[suckitsam]

These models are very unsafe because people (not me) aren't able to tell fact from fiction. Just think of how some gullible fool who can't make heads or tails of situations in real life, let alone when a computer tells them it's the truth (again, not me).

There are so many people out there who haven't had the benefit of a liberal arts education from an average midwestern university, so I think it's upon all of us to protect them from the mis/disinformation and Russia's election interference (but only every other cycle).

For example, you could have accidentally been led to listen to Kanye (who is bad), had AI not fact checked you. Please, think of all the children hospitalized each year while trying to melt eggs.

[didntcheck]

Yep, that's what always gets me about those advocating for "responsibile" restrictions on AI or other technology - the writer always seems to have the base assumption that they are one of the esteemed few with the lofty intelligence and morals required to tell the plebs how to think. It's no different then the nobles wanting to keep literacy and printing presses away from the lay people

[rcbdev]

You're infantilizing an entire subgroup of humanity based on nothing but your perceived stupidity of them.

[klibertp]

If the last sentence wasn't enough to tell you the GP is being sarcastic, then the "infantilization" you mention might not be completely baseless...

[gmerc]

respectfully, the same would make video games and movies and fox news dangerous.

[suckitsam]

Yes, they should also be outlawed.

[allarm]

And knives. And other sharp objects. And anything that can serve as a blunt weapon. Ropes can also be dangerous. Communication between people can lead to dangerous thoughts and ideas, we must censor it. In fact we should ban everything.

The only fucking things that should be allowed are small solitary cells with padded walls covered with Mickey Mouse pictures and sugary drinks with testosterone-reducing drugs.

[_lvbh]

I think you would like China

[MetricExpansion]

The AI safety people should be ashamed that their legitimate views cannot be easily told apart from the sarcasm of the GP.

[dragonwriter]

> These models are very unsafe because people (not me) aren’t able to tell fact from fiction.

People who aren’t able to tell fact from fiction are unsafe, not the model.

[crooked-v]

As a fictional Austrian movie star once said, "That's the joke."