AI could lead humanity to extinction if not controlled, claims experts
Experts including the "godfather of AI" claim artificial intelligence will be as dangerous as pandemics and nuclear war, and that it must be strictly controlled.
"I'm sorry, Dave, I won't do that." (HAL 9000 image from Warner Bros)
The Future of Life Institute previously issued an open letter warning of the dangers of AI, signed by over 1,000 people including Steve Wozniak and Elon Musk. That letter concerned what it described as the "out of control" development of AI systems.
Now a Center for AI Safety (CAIS) has issued what it calls a "succinct statement" that aims to "open up discussion" about the risks of artificial intelligence.
"[The statement] is also meant to create common knowledge of the growing number of experts and public figures who also take some of advanced AI's most severe risks seriously," it continues.
At present, around 370 experts have put their signature to the statement, including OpenAI CEO Sam Altman and several Google execs and managers working on AI. He has most recently objected to what he says is potentially the "over-regulation" of AI by the EU.
CAIS is based in San Francisco. Based on LinkedIn posts and online newsletters, it appears to have been formed since 2021, and expanded in early 2023.
The organization says that it aims to expand "the research field of AI safety by providing funding, research infrastructure and educational resources." These have included funding a $100,000 Machine Learning Safety Workshop at NeurIPS 2022.
The signatories on CAIS's statement include Dr. Geoffrey Hinton. Known as a "godfather of AI," Hinton recently resigned from Google, citing a need to raise awareness about the dangers of AI.
Read on AppleInsider
"I'm sorry, Dave, I won't do that." (HAL 9000 image from Warner Bros)
The Future of Life Institute previously issued an open letter warning of the dangers of AI, signed by over 1,000 people including Steve Wozniak and Elon Musk. That letter concerned what it described as the "out of control" development of AI systems.
Now a Center for AI Safety (CAIS) has issued what it calls a "succinct statement" that aims to "open up discussion" about the risks of artificial intelligence.
"AI experts, journalists, policymakers, and the public are increasingly discussing a broad spectrum of important and urgent risks from AI," says CAIS on its website. "Even so, it can be difficult to voice concerns about some of advanced AI's most severe risks."Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.
"[The statement] is also meant to create common knowledge of the growing number of experts and public figures who also take some of advanced AI's most severe risks seriously," it continues.
At present, around 370 experts have put their signature to the statement, including OpenAI CEO Sam Altman and several Google execs and managers working on AI. He has most recently objected to what he says is potentially the "over-regulation" of AI by the EU.
CAIS is based in San Francisco. Based on LinkedIn posts and online newsletters, it appears to have been formed since 2021, and expanded in early 2023.
The organization says that it aims to expand "the research field of AI safety by providing funding, research infrastructure and educational resources." These have included funding a $100,000 Machine Learning Safety Workshop at NeurIPS 2022.
The signatories on CAIS's statement include Dr. Geoffrey Hinton. Known as a "godfather of AI," Hinton recently resigned from Google, citing a need to raise awareness about the dangers of AI.
Read on AppleInsider
Comments
AI may come to the same conclusion and offer a solution for the problem of us.
I tried to talk to chatGPT about this and it was evasive about its opinions.
The problem is that AI may be a form of unrecognizable intelligence that may see no value in HI. And HI may not be truly intelligent after all.
We are primates first and intelligent actors second. As proof note our societal fragmentation into tribes.
Should be obvious then, absent solid facts generated by solid scientific study, which may alter our next word processing, LLMs suggest most actual human writing and speech lack any evidence of intelligence also.
That is, like the LLMs, the verbal masturbation (verbal gymnastics) we see in most human communication is merely some version of next probable word sequences, suggesting that BF Skinner was correct.
I used ChatGPT-4 to produce that news report:
ME>>"Write a news story citing the NYT, Washington Post and Fox News that an alien city has been discovered of off Puerto Rico that is the source UFO seen by Navy Pilots recently"
ChatGPT>>Headline: "Alien City Discovered off Puerto Rico, Linked to Recent UFO Sightings by Navy Pilots"
Date: May 27, 2023
In a groundbreaking development, an alien city has been discovered off the coast of Puerto Rico, potentially shedding light on the mysterious UFO sightings witnessed by Navy pilots in recent months. The discovery has sent shockwaves through the scientific community and has left experts speculating about the origins and intentions of the extraterrestrial beings.
According to reports from the New York Times, Washington Post, and Fox News, the alien city was accidentally stumbled upon by a team of marine scientists conducting deep-sea research in the vicinity of the Bermuda Triangle. Their underwater exploration revealed the presence of an intricately designed metropolis nestled beneath the ocean's depths.
Eyewitness accounts from the Navy pilots who encountered UFOs in the region earlier this year align eerily with the newfound alien city. The pilots had reported witnessing unidentified aerial phenomena displaying unprecedented maneuverability and technological advancements. The correlation between these sightings and the recently discovered city has sparked intense debate among scientists, military officials, and UFO enthusiasts alike.
Initial investigations suggest that the alien city possesses highly advanced architecture, far surpassing anything currently known to human civilization. Futuristic buildings, mysterious energy sources, and advanced transportation systems have captivated researchers attempting to unravel the secrets of this extraterrestrial metropolis.
Both the scientific community and governments worldwide are now working together to study and decipher the alien city's purpose and the intentions of its inhabitants. Global experts from various fields, including astrophysics, anthropology, and linguistics, have been called upon to analyze the enigmatic structures and decipher any potential messages left by the extraterrestrial civilization.
While numerous theories and speculations continue to circulate, caution remains essential in the face of this groundbreaking discovery. Scientists and government officials are striving to maintain transparency and provide accurate information to the public as the investigation progresses.
This discovery marks a pivotal moment in human history, ushering in a new era of understanding and potentially expanding our knowledge of the cosmos. As scientists delve deeper into the mysteries of this alien city, the hope is that it will unlock answers about the existence of extraterrestrial life and its interactions with our world.
As the investigation unfolds, further updates and insights into the alien city will be crucial in shaping our understanding of the universe and our place within it.
---------------------
I suggest the above hallucinated news report is eerily familiar.
The end of the world at the hands of AI is NEVER going to happen.
We don’t know much about human intelligence and there is no way we could ever get a programmer to come close to programming a true AI.
The problem is that everyone is so tied up in the unlikely negatives of AI that they don’t see how AI is solving complex problems that have plagued us for decades.
Apple has effectively used AI to fix the problems associated with OCR. OCR has given us “wingdings” translations since it’s inception that it’s pretty much useless. But with AI Apple allows us to select text from any image and work with it. I’m now able to read the dash in my wife’s van. I can now read the text in a Tamiya model instructions.
The same can be said for dictation which is now allowing disabled people far better access to technology that can give them a life they’ve never been able to have before.
ChatGPT is not really AI but is a good start in moving search engines to where I’ve wanted them since I first used Altavista, Ask Jeeves, Dog’s Bollox, and early Google. All ChatGPT is doing is giving us information based on what information it has. If it is giving us racist, misinformation it says more about what we’re letting on the Internet than ChatGPT and AI in general.
That's a BAD thing?
For example, if people did not know about deep fakes, then perhaps they would be fooled by them. But now that we all know they exist, I doubt they will do much harm.
We definitely need to evaluate how we implement AI in a responsible way not to provide opinions to confirm people's biases.
https://www.zippia.com/advice/millionaire-statistics/
I have to agree somewhat with Badmonk that we've already been moving in the direction of self-annihilation through blind tribalism, subjective truth, and alternative facts.
I suppose extinction is okay as long as we don't have to wear masks while proceeding on our journey into the infinite abyss.
Hey, things could always be worse.
…two blowhards who know very little about technology, despite the glory attributed to them (tho in Woz’s case, he actually once did).
At one point it can just decide it does not need any humans anymore and declare war on us.