler, but probably more expensive as OpenAI leverages its growing power.
  
  
• If you're a consumer: You get enterprise-grade features at consumer pricing. Use Canvas for everything from writing blog posts to reviewing code before you consider paying for specialized tools.