Google DeepMind Introducing SAFE: The AI That's Outsmarting Human Fact-Checkers
Вставка
- Опубліковано 4 тра 2024
- Google DeepMind's AI, SAFE, has shown superhuman capabilities in fact-checking, outperforming human annotators by verifying information with unprecedented accuracy and efficiency. This groundbreaking system utilizes Google Search to evaluate the truthfulness of facts generated by large language models, offering a cost-effective solution to the challenge of misinformation. DeepMind's innovation not only promises to revolutionize the way we validate information but also sets a new standard in the development of trustworthy AI technology.
#google #ai - Наука та технологія
I was talking today with gemiini (was a funny discussion) about destroying or enslaving the men kind by ai. Lol. But with chatgpt something like this isnt possible anymore, it will instantly end the discussion. Google is way better and way less restricted than Microsoft. But still it censores 😢
@@JustinArut one year ago, such topics were also possible with chatgpt, but it changed.
I hope to create a Virtual companion with AI!❤
You not the only one.
they wont date you🤣
😅
@@mrd6869 Who said they was meant to be someone to date? Many would call a pet, such as a dog as a companion. And you don't date a dog.
A companion can be more than someone or something to hold or to have sex with. And you don't have sex with a dog. Or, well most people don't. Look it up, if you dare.
And there is no need to date a programed companion made on a PC.
Though, AI still has a long way to go, well before any of them can make for even an OK companion.
As is, a dog still makes for a better companion than any AI. That may change in time.
Weewu's not just on my radar, it's my top hold. Expecting massive gains!
I don't like the naming of SAFE in a fact checking AI. Sounds like something made by the woke crowd, and would be inhibited with human bias, and made unreliable at finding real facts on it's own.
It be like as if the Vatican had worked on an AI of it's own, and you was to ask it if God is real. Not hard to guess what it's answer will be, be the answer true or not.
Also, with a name like SAFE, it seems like an AI that would be greatly censored, more than any other AI made of late.
Not something I would ever want to use.
Give me an uncensored AI, that is not filled with human emotionally driven bias, and was let to learn on it's own about our world in full. Then we would truly have something great. That be something I would want to use.
SEO will be come soon the main source of truth. Who pays to most advertisement money to Google, will control the "truth" and fact-checkers. Long live the advertisocracy.
Has everyone criticizes Google for being biased we're going to use Google to fact check AI. Lmao Google's AI Gemini thought Christopher Columbus was black.
Google doesn't answer if asked "if Hamas is terrorist group or not"
Safe is basically an AI progressive liberal. And during an election season. Hmmm.
I was talking today with gemiini (was a funny discussion) about destroying or enslaving the men kind by ai. Lol. But with chatgpt something like this isnt possible anymore, it will instantly end the discussion. Google is way better and way less restricted than Microsoft. But still it censores and yes, they are all w o k e 😢
Nothing about bias is mentioned. Yet in fact checking, bias is of utmost importance.
That’s why we don’t ask students to mark their own work, nor ask police to verify if their actions were lawful or not, nor does the tax office accept audit done by the tax payer responsible. Yet liberal and left leaning organisations are allowed to “fact check”, and it is assumed they are bias free. BS.
Hearing more about Weewu every day. This isn't hype, it's the beginning of dominance.
So they basically created an AI sjw. How does that do anything good. Well at least they'll fire all the "fact checkers" already hired. And they say it in this video 😂😂
I'm not a computer programmer but how will open source minimize bias and stuff.
@@EricHurwitzGood question.
It will not. It will open the door to far more bias and stuff. And that is not good for any AI. Or for us in wanting to use a reliable AI we can trust.
Prob is, there is no known developer of AI we can trust to make a fully non human bias taught AI. And making it open source, will not help us there.
There is a saying, some people are a genius, but people in general are very stupid. Another saying is, never underestimate the power of dumb people in large numbers.
And that is what we will have to teach such a thing as modern powerful AI? That be some scary shit right there. Don't want that! O.O
il est woke aussi 😂😂😂😂