Hey,
Welcome to a new issue of Experiments in Progress.
This week’s issue started with a simple question:
How much do we really trust what we’re seeing now that AI is everywhere?
From there, it went in a few directions.
A quick experiment to test how confidently you can spot what’s real anymore
The part of AI that genuinely worries me (and it’s not the flashy stuff)
Two recent tech moves worth paying attention to
Let’s dive in👇
🤖 How Worried Am I About AI?
This came up after a short experiment we did recently.
I spent about $20, and in not very long at all, I could generate a video of almost anyone moving, talking, saying whatever I wanted them to say.
That’s the context for this question.
👀 Quick experiment (before you read on): Can you spot the REAL photo?
Before you move on, I want you to try something.
Below are four photos. Three are AI-generated. One is real.
Spot the real one.
➡️ Reply to this email with which one you think is real.
If that felt harder than you expected, that’s kind of the point.
This is a small example of a much bigger problem.
😬 The part that genuinely worries me
It’s impersonation that worries me the most.
Voice cloning is already trivial
Video impersonation is good enough to pass at a glance
It’s cheap, fast, and improving constantly
You could very easily pretend to be:
- someone’s child
- a parent
- a colleague
- a boss
And then ask for money. Or access. Or information.
👀 Why this works
Most people can’t reliably spot AI yet.
I still get messages all the time like: “Have you seen this funny reel?”. And it’s obviously AI if you know what to look for.
Slightly off sound quality
Hands or fingers behaving strangely
Objects appearing or disappearing
Movements that don’t quite track
Most people aren’t looking for these signs.
🌊 The bigger, quieter problem
Beyond scams, there’s something else happening.
Platforms are filling up with AI-generated content at scale.
Entire channels now publish:
- 20–30 minute AI-made videos
- imaginary stories
- fully synthetic narration
Some of them have hundreds of thousands of subscribers and millions of views.
It’s what people have started calling the shittification of the internet but people are still watching, clicking, and paying attention.
Which means:
creators make money
platforms make money
nobody has much incentive to slow it down
🔁 The feedback loop problem (my real example)
We saw this very clearly while working with a brand recently.
The brand shared a high-performing ad
The ad was almost certainly AI-generated
We copied the structure for our video
The brand then replied to us using AI-written messages
Likely analyzed our script with AI
Then made AI-based recommendations
It’s AI creating content → AI analyzing performance → AI shaping the next version. A loop feeding itself.
And we’re all inside it, whether we want to be or not.
📱 Enjoying this newsletter? Share it with a friend who’s as obsessed with tech as you are:
📰 News worth knowing
🏈 Super Bowl heads to Silicon Valley (with tech in tow)
The San Francisco Bay Area is adding a tech innovation summit alongside the Super Bowl experience this year — blending sports, startups, venture, and local tech ecosystems into the event.
Think less “sports tech demos,” more positioning: Silicon Valley reminding everyone it’s still the center of gravity when culture, money, and attention collide.
Source: The Economic Times
📱 Samsung Galaxy S26 reportedly lands Feb 25
Leaks suggest Samsung will unveil the Galaxy S26 lineup and Buds4 at an Unpacked event later this month, with February 25 floating around as the likely date.
By locking in timing early, Samsung frames upgrade decisions before competitors (and before people talk themselves out of spending). The hardware matters, but the real move is owning the moment when people decide whether to upgrade at all.
Source: Android Central
🧺 Next Issue: My Current “No-Regret” Shopping List
In the next issue, I’m sharing my current no-regret shopping list. These are the items I’d rebuy immediately if I had to start from zero tomorrow.
Here’s what I’ll cover:
The categories where spending more actually paid off and where it didn’t
The few items that survived multiple upgrades and trends
What I no longer buy at all (even when it’s “recommended”)
The difference between useful and emotionally comforting purchases
How this list changes (and stays surprisingly stable) over time
🗓️ See you Thursday.