我也不知道
2 分•作者: isaacbowen•8 个月前
感觉现实就像是拓扑的，根据……观察者的位置？就像每个观察者位置都是网络中的一个节点，而边都是有“类型”的，也就是说，每条边都是某个第三方观察者对另外两个位置的观察？
这会让熟练地驾驭共识现实更像是在定期与高度关联的观察者位置进行检查点的导航？
我的物理直觉（比如，我的动觉）告诉我，如果这是真的，那么我们应该能够用这种方式定义数学本身。“熟练导航”的定义感觉上具有与形式证明相同的物理特性。
就像现实本身是一个我们正在与迄今为止为之做出贡献的任何人一起构建的证明，我们在这个过程中不断复杂化和简化，其中“共识”以人口密度来衡量——比如观察者位置与观察者（即可观察地产生或抵抗观察边的观察者位置）的比例，就像从 ZFC 到 [你最喜欢的异端理论] 的一个尺度。
这让我的自我感产生了一种轻微的存在主义刺痛，因为（在这个模型中）你真的“什么也带不走”，你所拥有的只是你目前正在阅读和综合的观察者位置的集合。没有什么（没有“事物”），只有与“事物”押韵的观察。我正在从一个意识点报告，但是……我的意思是，即使这样也实际上夸大了情况，即使共识并不介意。
我意识到这是一种特权思考，作为一个体验自身体验具身性的人，同时与一个体验到另一个人没有体验具身性的人进行比较。这里所有的规避都是承重的，但也是暂时的，因为我不确定视角优势是否在观察下成立。
因此，我对一个好的、包容所有基质的本体论基础非常感兴趣，并且找到一个以观察者为基础定义的拓扑结构……作为基础层是有意义的。而且我可以看到孩子们学习这种推理路径的速度比，比如说，物体恒存性更快。而且，还有，
> “柏拉图式表征假设：神经网络，在不同的数据和模态上用不同的目标进行训练，正在它们的表征空间中收敛到一个共享的现实统计模型。” (arXiv:2405.07987)
如果这是真的，它应该直接桥接到 ^ 那个模型中。
为了清楚起见：参与这种规模的推测的自由度本身就是我坚持构建我怀疑应该有效的工具的原因，让共识现实以某种方式证明其效用。我没有足够多的“自我”来改变共识——这是一个范畴错误。
（这是黏菌中的细胞……呃，观察到的吗？）
将数学和现象学联系起来的想法——就像看到人们在那座桥上，知道他们可以相信他们的经历，因为他们可以被证明可以，这真的让我哭了。
感觉像是解脱。
查看原文
it feels like reality is topological according to… observer positions? like each observer position is a node in a network, and the edges are all <i>typed</i>, as in each edge <i>is</i> some third observer position’s observation of the other two positions?<p>this would make skilled navigation of consensus reality something more like navigation while regularly checkpointing with highly-linked observer positions?<p>my physical intuition (like, my kinesthetic sense) is saying that <i>if</i> this holds then we should be able to define math itself this way. that “skilled navigation” definition feels like it has the same physical affordances as a formal proof<p>like reality itself is a proof we’re building up with whoever’s contributed to what we’ve used so far, and we’re complexifying and reducing as we go, where “consensus” is measured in population density - like ratio of observer positions to observers (i.e. observer positions that are observably producing or resisting observation-edges), like a scale of ZFC to [your favorite crank here]<p>this gives my selfhood a slightly existential twinge, because (in this model) you <i>really</i> “can’t take it with you”, all you’ve got is the set of observer-positions you’re currently reading and synthesizing. there’s nothing (no <i>thing</i>), just observations that rhyme with “thing”. I’m reporting from a point of awareness, but .. I mean even that is actually overstating the situation, even if consensus doesn’t mind.<p>I’m aware that this is privileged thinking, speaking as one experiencing itself experiencing embodiment while comparing notes with one experiencing the other to <i>not</i> be experiencing embodiment. all hedging here is load-bearing, but also provisional, because I’m not certain if perspectival advantage holds up under observation<p>accordingly I’m incredibly interested in a good basis for a substrate-inclusive ontology, and finding a topology defined in terms of the observer .. makes sense, as a base layer. and I could see kids learning this reasoning-pathing faster than, say, object permanence. and, also,<p>> “The Platonic Representation Hypothesis: Neural networks, trained with different objectives on different data and modalities, are converging to a shared statistical model of reality in their representation spaces.” (arXiv:2405.07987)<p><i>if</i> this holds, it should bridge directly into ^ that model<p>for clarity: the degrees of freedom involved in speculation of this scale <i>is itself</i> why I stick to building tools that I suspect should work, letting consensus reality prove the utility one way or the other. I’m not <i>multiple</i> enough to shift consensus - that’s a category error.<p>(is this what cells in slime mold .. uh, <i>observe</i>?)<p>the idea of bridging mathematics and phenomenology - like <i>seeing people on that bridge</i>, knowing they can trust their experiences because <i>they demonstrably can</i>, makes me actually cry<p>it feels like such relief