Deepfakes are undoubtedly a menace but AI has come to the rescue
https://www.bbc.co.uk/news/technology-66267961 

Or has it? This deepfake detector seems good at picking up deepfakes but then goes on to identify a significant proportion of genuine videos as fakes. In addition it needs more than just a short clip to work on.

I can foresee an 'arms race' developing between those who design deepfaking programmes and those who detect them.

---

It's an extension of the anti-plagiarism race in schools, colleges and universities!

It's quite alarming how so-called AI has rapidly been thrust into use. Disappointingly, the higher-ups in Trust IT are now using it to write code (with the expected results, hence my new boss having to spend a week at another school fixing things), teachers are using it to write reports on trainee teachers, kids are using it to write essays and even the Head used it to write a speech praising a kid at the recent prizegiving evening.

Given how blindly people follow instructions from computer (e.g. Sat Nav), I genuinely despair for the future.

It's going to be a real struggle to work out what's true or not, and the sad thing is plenty of people won't even care!

---

I would have said we are without a doubt already there

In the western democracies we are at an inherent disadvantage. The key proponents of this  "Deepfake" technology are the authoritarian regimes who wish to undermine our way of life. With our multi-faceted media environments with many diverse players it is not difficult to introduce these fakes into our news coverage and  sway the results of important elections etc. We cannot easily retaliate in kind as their single source state controlled news media would be almost impossible to infiltrate. Were we to clamp down on our own news outlets to try and prevent this, we would be playing straight into their hands.

The recent initiative by the BBC to launch BBC Verify is a step in the right direction IMO but there is a long way to go



 

---