In this week’s newsletter, I talk about how I wrestled with early technology over 25 years ago and how that hasn’t really changed today with AI. I also share my week’s postings about my latest CNA commentary for parents, how Google’s “nano-banana” gave me new hairstyles, addressing boring copywriting in recruitment ads, and I vibe-coded a new app with a sprinkle of flower power.
In 1999, I was an intern photojournalist at The New Paper. When I was not out for hot and sweaty assignments, I would spend hours at the office’s Apple Mac desktop figuring out how to use Adobe Photoshop to edit my photos.
Back then, our photos were published on newsprint, and we had to artificially saturate the images so the colours would look acceptable on the recycled, greyish paper. There was no user manual available, so we had to do this by trial and error – edit the photos, send them off to print, then examine the printed version to see if our experimentation worked. Over time, we figured out the best settings, saved them as “Photoshop actions”, and applied the actions to our photos.
A year later, we started using Nikon’s D1 digital SLR for work, and this camera produced ugly purplish images. Again, I spent a long time figuring out what was wrong with the colour profile (it used the NTSC colour space instead of sRGB) and I designed new Photoshop actions to fix the hues as much as I could. I shared my technique with my colleagues, but few were as bothered about the problem as I was.
Today, newsprint is becoming a relic of the past, and we don’t worry about poor color reproduction or even do that much editing anymore. Our AMOLED phone displays can be too accurate, and our photos are automatically adjusted by algorithms the minute we take them.
But there is still much experimentation to do in the age of AI!
This week, I started to pay more attention to Google’s AI Studio after they launched a few demo apps that utilised the new “nano-banana” image engine. Without getting too technical, AI Studio makes it possible to create advanced AI apps and tools that far surpass what ChatGPT can create. That means I need to invest time experimenting on this platform to create useful AI tools, and also teach the same skills to my Gen AI workshop participants.
Technology gets dumbed down over time for the masses, but if you want to wield new technology quickly, you have
to get your hands dirty, write your own user manual and share your learnings with others.
(FYI, this week’s image banner was made using Google Gemini, Google Whisk, and…Adobe Photoshop – we never give up what we have learned decades ago.)
What I wrote this week
CNA Commentary: I write about the three things that parents need in order to deal with Gen AI.
Learning AI: Why you need to have healthy skepticism and also invest a lot of time.
Copywriting we ought to see in recruitment ads.
Google Gemini’s new “nano-banana” image editing is pretty good… for adding hair to my bald head. It’s now available in the consumer Gemini app.
I vibe-coded a new app this week and embedded “nano-banana” in it to pixelate anyone and anything.
Claude AI cannot make images natively but it can tap on open-source imaging tools with its clever MCP “connector” capability.
NYT stories I’m sharing
The only real solution to the AI cheating crisis.
How to rekindle your love of reading (I think if you’re reading this, you don’t really need this advice).