How to choose which tasks should be automated with AI.

Each morning, some people typically sit down at their computer with a cup of coffee and type /today into Claude Code . Then, they sip their first few sips of coffee while watching Claude retrieve new tasks from the Trello board, create a today.md file with a to-do list for the day, and generate a research summary based on the latest academic research on topics they are interested in.

 

After reviewing the today.md file, they began planning their day. Some were getting into the habit of asking Claude, "Is there anything on my to-do list that you could help me with? "

 

So which tasks should you automate using AI?

Understand what AI workflows can do for you.

Many people started using ChatGPT on their web browsers shortly after its launch. After a few months, they began challenging themselves to perform more tasks with AI. They started wondering, "Can ChatGPT help with this? "

As they explored more use cases and grew tired of copy-pasting, they switched to Claude Code. But the core philosophy remained the same. They continuously strived to see what LLMs could accomplish today.

The default approach is to try to do everything with AI. But there are only so many hours in a day, and you have to prioritize which tasks should be turned into workflows and which should be done individually.

A workflow is a sequence of steps necessary to complete a task, some of which can be automated by AI, others can be assisted by AI, and still others can be performed by you.

When reviewing the workflows you've created, you may see some clear patterns. Let's use AI to help you:

1. Do more than what you're already good at.
2. Remove obstacles in the tasks you regularly perform.
3. Eliminate things that exhaust you.

The goal is to use AI in a way that helps you scale your impact. Let's look at a few examples.

You're writing around 35,000 words per month. That's up from about 8,000 words per month. You write more frequently and in more detail. You refer more to academic research. You share more stories—both personal stories and stories of others.

 

The increase in productivity comes from the way you work with Claude. Maintaining writing motivation becomes much easier because Claude always asks if you're ready to write the next part. It's a small reminder, but enough to help you stay focused on the work for longer.

An academic research process makes it easy to stay updated on the topics you're most interested in. You spend more time editing your writing structure with Claude before you start writing – so you have less need to remove unnecessary content.

Writing is one example of how AI is helping you do better at what you're good at.

There are some tasks you'd be happy to automate completely. Many people enjoy reading interesting academic papers on a variety of topics: creativity, collaboration, the effectiveness of AI, etc., but hate searching for relevant articles, so they've completely automated this. Every morning, the automated research process searches for relevant articles and adds them to the daily research summary for review.

Let's discuss in more detail how we can choose which tasks should be automated using AI!

Choose your first AI workflow.

As you go about your daily tasks, ask yourself, 'How can I use AI to help with this?'

There have been countless blog posts and even books written about how and when to delegate. But ultimately, the advice boils down to this:

1. Develop awareness of how you use your time.
2. Identify which tasks you can delegate and delegate them.
3. Spend your time on the tasks you're best at.

 

This is a good starting point when thinking about how AI can help. We often emphasize automations that save time and increase efficiency.

But don't overlook the things you enjoy doing and want to do more of. This is where you'll often get the most benefit from using AI.

You should automate what you can, but more interesting use cases involve using AI to support what you do best.

To explore potential workflows, start by asking yourself a simple question: Can artificial intelligence (AI) help with this? Ask yourself that question repeatedly while performing different tasks and use it to drive your experimentation process.

You can also start this way. As you work today, ask yourself: How can AI help with this?

Assess whether a task is a good candidate for an AI workflow.

Through trial and error, you'll identify a few more questions that will help you choose where to focus your efforts on AI-powered workflows.

When starting a new task, ask:

1. Is this a one-time task or one that needs to be performed regularly?
2. Would you enjoy doing this task yourself, or would you delegate it to someone else if possible?
3. How complicated is this task?
4. Can you describe how you would complete the task, step by step?
5. Does completing a task require human judgment?
6. Can you define what "successfully completed" looks like?

What is the level of risk if the task isn't done well? This list may seem long. But it usually only takes a few minutes. The answers help you decide whether an AI workflow is worth the investment, what you should automate or add, and what you need to continue doing manually.

Deciding where to start can be challenging. The goal is to provide you with plenty of inspiration to choose from, so you can begin building your first AI workflow today.