By now, everyone has heard the rounds about AI, which improves productivity, saves time and automates repetitive tasks. But the question is, how do we leverage AI and use it effectively to achieve these things? Every week, there is a new AI tool to use, promising to elevate your workflow, leading you to believe you need a wide assortment of apps. From chatbots, content creation and writing assistants, all to accomplish each part of your business well. What they don’t tell you is that it leads to context switching, inconsistent results and tool fatigue.
Context switching is shifting your attention from one task to another, a seemingly trivial function as we balance many aspects of our job, from switching language for various stakeholders, switching between projects or balancing life outside work, but as we continue to do this, it comes at the cost of slower performance and a decrease in accuracy in the following switched task. This is known as the switch cost.
Having different apps for different tasks means each AI will lose context on the overall project; this can lead to unreliable data and responses if the prompts provided do not include all the necessary information.
Digital tool fatigue, a term coined by Bryan Robinson, describes a study on work overload, stating that 96% of workers using digital tools experienced burnout rather than helping them keep up. The multitude of tools and switching led to frustration over the lack of data between apps, causing mental fatigue and stress, working on everything rather than the important tasks. Over half the participants stated they wished for a one-stop solution that could handle all their different tasks. This Forbes article and term help explain the problem that more AI isn’t always better.
How does Shape deal with this digital tool fatigue? The first step is to define the core needs your business wants. Offering personalised software development services at an agile rate, an AI-powered code editor is the main tool needed. To help with this process, we also look at content creation in the form of images to help speed up the design process. With the two core needs chosen, defining a criterion for what you require in a tool is necessary to optimise a focused AI stack. We aim for simplicity, ease of integration and a low learning curve.
Cursor checks these off the box, everything is in one place, coding, debugging and documenting. It reduces context switching by embedding the AI into the code editor, meaning you no longer have to transfer code and files into an external app. Integration is convenient and fast, as there is minimal setup. Once installed, AI can be used on a project instantly. The barrier to entry is practically zero as the editor is a VSCode fork meaning the team is already accustomed to. Cursor integrates seamlessly into our existing workflow as it is almost one-to-one, but with the power of AI built in, allowing us to keep the mental overhead low.
To help with quick design and proof of concepts, FigMake is used. Already using Figma makes this AI tool simple, as there is no separate interface to use. Helping reduce creative clutter by combining design generation into one system in a single command, rather than needing multiple plugins and learning them individually. Additionally, to aid with assets, Midjourney is also used; its focus is on generating visuals, and it doesn’t contain unnecessary dashboards or features. The simplest workflow of the three, a prompt and an output, as easy as that. Together, FigMake and Midjourney streamline the entire design workflow from fast concept generation to clean, product-ready designs.
In hindsight, more AI appears to be the straightforward choice, but the cost, both financial and mental, calls this into question. Choosing fewer and better ones is more optimal, focusing on the task at hand with the right tool that supports your workflow. So, evaluate your current AI stack and start by limiting yourself to 3 tools you need and track how much smoother your operations go, and live with the future of productivity, not be consumed by it.
