At Home with Tech

Unlock the power of all your technology and learn how to master your photography, computers and smartphone.

Category: AI

Using AI to Bend Reality in My Vacation Photography

I enjoy taking lots of photos of my life. Why exactly? Well, why does anybody?

  • To remember. To reflect. To share. To prove that it happened.
  • Family. Vacation. Adventure. Misadventure. Home. Passion. Life.

But now with a little help from generative AI, you can whip up your own life’s photos without having to actually experience… your life. Now, you can document your imagined life and share this alternate version if you want.

Sure, I know this all sounds rather absurd. But the fact that it’s possible now… easy in fact, should give us all pause. What is real anymore? 

This is, of course, a big topic of discussion on any number of fronts. For the moment, I’m simply directing the focus inward from societal to individual impact.

Google Whisk’s ‘Precise Reference’ Mode
Okay. So, with that set up, here’s how to have some ‘fun’ reinventing your life in pictures.

I’ve been experimenting with Google Whisk (one of several players in this disruptive and quickly evolving digital sandbox). Here’s the game-changing trick I’ve recently learned that turned this AI image generator into a reality-blending tool.

  • Activate ‘Precise Reference’ mode in Settings.

From there you simply need to upload at least one pictures of yourself for Whisk to see. That’s the critical reference point that puts ‘you’ in the new scene.

You can also upload photo backgrounds to help art-direct your shot or create them via text prompts.

Then, everything is ready for you to prompt your new photo into existence…starring you.

And then just click to generate.

Photos from My Vacations Not Taken
I followed the above steps, and within seconds, I received back each of these vacation photos from my alternate universe.

Sailboat Racing Fun

Seeing is Believing?
Whoa. This other guy sure is having fun. Maybe he should dial it back a bit. No, these AI-generated shots aren’t perfect. But they’re close enough to prove my point.

Creating a fake photo isn’t exactly new. Other tools have been available to do that for years. But it used to take a certain amount of skill and effort. Now, with a couple reference photos, a few clicks and a basic understanding of the process, everyone can access this great power. 

And we all know the line from “Spider-Man.”

Time to Meet Your Doppelgänger
I am fascinated. I am concerned. I am confused. My creative center feels in flux. My very existence can be morphed (as can yours).

But I’m determined to figure out how to properly integrate this AI-led creative revolution into my own reality (as we all should).

To truly understand it, you have to know how to operate within it. This is no time to ignore what’s already happened.

That’s why I’m spending time creating a vacation album from my alternate universe. Yes, it’s been a fun exercise. 

But I couldn’t be more serious.

How to Magically Turn your Photo into a Video Using Generative AI

The creative realm is no longer inhabited exclusively by human minds. Generative AI tools have revolutionized how you and I can develop our own creativity. Yes, AI may still require our inspiration, but then it magically does most of the work.

One way to quickly immerse yourself in this new creative workflow is through a simple shortcut. Just start with a real photograph/image that you’ve already created as a reference point. Then, it’s much easier for an AI app to develop it further as opposed to having to start the process from scratch through extensive prompts.

For me, that’s been the key to easily unlock AI’s visual powers.

AI Follows the Creative Direction from your Photography
After uploading your own photo, you can create an AI-generated clone in one click that looks remarkably similar. The AI takes certain creative liberties, but it nails the framing and essential visual elements.

And then, with just a few more prompts and a click, you can generate short video clips that bring your photos to life.

So yes, we can now create videos out of thin air based on our photography. 

Here are a few examples I generated after feeding my photos through Google’s Whisk and Veo generative AI models. (Other companies offer similar fast-developing technologies.)

Maine Sunrise
I snapped this sunrise photo during our Maine vacation:

Here’s the Google Whisk version:

And here’s the Google Veo video:


Alaska Sunrise
Here’s my sunrise shot from Homer, Alaska during our 2023 trip.

Whisk photo:

Veo video:


Baltimore Sunrise
Here’s my photo of people walking by the water in Baltimore, Maryland.

Whisk photo:

Veo video:


Two Paddleboarders on the Ocean
I photographed these two paddleboarders in Maine last year.

Whisk photo:

Veo video:


A Man and his Dog
During our vacation in Alaska, I took a photo of a man with his beautiful golden retriever. I processed it through Google Whisk and Veo and generated this:

Whisk photo:

Veo video:


Generative AI Provides the Paint and Canvas
I find these examples remarkable and clearly disruptive. I’m still adjusting to the massive implications to all this. 

Generative AI tools have quickly become our new paint and canvas to bring our creative ideas to life. And the results will only get better.

So, it’s time for all of us to relearn how to paint, even as photographers.

Here’s What Happened When I Tried to Create an AI-Generated Podcast about my Blog

I invited Google’s NotebookLM to create “deep dive” podcasts based on a few of my blog posts, and here are the stunningly good results.

The written word, though certainly enduring, has plenty of competition out there. We all consume information in different ways. But quite frankly, photos, videos and podcasts are often more compelling and clickable.

Barrett’s Podcast?
I’ve occasionally thought about creating a companion At Home with Tech podcast, but that takes a fair amount of work to maintain. (It’s enough of a challenge for a busy dad to pump out the written version every week.)

That said, I did create this pilot podcast episode some years back.

How to Create Your Photo Archive in the Cloud: The Podcast

It was great fun to do… but too big of an ongoing lift.

ElevenLabs
Last year, I revisited the idea and considered applying a different solution using A.I. to more quickly generate my podcast. How? 

I first cloned my voice by digitizing it through ElevenLabs’ website. Then, I simply copied and pasted my blog’s text into ElevenLab’s interface to magically create a spoken version with my cloned voice! It was certainly much quicker than doing the voice work myself… not that I can’t do that.

Yes. Text to voice… BAM!

Here’s that pilot:

Should You Clone your Voice to Help Preserve your Legacy?

While certainly simpler and faster, I still didn’t greenlight my own podcast series. (I decided to stick with my core product.)

But I haven’t stopped pondering the challenge… and opportunity.

NotebookLM
A few months back, a friend of mine showed me a nifty Google trick using A.I. to magically create an audio podcast. Google’s tool is NotebookLM. He put it to work and generated an authentic sounding A.I. conversation about my career by simply loading my LinkedIn profile. 

Within minutes, a breezy 10-minute audio podcast appeared on his iPhone with two relaxed A.I. personalities chatting about my awesome life to date.

It was flattering, but also a bit weird to hear ‘people’ talking about me in this way. And beyond marveling at this parlor trick, I felt it wasn’t usable in a broader sense. (I wasn’t going to post this over-the-top publicity anywhere.)

I can Now Create my Podcast in One Click!
But I thought about NotebookLM again recently and navigated over to the URL: notebooklm.google to see how it’s been evolving.  

I realized this virtual research assistant actually has plenty of uses (like summarizing marketing plans, course reading, research notes, meeting transcripts and sales documents). 

And then, I spotted the ‘Audio Overview’ section in the top right of the page.  That’s the place where you generate the A.I. conversation about your topic. And there are any number of ways to feed in what you want the A.I. to absorb and then talk about (websites, PDFs, Google Docs and even simple text).

So, I uploaded a link to my recent blog post, and within a few minutes, my podcast was ready. Click. I listened to the likable pair of podcast hosts effortlessly discussing the detail from my blog. 

My jaw dropped. It all felt like magic. I immediately downloaded the file.
Here it is:

Safe Garden Hose Watering Solutions

Yes, I’ve done a bit of light editing to it using GarageBand. I’ve also trimmed out a few sentences where the A.I. rambled on in a few places. (I reserve that luxury for myself, thank you very much.) Then, I added in some music at the top and back. But that was it. Easy.

Here are a couple more…

My Kitchen’s Unexplained Ping: A Tech Mystery


UPS Battery Replacement: Protecting Your Computer from a Blackout


The Voices Sound So Real
NotebookLM’s A.I. voices are remarkably life-like. The casual banter spoken between this virtual woman-and-man team seems especially friendly and so authentic. 

The result speaks for itself. That said, as incredible as this may appear, some of the ‘summarizing’ occasionally offered additional ‘thinking’ that fell slightly outside of my core perspective. (I suppose a real person could also do that.) 

Still, I think this auto-generated podcast could be a nice companion piece that offers an alternate way to consume the essence of my blog’s content. 

Will You Enjoy Listening to This?
We’ll see if I add in this A.I. podcast as an ongoing feature to my blogging. It’s hard to know if everyone (anyone) wants to regularly listen to two virtual coffee-talk personalities doing a deep-dive exploration of my blog.

But for now, I’ve got to admit… NotebookLM is much more than an amazing parlor trick.