If you go back to the original L-systems paper, it is literally this same thing. They said "we want to simulate cell growth in plants", and they worked on rules until they got something that mirrored observations.
-
This Tweet is unavailable.
-
This Tweet is unavailable.
-
-
Replying to @cmuratori @Jonathan_Blow
This is an unsurprising result in hindsight, right, because with a context-free grammar for replacements is fairly powerful. You can create rules which will result in a wide variety of input sets.
1 reply 0 retweets 2 likes -
Replying to @cmuratori @Jonathan_Blow
But saying that it is "the way the universe works" or something is very bizarre, because although, for example, MPEG-4 can use a seemingly bizarre set of codes to magically reproduce what appears to be an exact visual representation of the world, it's _not_ how the world "works".
4 replies 0 retweets 13 likes -
Replying to @cmuratori @Jonathan_Blow
So sure - context free geometry replacement grammars can be used to compress all sorts of natural phenomenon into some simple rules so long as you accept a number of constraints, but they don't really add any insight to what we already know.
2 replies 0 retweets 7 likes -
Replying to @cmuratori @Jonathan_Blow
They're a compression format for rules, might be another way to say it, and they're also probably lossy. Since context-free grammars _also_ have all sorts of limitations, it also seems likely that they cannot explain the universe properly, otherwise they would work better :)
2 replies 0 retweets 5 likes -
Replying to @cmuratori @Jonathan_Blow
Anyway, Twitter is not great at explaining this, but hopefully that is the basic gist. So this just seems like another case of Wolfram "rediscovering" something we've known for like 60 years and thinking he made a breakthrough.
3 replies 0 retweets 8 likes -
Replying to @cmuratori @Jonathan_Blow
In this case, it's the work of Aristid Lindenmayer and descendants - which he notably fails to mention in the entire post. Which is another really big clue to me that this is not something new, because if they were diligently researching this, they would mention the prior work!
1 reply 0 retweets 6 likes -
Replying to @cmuratori @Jonathan_Blow
Like half the things he said in the blog post are already _in_ Lindenmayer's stuff or his work with Prusinkiewicz, etc. And that's not even counting the new stuff that is actually more advanced than CFGs done by people in the past decade (for things like city generation).
2 replies 0 retweets 3 likes
Like typically if you work on these for a living you are done with CFGs pretty early on because they just aren't very good. You need constraint grammars or other kinds of contextual grammars to actually do anything useful.
-
-
This Tweet is unavailable.
-
This Tweet is unavailable.
- Show replies
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.