Where and how I use AI to address the one finite resource we all have, time.
I think a lot about time these days.
The time it takes to get my two-year old son ready for daycare. The time it takes to get him ready for bed. And everything in between.
I’m also currently looking for my next product design role so I’m thinking about how I’m using my time throughout the day. Am I spending too much time job searching? Am I not spending enough time networking? What about learning a new skill?
Oh shoot, what are we having for dinner? I would love to get a workout in before I need to find/cook a recipe and pick him up. How many podcasts can I fit in while doing all of that?
Do I have enough time?
It’s “Create/Consume”, not a “Work/Life” Balance
Having the time to think about…well time….has led me to no longer seek out “work/life balance”, but rather reflect on how much time I’m devoting to creating something, or consuming it.
For instance:
Am I creating a new memory with my family, or scrolling my phone?
Am I trying to write a new post/blog, or reading someone else’s?
Am I trying to cook a meal, or order DoorDash?
Am I spending time working on my next project, or is that being consumed by one meeting after another (or in my current state, one job application after another)?
So on and so forth…
With sooo many things drawing (even demanding) our attention…we all need to start asking ourselves if we’re wasting the day consuming …or do we want to focus more on creating…regardless if it’s at work, or in your life.
Where AI is fitting in my life
AI and time is a bit of a paradox.
You can ask it to create anything, thus saving you time…only to find yourself going deeper down its rabbit hole of chat results, “artifacts”, generated images/mocks, etc. etc…which can be a big waste of time.
After all, you’re conversing with a machine that is built to cater to your every question and idea. That’s why I keep this quote in mind:
“Purposeful play is a powerful thing.”
I’ve started to keep a running list of ideas to see how well I can execute on them by using certain AI tools within a certain amount of time/money.
The more I play around, the better sense I get in how much time I want to devote to that idea.
Some have come together
Image: Having some fun with animation: aniletter.figma.site. View on Medium ↗
Image: A tribute to one of my favorite songs/albums: junglelandtribute.figma.site. View on Medium ↗
Some not so much
Image: Wanted to recreate the feeling of watching an old movie trailer: vcrexperience.figma.site. View on Medium ↗
Other examples
Job searching
In preparation for job interviews, I used to do hours upon hours of company/role research via lots of Googling and assembling into word docs. Now, I use Claude to help do that:
Image: Recent example. View on Medium ↗
Applying what I’m consuming
I also read lots of design articles but I would be the first to admit it’s hard to remember key details as time goes on, especially as the content machine keeps rolling.
That’s why I wanted to leverage Claude’s Project feature to help me leverage that content to help me save time on evaluating designs in the future.
Image: An example of a Claude project/instructions I set up for future work use cases. View on Medium ↗
Managing my podcast addiction better
My last example comes from a hobby of mine.
I love podcasts, and yes I am that guy that says in public “I heard this in a podcast…” (I’m trying to be better!)
However, I’ve found myself prioritizing certain tasks in my life based on if I’m able to listen to a podcast while I do them.
Enter AI agents.
Once I found a tool I liked, I created an AI agent to do the following:
- Track any new video posted by a YouTube channel I follow via their RSS feed (most of the podcasts I follow also publish YouTube videos)
- Ask me for approval before continuing the run (since I may not want summary for every video that gets published on that channel)
- Find the transcript
- Summarize the YouTube video (based on my prompt)
- Send that summary to my email
Image: Done with Relay.app. View on Medium ↗
Although I would still like to listen to the podcast, I don’t always have the time, nor do I really want to let it dictate how I spend my time throughout the day.
This approach gives me the choice to either listen to it, or read the summary in order to get the insights I wanted out of it in the first place.
What I’ve learned
Through the course of using these tools, I’m starting to form my own set of principles to help me save time on future ideas/projects regardless of which AI tool I’m using.
Learn the language
Don’t be fooled by the simple nature of these chat interfaces, or any AI-company marketing that is out there. To get the results you actually want, it’s not always as simple as typing in your question.
Although these tools are getting better at addressing this articulation barrier, you will be far better off learning how to actually communicate with these machines. That includes, first and foremost, learning how to prompt effectively (my top recommended resource: Prompt Engineering by Lee Boonstra).
Next, you need to understand that for AI, context is king. Without it, these tools will likely underperform your expectations. So if you’re building a website with one of these tools, for instance, it would be helpful to have a working knowledge of CSS for starters.
If not, always have a goal in mind, and better yet, some thoughts about how you want the AI to approach your idea/question.
Be economical
I love how there are so many posts and recommendations of AI tools and yet none of them really mention the costs if you were to actually sign up for them all.
That said, once you know how to better communicate with the AI tools these companies offer, you can leverage their free (or lower priced) plans in order to better evaluate their performance, in a more efficient way.
After all, the AI will never tell you your idea sucks. It will just notify you when you’ve run out of credits.
Be judgemental
To that end, I keep in mind two psych concepts when I’m using AI to co-create something. They are:
The anchor effect: which “is a cognitive bias where people rely too heavily on the first piece of information they receive (the “anchor”) when making decisions, even if that information is arbitrary. This initial piece of information influences subsequent judgments and estimates, causing people to make choices based on this anchor rather than objective analysis.”
AND
Sunk-cost fallacy: “whereby a person is reluctant to abandon a strategy or course of action because they have invested heavily in it, even when it is clear that abandonment would be more beneficial.”
Reason being, I’ve found it best not to rely on the first answer the AI gives you. Of course, this depends on the prompt, in which I would then say, don’t rely on your first prompt.
And if you’re not satisfied with the results you’re getting…save yourself some time by cutting the idea loose, stepping away from the computer, and trying again later.
Stick with it
Time is the one thing we can’t get more of and yet, everything seems like its competing for it these days.
AI can help combat that, but anyone that uses these tools on a regular basis goes through similar learning journeys like the one I described above. Stick with yours and try to use these tools to give you the time to create something new.
Originally published on Medium.