The first comment (kudos for open review) links to a post that says some of this at greater length, but to repeat my own reaction: "There's nothing in there about alignment. The proposed motivational system is internal-system-health reward with nothing about caring for humans."
SAFETY
-
Zero Accidents: When Autonomous Technology Becomes Trustworthy
By
–
If the technology becomes so robust and reliable that, over a long time period there is essentially zero accidents, I’d say this is something that could be considered #TeamHuman
-
Biden Issues Executive Order on AI Safeguards and Regulations
By
–
Biden Issues Executive Order to Create A.I. Safeguards: The sweeping order is a first step as the Biden administration seeks to put guardrails on a global technology that offers both great promise and significant danger. https://
nytimes.com/2023/10/30/us/
politics/biden-ai-regulation.html?utm_source=dlvr.it&utm_medium=twitter
… #Technology #Tech #FutureTech -
LLM Capabilities and Existential Risk Concerns
By
–
I'm sure we can find a few LLM fanbois who believe this could actually work.
But they won't try it for fear that humanity will immediately be destroyed thereafter -
AGI Safety Victory: Ensuring Beneficial Superintelligent Expansion
By
–
I can't quite say that I'd declare victory, because I need to know what use is going to be made of the rest of the accessible galaxies, if they all turn into cities of sapient life having fun. But that those AGIs were at least safe for humans–I am happy to call that game won.
-
Greatest threats to US national security in AI era
By
–
Which is the greatest threat to US National Security?
-
Subtle Differences in AI Assistant Greeting Responses Analysis
By
–
The difference between a world that starts with 'Hi there! How can I assist you today?' and one that begins with 'Hello! Is there anything I can assist you with today?' would be quite peculiar.
-
ASI Competition for Resources Threatens Human Survival
By
–
We compete for matter, negentropy, and humans creating actually-competitive additional ASIs if we're allowed to stick around. Or more plainly, if an ASI boils Earth's oceans as coolant for computation, humanity doesn't survive that.
-
Prioritizing Harmlessness After Helpfulness in AI Model Training
By
–
This is precisely why we should work towards harmless after achieving helpfulness, than the other way. Models trained by Perplexity are going to be doing this.