The ChatGPT Color Challenge

As part of the renos, we need to pick out new paint colors. I’m notoriously bad at this task. Not only do I feel overwhelmed by the color samples, it’s not something I enjoy. I can’t appreciate nuance in colors and after a while everything starts to look the same. In the past, I’ve faced some harsh criticisms for my color choices, which also make me pretty self-conscious about the whole thing.

Normally I hire my designer friend who is exceptionally skilled with colors for this kind of task. However, there wasn’t time for the consult with the tight deadlines. Feeling a bit lost, I decided to upload a photo of our samples to ChatGPT for some recommendations. Within minutes, a list of 2-3 color options from different color families appeared, along with suggestions for ceiling and trim. For fun, I asked about an accent color for one wall. Within seconds a few options appeared. I tried it a few times until my free account cut off my uploads for a day.

What I appreciated about the experience was getting a few solid starting points. Within minutes, my options were reduced significantly. We still had to compare the sample paint chips with our floor and tile samples in the natural light. However, the task felt less daunting than fanning out options with the color sample book. Even narrowing down to one section in the color sample book still leaves dozens of strips to review.

I felt surprised that ChatGPT also offered to create mockups of the samples I uploaded and the paint color suggestions. However, when I tried this prompt, it took a really long time and nothing ever appeared. Equally annoying was when I reached my maximum amount of uploads, ChatGPT wiped out all the previous recommendations. Luckily, I had been making a list as I went along to see if it was making reasonable suggestions. I also discovered, it picked up too much detail sometimes. I carefully considered the background when uploading photos of the floor and tile samples.

In terms of selection quality, I think my friend would do a better job. I don’t think there’s any danger of ChatGPT replacing interior designers anytime soon. Some things still need that human touch.

Taking Out the Trash

This morning I watched a hummingbird in my backyard. The lucky spotting happened leaving the garage to return to the house. I had just finished taking out the recycling and garbage for pick up. A particularly invisible task that rarely feels gratifying. Even though it was early in the day, I could feel the heat creeping up. A slight sheen of sweat already starting to pepper my forehead and collect at the back of my neck. Then I spotted the little hummer zooming around. Instantly my mood lightened.

I stood mesmerized by this tiny flyer, watching her flit effortlessly from flower to flower. The core of her form remained almost motionless in stark contrast to the rapid fluttering of her wings. They appeared as vibrating, grayish smudges on either side of her body. Her beautifully choreographed movements were precise and efficient. She zoomed from one flower to another covering an entire section of the shrub within seconds.

Ever since the Rose of Sharon started to bloom in the backyard a couple weeks ago, I’ve been watching for hummingbirds. I have a love-hate relationship with the Rose of Sharon. The flowers are beautiful and abundant. They attract all kinds of amazing creatures, including hummingbirds. But often I can hear bees buzzing in the flowers and see butterflies.

We have several mature ones in the backyard in several different colors. The ones right outside my bedroom window are so large they cover it completely in full bloom. I’ve discovered over the years that their seeds pods are as prolific as their flowers. Now I spend weeks every fall deadheading the shrubs to prevent them from spreading further. Yet, I look forward to their blossoms each year, mostly because they attract the hummingbirds.

In many cultures, the hummingbird is a sign of joy, luck, and magic. They are said to be messengers, sometimes from loved ones who passed away. Today, it felt like a small gift, just for me. Perhaps as a reward and quiet “thank you” for taking out the trash.

AI’s Overreach

Last week I attended a workshop focused on building an AI strategy. It included a road map, outlining the major considerations and a progression of steps to follow. A highlight included lively conversation with other industry peers. It was affirming and reassuring to hear about others struggling and triumphing with AI.

Usually at these events I walk away overcome by the speed, power, and magnitude AI as a presence in our lives. AI is seeping into everyday things we use. Sometimes we may not even realize it’s there. Other times, we may notice, but there may not be an easy way to disengage. This time, however, I left the event feeling different. Although AI is advancing rapidly, it’s important to focus on the real and tangible ways it’s available and useful to us.

The uses and options for AI are wide, varied, and only limited by our imaginations. However, some of the best benefits are realized by using AI as a humble, assistant. When I started my consultancy, many people advised me that the first position to hire would be an assistant. I never made the leap. If I thought back about that time, part of it was that I didn’t understand how to parse and delegate tasks. I couldn’t separate the mundane, tedious parts of my work eating away at my time and profit margin. It took effort to figure out which aspects of my everyday routine could, in fact, be handled by someone else.

I never did manage to hire that assistant, though I did try a virtual service a couple of times. The service was okay, but the research I needed took a lot of explaining and resulted in subpar results. Now, of course, I could tailor the research myself. AI is available to assist me with a few carefully worded prompts. Within minutes AI can produce similar results to the virtual service and still leave me with the same amount of effort required to review.

Now, older and wiser, I could actually use an assistant. And not for anything grandiose or fantastical. Rather I find AI useful for everyday, mundane tasks. This is especially true for ones I don’t particularly like doing, or can’t manage to find adequate time to focus. Examples include creating meal plans, shopping lists, or even routine email replies. The reach is basic and the results are real.

The Prevalence of AI

It’s impossible to do anything these days without AI. Either somebody is talking about it or you’re interacting with something AI generated. Even more commonly, I’ve started to notice AI prompts appearing. In some places, the AI option remains always visible and ready. Other places, the AI takes a more active role.

Earlier today I realized I forgot to reply to an agency that reached out regarding a new job opportunity a few days ago. The job posting wasn’t a good fit. Replying felt low on my to-do list. I also didn’t want to ignore the email and leave it unanswered. Even though I’m not actively job searching, I always like to keep up the relationships with former employers. In short, I was kind of dreading writing this small reply, even though I knew it would only take about a minute.

Today, however, when I opened it with the intention of replying, AI swooped in for the win! When I hit reply, a small pen icon appeared with the magic words written in a faint gray font “Help me write” Swipe –>.

I swiped right and a prompt box automatically appeared.

I initially typed in “regrets”, but the prompt needed at least three words to perform. Next, I typed something like “turn down offer.” Almost before I finished typing the “r”, a message appeared politely thanking the agency for the offer before turning it down. I read it, added in a couple lines, and fired it off. Tedious, boring task completed with minimal procrastination! Impressive!

Normally I find AI’s presence obtrusive, most of the time because it’s just there. Often, there’s no obvious way to disable it. Most of the time, there’s no choice about if I even want to engage with AI. This time, however, I felt appreciative for its ease of use and quick, decent sounding reply with minimal effort.

Getting caught up in all the fantastical and futuristic uses for AI is easy. In reality, the biggest impact is likely handling everyday, routine, somewhat boring, tedious things. Low priority tasks that need to be finished, but that nobody wants to do. Though it does make me wonder how our expectations will change to accommodate these new easy, features. Will instant replies be the norm? What if me and the recipient both use automatic AI-replies? Our AIs would have an exchange without our human participation. How weird is that!

Time to Thin it Out

As part of my “Renovation Readiness“, I employed a new strategy. Rather than focusing on getting rid of everything, or keeping too much, I used a different tactic. For this one, I concentrated on reduction. The goal was simple, thin stuff out. In this way, I could trim down what I had without making any hard decisions in a time sensitive situation. Or when energy levels were low, not feel pressured or overwhelmed with too many decisions.

I’ve also found this strategy to be useful with others. Going through a big pile of stuff with a goal of complete organization and resolution can feel daunting. However, switching the mindset to thinning it out a little, alleviates some of those feelings. It helped reduce the pressure to “get it done” and instead reframed the task to “getting some done.” Ultimately, this can be effective for keeping energy moving and maintaining momentum. Although it can create future make-work projects, at least they won’t be as big. Plus the work done to thin something out can also be a good launch pad for the second round or purging and organizing.

I discovered this new tactic provided easy to set goals. For example, instead of going through all my files, I only go through enough so the drawer closes smoothly again. Or all the papers fit in the right folders. There’s no stress to reorganize my entire filing cabinet when I can remove just enough to keep it in good shape. I found it easier to keep moving by focussing on trimming down the overall content to prepare for the renovations.

I still created the same piles with the thin it out strategy. Primarily, shred, toss, donate (or sell), keep. Many loads of things were donated, sold, or given away. We threw out several large garbage bags of trash, unusable, and broken items. Still more papers and boxes made it to the recycling and shred bins. In total, I lost track of how much stuff we thinned out. All I know is that there’s still more to go! But now, I have a new approach to help me through the hard moments. I don’t have to decide everything in one go. I can thin out the current area enough to move on to the next one.

The Impacts of AI

This morning I failed an AI detection test featured in the NYTimes. The quiz included 10 videos, a mix of AI-generated and human-made ones. My goal: correctly determine if AI or a human made the videos. I felt good after the nailing the first four and then things took a turn. I got the next 5 wrong and finally guessed the last one right. Total score, 50%.

AI-generated videos are getting much harder to detect. The small glitches and weird imperfections, telltale marks of AI, are largely undetectable. The technology has improved rapidly. Also, AI trains on real recordings, making it possible to create something new from an existing persona. All of these factors combined make it increasingly more difficult to detect fact from fiction.

On the flip side, technology has also improved the capabilities of making films and movies. In some cases, AI may even be used on the backend to expedite editing or creating effects, even though the main content is real. Sometimes AI is used to fill in missing elements or add something to the background that wasn’t there initially. This also blurs the line between AI-generated and human made. I suppose at a certain point, we’ll consider the percentage of a work made by humans vs. AI to determine creation.

Another huge impact of AI is more on the environmental side. AI generating centers require huge amounts of space, time, resources, and energy to maintain. As more of us start using AI, or rely on devices using it, this creates more demand for the centers. The location of these centers and their management also creates impacts. Living in Canada, we’re mindful in which country our Canadian resides. This can create challenges and concerns about data management and under which jurisdiction. For example, the privacy laws in the US are different from those in Canada.

All of this to say, the lines are blurring.