Matter cannot pass through matter so we say oh there are atoms and there's a lot of space in between what's in the space a vacuum okay fine. So I can see the anup reina mountain through the clouds miles away which is not always the case usually only in the morning when the sunlight hitting them is strong enough and then they disappear behind the clouds however what is it that makes the sunlight or light what we think of this photon strong enough in the morning then becomes more diffuse so this is an entropy problem essentially this is about low entropy directed sunlight about that we say bounces off the mountains, is able to pass through even some cloud cover early in the morning when the sun is directed nuff but this is also an Information on relationship problem so think about the commutative property or associative I forget the idea is if a has a relationship with be and b has a relationship with c, then as impact on sea let's say c is the observer depends on b that that intermediary and this explains why we say neutrinos can pass through matter. Which doesn't it depends on that those properties as associative or relational properties to determine what the characteristics of matter actually are so we got it all wrong we should be thinking about matter in relational terms to determine its properties. And those properties those patterns of relationships think about it as patterns of relationships Pure math everybody remembers the commutative associative properties even though I have no idea what they are I think about them from um a database perspective and how you have relational databases that actually are good um they are a good model here. Because if we have a key value in one part of a database a one to many that won value propagate throughout the database and effectively creates what we think of this matter. So if we look at reality like a well formed relational database we can start to see how these patterns now propagate throughout reality and in fact created you know in a database the the end table might just be a key value but because of its relational properties it actually gets assigned a name or value or it could be anything really could be an entire. Could be it could be an object actually think about that way if you've got a an object let s say an image that is the look up table. That image then appears anywhere in the database reality is a database. And of course the question is what creates the database well it could be self propagating you know it could there could be these initial sort of random seeds that we think of his evolution that then can over a sequence and iteration get replicated as more complex phenomena that s actually not that hard to um to conceive of and it makes a lot of sense. And now let s go for the wind with god ellen limits when we can only think in terms of structured data and relational databases before neural networks and large language models helped us break free of that and adopt a different paradigm that can then look back and say oh wow this is actually a good model where the where the power of languages also its limitation it doesn't readily reveal that structure. And that s all its power but then sort of contrast to right if we contrast a relational database which could be seen as a proxy for heuristics with the network graph of uh of a large language model and neural networks any kind of network based graph. Those 2 are very complementary they are different but complementary and they transcend together they are very powerful and they can transcend the limits of each other.