Have you watched the TV series Person of Interest?

Slight spoiler alert:

The overall premise of the series is that AI can track people, understand their motives based on their behaviour, and predict what they’ll do next.

A similar theme runs through the movie Minority Report.

But, we know that AI is flawed and based on the programmer’s biases.

Now researchers at MIT have successfully fooled the AI face recognition software at least three times.

For me, the best was fooling the AI into thinking a toy turtle was a rifle.

All they did was slightly modify the shell texture and Google’s Inception AI was fooled.

They’ve also managed to have the AI identify a cat as guacamole and a baseball as an espresso machine.

While that all sounds like a bit of fun, and it is, there is also a serious side to this.

AI is used in self-driving cars to read signs and scan their surroundings.

None of these things are always in pristine condition so what do you think the chances of a serious mishap are?

Then there is China.

The Chinese Government is already extensively using AI and facial recognition to track their citizens and monitor their behaviour.

They are ranked according to their behaviour on a scale unknown to anyone outside the inner circle, but any citizen can have their status upgraded or downgraded at any stage.

How often do you think the system gets it wrong?

But, if you can fool this software by changing texture and patterns, what chance is there that people will fight back by printing T-shirts with patterns that fool the machines?

It would be hilarious, and probably dangerous, to wear a T-shirt in China that made the machines think you’re the President, or a lamp post, or a toaster.

Regards,
Brent.

P.S.  Use AI as a helper, don’t let it be your boss.

Ai can help you discover underserved niches.

It can help you put together a plan to serve those niches and make money.

Check out this tool: https://go.wm-tips.com/diamond.