Grok's deepfake disaster & AI that could save lives

Описание: Grok's deepfake disaster and AI that could save lives.
Two stories. Two very different approaches to responsibility.
I unpacked both with Ahron Young on Ticker this morning.

On Grok's sexual deepfake crisis:
The complete lack of guardrails before rolling out this image editing feature is staggering. The EU, UK, and India are up in arms.

India gave Grok 72 hours to remove the feature or face legal consequences.

The UK is investigating which laws were breached.

Grok apologised, but come on. As if you wouldn't predict this was how this new image editing feature would be used.

This isn't the first time either. Grok in Tesla cars was telling underage users to send nudes. Absolutely bizarre.

The real problem? The race for AI is prioritising speed over safety.

Every platform wants to be first. Risk factors and guardrails are taking a backseat.

I've been watching content from CES in Las Vegas. The tech is mind-blowing, but there's very little talk about risk and ethics.
That's concerning.

On the University of Delaware AI model:
This is the flip side. AI being used for good.

Researchers developed a model that can identify potentially harmful social media videos before they go viral.

Specifically: content that could trigger suicidal responses in vulnerable viewers.

How it works:
It uses medical risk factors based on years of research.
Analyses what's being said, the tone, the music, the combination.
Runs sentiment analysis on comments.

The aim isn't to stop the video from being posted. It's to catch it before the algorithm pushes it out and it goes viral.

The freedom vs safety debate:
Should we be predicting harmful content or leave it to users to make their own decisions?

I think if we can identify content that poses medical risk to vulnerable people and stop it from being amplified, we should.

TikTok already uses similar technology on live streams.

This research is early stages. But if it can be developed further with accuracy, platforms should adopt it.

Ahron and I discussed two very different stories.

One shows what happens when tech moves without guardrails.

The other shows what's possible when AI is designed with responsibility built in.

The technology isn't the problem. The approach is.

#AI #TechEthics #AIGovernance #SocialMedia #Grok #AIforGood #Ticker #ResponsibleAI

Want to create live streams like this? Check out StreamYard: https://streamyard.com/pal/d/59843241...