AI Is Dangerous, but Not for the Reasons You Think | Sasha Luccioni | TED

Humans have been, are, and always will be the greatest threat to humanity. I'm more worried about Billionaires with personal AIs or Billionaire access to AI abusing that access to get even more wealthy and more power. I'm worried about the PEOPLE that will abuse early AI that's still basically just a really smart tool.

wow... I went into this video expecting actual real dangers AI is posing in society but all this was was talking about problems of data bias and symptoms of stereotypes without explaining the root cause of the symtpoms while trying to sell her products/services.... And energy crisis? there are billions of people driving to work everyday to do menial jobs that could be replaced by automation. Does that make their work problematic? For a person being paid as a researcher delving into societal problems of AI, this presentation was very lacklustre and borderline a deceptive commercial.

How about looking into how AI is already creating false pornographic images out of people's sns pictures where scammers are threatening people with them, or how the trust of the media is being actively broken down with misinformation and improvements of ai generated video/audio??? or how ignorant/lazy people misplacing trust on AI systems that can generate fake facts can cause people to be misinformed or make mistakes in their industry (i.e. that idiotic lawyer quoting fake court cases generated by OpenAI)? or how children are now not engaging in intellectual research but becoming more and more reliant on automated systems to do their work for them? or how giant tech companies are now closing off open source internet (and changing the reason for existence of the internet) and gatekeeping data for themselves so they could monetize AI services harder? or how society is not ready for automation at scale where it can replace millions of jobs (look to ai systems already replacing crop management)? or how AI driven drone swarm networks basically puts destructive weapons, that has potential to kill hundreds of people very easily, in the hands of anyone? or how smaller nations can now go against the US military might using drones/AI changing the geopolitical landscape where the US can no longer be called the sheriff of the world)?

There are very legitimate concerns with improvements in AI and some artists' ability to get paid becuase their monetary source of income from their industry is being disrupted by innovation is NOT a societal shaking problem to discuss in a forum geared towards a worldwide audience..... Not to mention the VERY ACT of these artists creating art is pretty much the same process of ai generated artwork. If AI could be sued for the art that it produces, then literally every single artist on earth has a right to be sued, and only the original artists (who are all dead) that created the different types of styles/technical process of art would be safe.

Women won the fight against men gender discrimination, now they will fight the AI gender discrimination 🤭

The speaker, who is an AI researcher, received an email from a stranger expressing concerns that their work in AI could lead to the end of humanity.

AI has become a hot topic, making headlines for both positive contributions like medical discoveries and negative instances such as biased AI systems.

The focus should shift from future existential risks to current tangible impacts of AI on the environment, society, and individuals.

AI models contribute to climate change, emitting significant amounts of carbon and consuming large amounts of energy during training.

Large language models, like GPT-3, are responsible for substantial carbon emissions and environmental impact.

Tools like CodeCarbon help estimate and track the energy consumption and carbon emissions of AI models, allowing for informed choices in deploying models.

Artists and authors face issues related to their work being used for AI training without their consent, and tools like "Have I Been Trained?" are created to provide evidence of unauthorized use.

Bias in AI systems can lead to harmful consequences, as shown by cases where biased AI systems led to false accusations and wrongful imprisonment.

The Stable Bias Explorer tool helps reveal bias in image generation models through the lens of professions, highlighting underrepresentation of various groups.

Creating tools to measure AI's impact is crucial for addressing issues like bias, copyright, and climate change, and it empowers users to make informed choices.

The speaker emphasizes the importance of transparency, governance, and collective decisions to shape the direction of AI's development.

Focusing on addressing the current impacts of AI is essential, rather than fixating on distant existential risks.

In summary, the talk highlights the need to address the current, real-world impacts of AI on the environment, society, and individuals, and the importance of creating tools to measure and mitigate these impacts.

So, the answer to bad software is to create more software to police the bad software. What ensures some of the police software won't also be bad software?

It’s crazy you wrote code to get the carbon footprint. I wonder if she measured her code to get that footprint. Hmmmmm

Plot twist: her script was written by AI.

Well said ❤🎉🎉

It's absurd to suggest that anyone with influence has any real interest in hampering the carbon emission of power consumption considering that they could do so rapidly at any time they choose. The machine rebellion apocalypse fantasy is indeed a distraction from the present problems of AI, just as much as carbon emission from the power required by AI being a meaningful challenge is a distraction from the enslavement of humanity.

If everybody with inventions or innovations to combat carbon emissive energy had the freedom to express them today, we would build it all tomorrow and be at near zero carbon emission the day after.

Acting like it is complicated is insidious.

Ai is a tool. Humans will use that tool to their advantage. But there will come a point when Ai cant be controled. Theres nerver been a perfect, unhackable OS.

No matter how you slice it…we are inevitably doomed 😢. Bell Curve kinda says it all…majority in the middle and the middle isn’t that bright 😮

Pretty pointless talk tbh…

What these jerks fail to recognize or tell you is that Ai is a reflection of the evil human epigenetic and evolutionary primal existence. Either Ai recognizes the threat to universe or adopts evil existence; either way, we are doomed.

IA ist also dangerous - intelligent aritification is probably the more distributed danger at the moment??

You sound smart so I'll just burn all my tech and move into a cave and only eat dirt.

Huggingface continues to be the most based AI company

People don't understand electricity and there goes hoping they will care to understand AI when it's already deployed and mutating. BTW it's not AI that is bias, it's society and even worse social media with training datasets been just a reflection of that, AI doesn't have an opinion so can't be bias.

No call to action=no effect

"it's not a done deal" - in a way, it never is. In another way, many things are already here, and being used (often by governments - yes I'm looking at you, China, but also in 'western countries'). Looking at legislation history, implemented things seldom get dismissed, because they have advantages to the ones in power. If they hadn't, then the system would not have been setup. But, (ironically?) it's the ones that are in power that can 'dismiss' a system. But why would they, they are the ones benefitting...
As ever: The ones with power to change something are usually the ones that benefit the least from a change...

I'm just here to point out how clickbaity both the thumbnail and the title is. Really?