4 ms·

That's also the beauty of evolution–there's no goal in that search algorithm, just open-ended incremental exploration which eventually grows beautiful things.

That's also the beauty of evolution–there's no goal in that search algorithm, just open-ended incremental exploration which eventually grows beautiful things. See Kenneth Stanley's book Why Greatness Cannot Be Planned: The Myth of the Objective: https://www.goodreads.com/book/show/25670869-why-greatness-cannot-be-planned https://www.goodreads.com/book/show/25670869-why-greatness-c...

lovestory 1y ago
This is the second time someone has recommended this book on HN so for sure it can't be a bad read. I will be starting it tonight
- michaelsbradley 1y ago
  The book may be a good read or bad one, I am unfamiliar with it. However, terrible recommendations are made on HN all the time, as well good ones, so I’d caution against “recommended a few times on HN” (on its own!) as an indicator of much anything. --- EDIT: changed wording from “multiple times” to “a few times” to be more clear.
  - nine_k 1y ago
    It may be an indicator of popularity / mindshare of certain ideas. It says little about the ideas' substance, but may still be very important in practice, if you deal with people.
    - michaelsbradley 1y ago
      The git cli tool is mentioned countless times over many years on HN, surely an indicator that it’s in popular use by the commentatorship here. I’ve seen many (seemingly informed) mentions of the jj cli tool in the last year — not unreasonable to conclude interest in it is growing among folks here, and enough to pique my curiosity. I responded to: This is the second time [I’ve seen that] someone has recommended this book on HN
- gsf_emergency_2 1y ago
  From Ken Stanley's wiki >He left both Uber AI Labs and the University of Central Florida in 2020 to lead the Open-Endedness team at OpenAI as a Research Science Manager.
ChrisMarshallNY 1y ago
I've always found that an "evolutionary" approach to design is important[0]. [0] https://littlegreenviper.com/evolutionary-design-specification/ https://littlegreenviper.com/evolutionary-design-specificati...
ants_everywhere 1y ago
> That's also the beauty of evolution–there's no goal in that search algorithm, Yes there is, there's a fitness function, both in biological evolution and evolutionary algorithms
- IAmBroom 1y ago
  That is not a goal. It is an adaptive strategy to meet changing constraints.
  - layer8 1y ago
    I’d argue it isn’t a strategy either. You can’t not be subject to it.
- layer8 1y ago
  There’s no goal in biological evolution. Being subject to natural selection based on the present environment is an inevitability. You could argue that mutation isn’t inevitable, but its presence and degree is subject to natural selection as well.
_lex 1y ago
Incorrect. The search algo has a goal: to find out. You just think it has no goal because you didnt think in quantum.

lovestory 1y ago

This is the second time someone has recommended this book on HN so for sure it can't be a bad read. I will be starting it tonight

michaelsbradley 1y ago
The book may be a good read or bad one, I am unfamiliar with it. However, terrible recommendations are made on HN all the time, as well good ones, so I’d caution against “recommended a few times on HN” (on its own!) as an indicator of much anything. --- EDIT: changed wording from “multiple times” to “a few times” to be more clear.
- nine_k 1y ago
  It may be an indicator of popularity / mindshare of certain ideas. It says little about the ideas' substance, but may still be very important in practice, if you deal with people.
  - michaelsbradley 1y ago
    The git cli tool is mentioned countless times over many years on HN, surely an indicator that it’s in popular use by the commentatorship here. I’ve seen many (seemingly informed) mentions of the jj cli tool in the last year — not unreasonable to conclude interest in it is growing among folks here, and enough to pique my curiosity. I responded to: This is the second time [I’ve seen that] someone has recommended this book on HN
gsf_emergency_2 1y ago
From Ken Stanley's wiki >He left both Uber AI Labs and the University of Central Florida in 2020 to lead the Open-Endedness team at OpenAI as a Research Science Manager.

ChrisMarshallNY 1y ago

I've always found that an "evolutionary" approach to design is important[0]. [0] https://littlegreenviper.com/evolutionary-design-specification/ https://littlegreenviper.com/evolutionary-design-specificati...

ants_everywhere 1y ago

> That's also the beauty of evolution–there's no goal in that search algorithm, Yes there is, there's a fitness function, both in biological evolution and evolutionary algorithms

IAmBroom 1y ago
That is not a goal. It is an adaptive strategy to meet changing constraints.
- layer8 1y ago
  I’d argue it isn’t a strategy either. You can’t not be subject to it.
layer8 1y ago
There’s no goal in biological evolution. Being subject to natural selection based on the present environment is an inevitability. You could argue that mutation isn’t inevitable, but its presence and degree is subject to natural selection as well.

_lex 1y ago

Incorrect. The search algo has a goal: to find out. You just think it has no goal because you didnt think in quantum.