Gemini’s Canvas Feature Is Surprisingly Good for Word Processing and Coding
AI will not come to your job . Probably ? Look, it’s complicated. And anyone who’s ever played with a chatbot has probably wondered, ” Is this what they want to replace a human with?” This is a major problem because generative AI chatbots like Google’s Gemini aren’t very helpful. However, to my surprise, the Canvas Gemini feature can do just that.
Let me be clear: I’m coming at this as a generative AI skeptic , but I’m not an AI proponent either. In some situations (such as creating alt text for images to improve accessibility ), AI-based tools can come in handy. I prefer to find where tools can be truly useful and ignore the noise. In that sense, Gemini’s Canvas is the only thing that made Gemini truly useful for me. Instead of just asking the AI to write for me, it gives me a workspace where I can do things and occasionally ask for help or complete tasks as needed.
What is Gemini Canvas?
Canvas is an additional tool you can enable in your Gemini conversation. This additional workspace can function as a text editor and coding tool. Any documents or code you create with this tool can also be edited directly, just like in any other editor. You can then export the documents you create to Google Docs when you’re ready to move on to the next stage of the project.
If you use it for coding, you can run code directly inside Canvas. This is especially useful for HTML based applets when you use Gemini in a web browser since it is already designed for HTML. The window can switch between a code preview and a text editor to control it directly.
To use Canvas, simply click the Canvas button at the bottom of the tooltip window (or under the + icon on mobile). The first prompt will open Canvas, and I usually like to start with a blank window, so I’ll say something like, “Give me a blank Canvas window first.” You can also ask Gemini to generate something to start with, and depending on whether it’s a text document or code, it will default to the appropriate Canvas window for it.
If this sounds like a more verbose way of simply asking for a blank document, that’s because it is. LLMs are surprisingly good at understanding natural language commands, but are not always good at giving short instructions (more on this later). However, this means that he responds well to verbal commands.
I often use tools like Windows Dictation to write text rather than typing it by hand. This is a great tool in its own right to relieve the constant strain on your wrists that comes from typing all day or if you’re a stream-of-consciousness type. In some cases, dictating commands and having Gemini interpret them can be faster for me than doing things manually.
With that in mind, I’ve found a few use cases that are actually more useful than the usual actions.
Organizing brainstorming sessions and drafts
“AI for brainstorming” is often the first thing people suggest, and it’s not very useful if your goal is to make the tool think for you. I’d like to suggest something a little different. When I’m brainstorming ideas, instead of using Gemini to come up with ideas for me, I use it to organize my ideas and notes, and turn approved suggestions into outlines.
The trick here is that Gemini can create, manage, and most importantly understand the differences between many different documents. For example, I sometimes use Gemini to write abstracts of articles in one document and then turn them into a draft in another. While working on a draft, I can ask Gemini to “change the status of presentation [x] to Complete in the presentation document.”
When I tested this, I expected Gemini to do something ridiculous and wrong. My “status” tracker is just an extra line under a paragraph that says something like “Status: Approved.” But Gemini understood what this meant and changed the text while I was working on another document. It’s a small thing, but it shows how convenient it is to have an assistant who understands what you’re trying to achieve.
By the way, Google Docs also has an “Ask Gemini” tool built in, and you would think that this is the best way to accomplish the same task, but for some reason it is not. I tried asking Gemini to make small changes to the Google Doc, but it doesn’t really change anything unless I manually highlight the text. It also can’t link to other Google Docs in your drive, so it’s more limited than Canvas in Gemini. I still usually copy my drafts into Docs later, but for early brainstorming, Canvas is a handy tool.
Edit formatting and presentation (in Markdown!)
Gemini (and frankly, every generative AI tool I’ve tried) is absolutely terrible at writing anything interesting. Out of curiosity, I checked to see if Gemini could write a draft of the article for me while I simply made edits. In each case I had to rewrite every word. Not only because they are often wrong, but also because they are bland, boring and unpleasant.
Formatting the words you write is a different story. When using Canvas, Gemini can understand spoken commands such as “convert all subheadings to H2” or “move second section of subheading above first” and apply them directly. It’s also great for more complex style options like “convert all H2 subheadings to sentence case” or “if there are semicolons in this document, remove them.”
As a bonus, Canvas uses Markdown formatting by default ( unlike Google Docs ), making it especially useful for creating drafts that you’ll later copy elsewhere. This isn’t really a feature of Gemini or Canvas per se, it’s just a nice touch that I think makes things easier.
Create your own shortcuts for complex changes and commands.
Everything I’ve said so far is true and useful, but if I’m honest with myself, it ‘s not enough to convince me to fully integrate Canvas into my workflow. What pushed me to despair was that I figured out how to create my own shortcuts. This trick is hidden in Settings > Saved Information .
Google presents it as a way to store “information about your life and preferences,” but in reality it’s a place to put any instructions Gemini needs to remember at all times. So, for example, I have a saved block that says “If the entire prompt is H2, convert all subheadings in the Canvas window to H2.” From now on I will only need to enter “H2s” to perform the action. This is a convenient way to create your own shortcuts for any instruction you like.
You can use this for more complex recommendations, such as: “When I ask for a diagram, format it [however you want].” Or, if you’re into programming (more on that below): “When I ask for an applet, let’s assume I want it to be written in HTML5.” You can make these instructions conditional, as I have done here, so that they only fire when you need them.
Create small, personalized applets for niche use.
Some time ago, a friend of mine was trying to imagine parking in a confined space. At that time, I opened Blender , a 3D modeling and animation tool, to render an example for her. Just recently I was trying to see if I could create a similar tool in Gemini. It took one hint. I asked for an animated HTML5 tool to demonstrate parking, gave some specific examples of what I needed, and soon I had a little applet that was much easier to make than the one I showed my friend.
One could quite reasonably argue that there are easier ways to code this, or that the resulting code is bad for some reason. But I can’t program, at least not that well. However, Gemini does a great job with very simple little applets like this one. Just don’t expect to return everything it generates to your boss if you’re a developer.
Do not use it as a substitute for creative or painstaking work.
It goes without saying, but Gemini (or any AI chatbot for now) is no substitute for thinking for yourself. He is terrible at writing anything interesting or creative, he can get facts or sources wrong , and, most damning of all, he confidently claims that he can do or has done something that he is not capable of.
In the end, you will still have to work. Leaving aside the fact that AI tools often perform effortless work, the question remains whether it is even ethical to use AI for work or creativity . Most AI tools were created by collecting copyrighted data , including from humans, which the same tools then attempted to replace. (Not to mention the rising energy costs required to train generative AI models .)
Artificial intelligence tools can be extremely useful if used for the right purpose. I admit, partly I write this with hope. Using Gemini’s Canvas allowed me to look at how useful AI tools could be if they were an interpretive layer for applications rather than a replacement for real work.