Are you ready to experiment with using AI to take control of your workday?
We all know that technology is constantly evolving and AI is no exception. In this essay collection, we're going to explore the potential of using AI to reduce your workload and increase your productivity. But keep in mind, this is an experiment, and we can't guarantee that it will work for everyone.
We'll take you through the process of finding the right tasks to assign to AI, and explore the ethical considerations of reducing your work hours. But that's not all, we'll also dive into the exciting world of AI-assisted writing and discover how it can enhance your work. So, join us on this journey of discovery and experimentation and see how AI can help you take your job to the next level - while still getting paid for it.
How to reduce my workload by 50%
What is my situation now?
I have a regular 8 hour job now. I write about 45 minutes a day. I don't measure how much I work. I don't measure how much I write. But it feels off, not the way I want it to be.
What is my goal?
My goal is: half the work, double the writing. My goal in numbers: max 4 hours of work, min 1.5 hours of writing. More specifically: max 2 hours of software development, max 2 hours of management. Not a hard goal, but a "let's see what happens" goal.
What is your goal?
How much do you want to work? Be realistic and idealistic at the same time. How do you want to spend your time you save of work? How much time do you want to spend with it? Be realistic and idealistic at the same time.
Be Transparent or Pretend?
I want to work 4 hours per day on a 8 hour job and get the same salary.
I have two options. Be transparent sharing it with my manager or pretend I still work 8 hours.
Pretend
Reducing work hours without informing management or colleagues can lead to a loss of trust. They may assume I am not committed. Not to mention other consequences if they find out.
This choice has ethical implications too. If I pretend to work twice as much, is it a form of lie? How authentic am I if I am not transparent about my intentions?
I have friends who do this and I do not judge them. Research shows people put in 4 hours of work a day on a full-time job.
It still does not take away my uneasiness.
Be transparent
It is possible to gather data to support your argument. Show your manager how reducing work hours can lead to increased productivity and better work-life balance.
You can provide examples of successful individuals and companies who have implemented shorter work weeks or flexible schedules.
Consider the potential concerns your manager may have, such as how your reduced hours will affect the team or company. Come up with solutions to address these concerns.
I cannot imagine having this conversation with my manager.
The choice
Do you have similar ethical concerns? Can you imagine that conversation with your manager? I choose to pretend.
What do you choose?
The Advantages of AI-assisted Writing
Let's start with the obvious advantages. AI produces text in a consistent style. It can shape your uncombed thoughts into nice and bland business English.
The text will have a good grammar and spelling which is twice as useful for us, non-native speakers. Chances are you can write in your mother tongue, but given less text on the internet written in it, the result is less sophisticated.
Most obvious of all, the AI can generate an insane amount of text in a matter of seconds.
Amount matters not only to sweatshops of words. When I am unsure about an idea, I ask the AI to come up with a handful of alternatives. Then I pick one or combine a few.
It has an attention to detail. When I leave out something, it will fill in the gap. I use it to make my argument stronger or to make unexpected connections.
It is my collaborative partner, not a tool. We create by having a conversation. It can play any role in the creative process, no matter where I am stuck.
When I write the traditional way, I focus on sentences. AI helps me focus on a higher level, on a few paragraph long chunks of text.
How has AI changed your writing process?
The Difference between Bullshit and Lazy Writing
They say AI is a well-read bullshit generator.
Bullshit can be defined as a mixture of lies and nonsense presented as truth. It's like fertilizer for the mind - it may help some things grow, but ultimately it's just a bunch of crap.
Bullshit is not exactly lie, it's not necessarily false. It can also include exaggerations, half-truths, or statements that are not supported by evidence.
A liar knows they're lying and wants to deceive, while a bullshitter doesn't necessarily know or care about the truth and just wants to impress or influence.
We call something bullshit because of the content.
Lazy writing is like a buffet of writing after their expiry dates - it's full of tired old phrases and ideas that have been served up one too many times. It's the literary equivalent of a "meat and potatoes" dish - it's easy to make, but it's lacking in flavor and creativity. There's nothing wrong with its content which has been true for decades, sometimes for centuries.
Sometimes even the most well-intentioned people can produce lazy, clichéd writing. This happens when a person is so focused on getting their message across, that they don't put in the time and effort to make it interesting or engaging. They fall back on tired old phrases and ideas, thinking that it will make their writing more relatable or easier to understand.
They say AI is a well-read bullshit generator. It's not true. AI uses a technique called "language modelling" to predict the next word in a sentence or phrase. It will often choose the most frequently used phrase or word, rather than the most creative or accurate one. This is lazy writing by definition.
At the end of the day, AI is only as good as the human using it. Now go back and weed out all the clichés from your text.
From Coding to AI-ing
As a software engineer or engineering manager, incorporating AI can improve your productivity and efficiency. One powerful strategy is to start tracking your time and thinking of each task as if you were communicating with an AI. This way, you'll be able to analyze patterns and identify which tasks can be automated with AI.
Imagine you're working on a project to improve website load times. Instead of jumping right into coding, take a moment to phrase a clear and specific prompt like "Improve website load times by reducing the size of images and minimizing code." This prompt sets a clear goal and serves as a guide for the task at hand.
Now, go ahead and do the work yourself. As you complete the task, pay attention to the challenges and difficulties you encounter. These are the areas where AI can come in and take over, making the process more efficient and less error-prone.
Once the task is complete, take a step back and look at the prompt. Ask yourself, "Which parts of this task could be handled by an AI?" In the website load time example, you might realize that image compression and minimization can be automated with AI.
But don't get carried away, introducing AI is a gradual process. Start small and manageable, then gradually move on to more complex tasks.
From Human Prompts to AI Tasks
I've been collecting a set of "human prompts" that define the tasks I want the AI to perform. I call them human prompts because they are instructions for myself. I can't wait to have my first task assigned to AI.
My first step is to analyze the prompts to see if they can be fed to AI. This means reading each prompt and checking if it means these criteria:
- "Clarity of Command" for human prompts. Is the prompt clear and understandable by AI? Or only by me?
- "Text or Code?" for determining the task type. Can this task performed by writing text or code? AI cannot make phone calls or cook pasta for you. That may change by time.
- "Data Availability" for the input data in written form. Do I have enough data for this task? Is it in written form? If the information is in your brain only, it will not work. The next step is to select a task that meets all these criteria. Now feed the prompt to the AI. Refine it until you get a response you can use.
Did it work? What did you learn? How could you write a better prompt next time?