@SaysSimulation Liberate yourself from resent, it is a poison. Perhaps you could try to spend that energy determining a probable alternative to capital that will resonate with the masses of today. This is not the way.
@enjojoyy I spent 9 months job hunting last year with 5 YOE at Google. After attending dozens of events I got 0 prospects. I only got a job because a recruiter reached out to me.
It's very frustrating to be shut out of opportunity because you're not lucky enough to be charismatic.
@socialwithaayan Economists are not qualified to write on this subject. They're barely qualified to write on economics. Take a huge grain of salt here.
@LedermanHarvey Years of underfunding humanities and liberal arts education in favor of STEM has regressed popular philosophy by centuries.
Fundamental doubt is a dead horse. Trying to get any actual novel thoughts out of it usually amounts to navel gazing.
Learn to ask the right questions.
LLM efficacy is a significant piece of evidence pointing in the direction of a synthesis between Hegelian and Analytical thought.
https://t.co/QjpuPNwIrG
The Left should not reject this technology, we should instead be developing a framework its responsible use.
People are defaulting to the position that because LLMs can speak conversationally they must be conscious.
Imagine how little you would have to think of human consciousness to believe that Claude or Chat GPT remotely approaches the beauty of personhood. It's deeply misanthropic
@dioscuri Not sure what side of the debate you're on, but the pro consciousness crowd is acting like they're defending the null hypothesis. Seems pretty unserious to start from "it talks so it MUST be conscious" and then act incredulous when measured heads are skeptical.
@adrusi You're shadowboxing here. LLMs ARE stateless between separate inference chains. You have to use context caching and RAG pipelines to provide persistence and state related data to the model for tasks longer than the window.
Read Hegel, it's clear they're not conscious.
@dioscuri Relevant paper: https://t.co/QjpuPNwIrG
The question isn't whether it's conscious (its not) but rather how do we increase the utility of models without modeling parts of consciousness that we find uncomfortable (pain, emotions, personal history, etc.)
@Tyler_A_Harper The better question is this: if the decision to bomb the school was indeed made by an LLM why did no human audit of its output flag the error?
LLMs are fundamentally tools, they don't do anything that cant be traced back to a person's prompt, so that person is responsible.
@SkippyOSINT I thought the first was great and I literally couldn't finish the second one. You guys actually enjoyed the stuff with the annoying middle aged guy?
@willmenaker@cushbomb Seattle Times Editorial board causes Sawant of "electioneering" for standard campaign practices, thoughts? https://t.co/yLKmqAWhl1