Center for Bits and Atoms: Difference between revisions
added link to long 2h interview that I've fully watched recently |
Focus on growth plan (cellular automata) |
||
| Line 2: | Line 2: | ||
Just some notes for now … | Just some notes for now … | ||
== Ready-fire-aim philosopy == | |||
Discoveries and scientific progress by | |||
* working on practical problems (where ideas come from (?)) | |||
* failing but discovering something different | |||
* which can be applied elsewhere eventually | |||
== Focus on growth plan (cellular automata) == | |||
Yes, I agree that the [[von Neumann bottleneck]] from <br> | |||
separating storage and compute (arithmetic logic unit ALU) is very bad. | |||
But personally I think maximally distributed compute (cellular automata) | |||
is only useful in special application cases. E.g. far term [[utility fog]]. | |||
I don't like the imperative nature of "what to do" step by step in a "growth plan". <br> | |||
I prefer the denotative nature in a direct description. <br> | |||
See: [[Decompression chain]] & [[Constructive solid geometry]] | |||
Well, a "growth plan" could be an automated step within a decompression chain. <br> | |||
Actually extrusion in a [[nanofactory]] chip is kind of a growth plan. <br> | |||
Just that the assembling nano and micro-machinery does not push itself along but rather stays put and pushes the product. <br> | |||
There is a bio-analogy to the (topologically 2D) "meristem" "germinal epithelia" "germinal zones" "germinal cells". <br> | |||
See: [[Complex surface nonplanar nanofactory chip growthfronts]] <br> | |||
Also see: [[Misleading biological analogies that should be avoided]] | |||
Reminder: Full volume growth would lead to | |||
* excessive productivity due to [[higher throughput of smaller machinery]]. | |||
* complications in feed-stock supply needing to cross product volume | |||
== Related == | == Related == | ||
Revision as of 01:24, 9 June 2023
Just some notes for now …
Ready-fire-aim philosopy
Discoveries and scientific progress by
- working on practical problems (where ideas come from (?))
- failing but discovering something different
- which can be applied elsewhere eventually
Focus on growth plan (cellular automata)
Yes, I agree that the von Neumann bottleneck from
separating storage and compute (arithmetic logic unit ALU) is very bad.
But personally I think maximally distributed compute (cellular automata) is only useful in special application cases. E.g. far term utility fog.
I don't like the imperative nature of "what to do" step by step in a "growth plan".
I prefer the denotative nature in a direct description.
See: Decompression chain & Constructive solid geometry
Well, a "growth plan" could be an automated step within a decompression chain.
Actually extrusion in a nanofactory chip is kind of a growth plan.
Just that the assembling nano and micro-machinery does not push itself along but rather stays put and pushes the product.
There is a bio-analogy to the (topologically 2D) "meristem" "germinal epithelia" "germinal zones" "germinal cells".
See: Complex surface nonplanar nanofactory chip growthfronts
Also see: Misleading biological analogies that should be avoided
Reminder: Full volume growth would lead to
- excessive productivity due to higher throughput of smaller machinery.
- complications in feed-stock supply needing to cross product volume
Related
- The three axes of the Center for Bits and Atoms
- Simultaneous prototyping across scales
- Robot substructure gradients
External links
Websites:
- Landing page of Center for Bits and Atoms, CBA about page
- MIT: Center for Bits and Atoms
- Wikipedia: Center for Bits and Atoms
Videos:
- 2023-05-25 (~2hours): Neil Gershenfeld: Self-Replicating Robots and the Future of Fabrication | Lex Fridman Podcast #380
Papers & theses: