YES!
One can believe that LLMs can do amazing things and are useful, *without* believing they are anywhere close to human-level intelligence (even if they are superior to humans in a few tasks). One can believe that LLMs will give new tools to people with bad intention *without*
ETHICS
-
LLMs capabilities limits and security risks balance
By
–
-
Cognitive Ability Decay: From Exceptional Memory to High School Normalcy
By
–
(My ability here decayed to something more approaching “has an anomalously good memory” by approximately late high school. I was both sad but also happy because it eased some social difficulties.)
-
Class Credentialing Institutions and Educational Gatekeeping Mechanisms
By
–
I sometimes wonder if I am too cynical about our class credentialing institutions and their funding and gatekeeping mechanisms. Then, I remember they are our class credentialing institutions, and so the baseline level of cynicism should be high.
-
Policy Design Benefits Elite, Harms Others: Known Side Effect
By
–
“Wow this seems like a policy designed to work spectacularly well for the class that wrote it and blow up terribly for other people.” More of the first half of that; the second is a side effect. It’s a known side effect, in the literature, but was not considered a blocker.
-
LLMs at 100x Capability: AGI or Advanced AI Automation?
By
–
If current LLMs were to become 100 times more capable, seamlessly integrating into workspaces to automate tasks, who could argue they're inferior to human labor? The public would hail it as AGI, though by @sama's standards—unable to perform nuclear fusion—it would merely be AI
-
Anthropic praised as EA’s most significant contribution
By
–
yep that's fair. Anthropic is the only good thing to come out of EA.
-
Tech Pioneer Focused on Making AI a Force for Good
By
–
A force for #AI good https://
wsj.com/tech/ai/a-tech
-pioneer-focused-on-making-ai-a-force-for-good-d60f8f10?mod=Searchresults_pos1&page=1
…
in today's WSJ @EmilyBobrow -
Research Leaders Should Work with Smarter Minds
By
–
I dunno, man. As a leader of research groups, I'm very familiar with *not* being the smartest entity in the room. Let alone on earth. One should strive to work with people who are smarter than themselves (or at least, smarter in some areas), and one shou,d enjoy doing so.
Why -
Pandemics and AI risks are fundamentally different phenomena
By
–
Some people are assimilating the risks associated with pandemics with the risks associated with AI progress.
There is a fundamental difference. Pandemics are natural phenomena.
We can take preventive measures before they happen and plan for corrective measures in case they do. -
AI Defense Techniques Mirror Offensive Capabilities
By
–
That's pretty much true for any nefarious use of AI: if bad actors can use AI for something, the same AI techniques can be used to protect against them.