I genuinely wonder how many refunds have already been issued for damage that never even existed.
We're at a point where someone can fake a broken product in seconds, yet retailers are still expected to make financial decisions based on a few photos.
How long before photos become completely useless as evidence?
This laptop looks damaged.
Except the damage doesn't exist. AI generated it in seconds.
And photos like this are still being used to make real refund decisions.
We built Verilo to verify the actual product, not just a photograph.
Would you have spotted the fake?
@tobi Canada wants more Shopifys. Yet C-22 goes beyond regulating when data can be accessed to specifying what technical capabilities designated providers must build.
That's not just privacy policy; it's government influence over software architecture. A precedent worth debating.
I don't think people have fully grasped what happens when a photograph can no longer be trusted as proof.
There are entire systems built around the assumption that it can.
I made a short film about one of them. I'd love to know your thoughts !
#AI#ecommerce#FraudPrevention
What's interesting to me is that the person reviewing the claim could do everything right and still get it wrong.
They can ask for another angle, a handwritten note, more photos. None of those things necessarily establish whether the damage is real anymore.
It's a problem with the process, not the person.
#AI #DigitalTrust
The cracked screen wasn't real. Neither was the crushed packaging or the handwritten note.
Every piece of evidence in this video was fabricated.
Yet retailers are still making real financial decisions based on photographs.
We're building Verilo because that needs to change.
#AI #Ecommerce
I almost miss when buying a work computer was just about what I needed it to do.
I made my 48GB MacBook Pro struggle today, and it got me thinking: if one AI agent sets off three more, and they each trigger more work, how do you even plan your computing requirements?
#AIAgents
I wonder how much of the actual cost gets missed when we focus purely on the tax revenue. If someone's wealth is largely tied up in equity, that 5% has to come from somewhere. So do they sell shares, borrow against them, or start moving capital elsewhere? Each has a different effect on liquidity, collateral and ultimately investment. I'd be interested to see if anyone's actually modelled that.
I genuinely can't believe what's happening with @getverilo right now.
The growth we've seen almost overnight has caught me completely off guard.
A few months ago, this was just an idea I couldn't stop thinking about.
Now I'm starting to wonder just how far it could go.
I think we're putting far too much emphasis on teaching people how to prompt AI, and nowhere near enough on teaching them how to critique it..
Knowing what to question, what to accept, and when another revision is actually making something worse.
That judgement is going to be incredibly valuable.
3 months ago I was living in Australia with an idea I couldn't stop thinking about.
Now I'm living in Bangkok and that idea is a real product, out in the world.
It's honestly a little surreal looking back at how quickly everything changed.
I think we're entering a period where seeing something and knowing it's real are becoming two very different things.
It's one of the things that got me thinking about Verilo in the first place. How do you actually prove something happened when a photograph alone isn't enough?
For 200 years, a photograph meant light had actually touched something real.
Now we can create images of things that never existed.
So when a customer sends a photo of a damaged order, what exactly are you verifying?
That's the question we built Verilo around.
https://t.co/9FpOtV4tXp