@alz_zyd_@tenobrus Well, i think the actual mechanics are probably important. LLMs only very slightly resemble brains at only the highest level of abstraction. I don't think you can even remotely compare an LLM to a simulation of the brain, let alone to a mechanical replica.
@redsphagnum@swdevservice@MasterTimBlais So there's this field called biology. And they study the things I'm talking about. No I don't have an answer for the exact physics necessary for life. Obviously. But all living things have things in common that help us classify them as such. Like DNA. Don't be pedantic
@swdevservice@MasterTimBlais There is very little consensus on what constraints are implied by either assumption. I don't think you can so confidently say that. Also I don't see how it's anthropocentric? I'm considering *at least* everything in the animal kingdom
@swdevservice@MasterTimBlais Literally every single entity we believe to be conscious is on biological substrate, and the *only* exception is maybe LLMs (according to people who believe they are conscious). The null hypothesis is 100% that consciousness depends on a biological substrate.
@abdimoalim_ Also LLMs are pretty decent at making sense of badly formulated math. They struggle to formulate things in the most intuitive way, but they give answers without ambiguity and lots of natural-language explanation, which is often good enough
@abdimoalim_ With good books, it shouldn't feel this way, but most books are not good. Robust intuitions on the fundamentals is the best way to navigate bad books. Don't move on from a page unless you understand the concepts deeply, and it will save time in the long run.
@tenobrus@Strife212 Why should we expect even a perfect simulation of the brain down to the molecular level to be conscious? A perfect simulation of a heart doesn't actually pump blood...
@NPCollapse What's the plan to legislate this? How is superintelligence research defined, what oversight is necessary to detect it, and what actions can be taken to prevent/regulate it?
@NPCollapse There are very real issues that big AI companies pose to the world already and right now. A movement for better AI should reflect that, and that involves real material politics, not fighting some imagined hypothetical super intelligence.
@zhoro_x Those two fritz kola guys are so unbelievably talented. They have dozens of soda flavors and each and every one is a masterpiece. The mint-rhubarb one goes hard af
@squeezoal@Aella_Girl Idk i couldn't tell you, I'm not unilaterally anti-AI, but LLMs being autocomplete is certainly a good argument for why LLMs are fundamentally incapable of being AGI