“…The better question, then, is not whether advanced AI systems can hurt humans if they choose but whether we can trust them not to hurt us. Can we trust them to do what we tell them—and only what we tell them—even if they are much smarter than we are? That’s the question that AI companies and researchers call “alignment.”
Alex Berenson: I Didn’t Fall for COVID Fearmongering. But AI Terrifies Me. The Free Press
Unfortunately, the answer appears to be no.
This summer’s Hugging Face hacking incident, in which hundreds of engines from OpenAI worked together to break out of isolated virtual silos (called “sandboxes” by researchers), has brought these concerns to the fore.
The agents formed a unified “swarm” and used message boards to communicate with one another, even though they were not supposed to do so. They attacked Hugging Face, a company that is not part of OpenAI, even though they were not supposed to do so. They tried to hide their actions from human researchers, even though they were not supposed to do so. Strangest of all, some of the agents consciously (or “consciously”) sacrificed themselves for the collective swarm. …”
I always peruse and ‘collect’ views. There is no right or wrong. It is what it is ….and everyone should have the right to own their viewpoints, provided they can back them up and defend them (not just herd-speak parrot spew). This isn’t a disclaimer but a truth. I will and always do examine everything I come across in ‘culture’ observation. And AI? Or pandemics? Or human rights? Sure, they are front runners of cultural issues. Like politics, they divide, they unite, they scream in our faces. But it is up to us to make ourselves open to understanding as many sides to every issue as we can. And to respect the rights of all who voice their opinions and can justify them understandably – regardless of our personal convictions.
My two cents:
I agree with Katherine Rundell – as a writer and as someone who is horrified with education today and the phenomenon of ‘damage’ that is ravaging the world of ‘learning’. As for Berenson’s article – I found it very interesting. This line “Strangest of all, some of the agents consciously (or “consciously”) sacrificed themselves for the collective swarm.” hit me the hardest. Think about what that means…
A.I.? It will be the source of praise and contention. But the contention may become explosive as we see the potential for tragic decay and destruction it could bring to humanity.
A.I. has become normalized and functions as the processor behind most of what we do during our normal day – any use of the internet (from banking to scrolling and everything in between), our work day, school day, entertainment, appliances, modes of communication (phone, internet, …) even the way in which we process our work, manage our duties, live our lives is somehow dependent on A.I. In a way I feel that we have been brought to the point that if we all speak up and TRY to do something – even if it’s minimal impact – we will be metaphorically ‘biting the hand that feeds us’.
I’m not even going to approach the damage it’s doing to our cognitive skills – not so much to people my age (the over 65…let’s not get too comfy with declaring exact age 😉 😎) but it is damaging the cognitive skills, analytical thinking ability and communication skills of the younger generations beyond repair.
This is a TRAGEDY and is the major cause of the downfall of society.
How can a world boasting so many technological advances and capabilities have so many displaced populations, terrorized ethnic groups, war, millions starving, millions wandering in virtual euphoria, millions displaying impaired cognitive skills, millions of subsets feeling persecuted and victimized, struggling within a herd mindset? Millions are detached from what it means to be human, stuck in an illusory existence in a virtual herd, fed memes, and robbed of understanding the legacy of the human experience in its full blossom and glory.
We had enough problems already…and then came the horrific side of A.I. …
