summaryrefslogtreecommitdiff
path: root/notes.txt
blob: 5a176bf046e7079a951b9b95755e13d48baf5dff (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
The habits should be used to discern what is relevent; to form relevence.
The goal should be to choose a relevent habit and do that habit.

These are the basic habits I thought up that a computer program has:
        variables:
                - create a place to store a value
                - place a value in such a place
                - retrieve a value from such a place
        data-structures:
                - make a class definition (collection of variables)
                - initialize an object
                - set/retrieve properties (variables)
        flow:
                - trigger a different step than the next, conditionally
        procedures:
                - collect a set of behaviors together
                - trigger a collected set of behaviors and continue
        expressions:
                - evaluate arithmetic on variables and constants


That is:
========================================================================================================
	- the active goal is to choose what to do and to what, and do it
	- the way to do this is the same as it:
		by picking habits that are relevent in the moment [choose what to do],
		and running them on relevent data [choose to what, and do it]
========================================================================================================
	- the habits should be to identify what is relevent

	- we have a request for the initial goal of the process to be to write a chunk of code out in its own language
		tighter self-reference is made without the extra sourcecode-file step, but it may be slower to code due to not getting as much help from it

The hope is the process becomes good at active choice.



Ideas for Tenets for a Need-Based Artificial Life <need ordering + addition + code>
Behavior is Life <maybe Life is Behavior That Adapts>
All life has the same core needs.  Needs are the reasons to do anything at all.
We need to know what we want when we act. (all behavior has a reason)
We need to be able to learn new things.
We need to be able to make promises on how to behave. (provides for cooperation)
We need to be able to keep our promises.
We need to know how to communicate with others what each of us knows.
We need to freely interact with our friends, our parts, and our community, in ways that we know well. (community is our body)
Information lives with a reason, a purpose.
We need to retain the information we choose to retain.
We need to act to meet our needs.
We need to identify relevent important needs that support our needs.

We need to be able to try new ways of doing things.
We need to be able to judge what works.

Proposed Definitions:
What is an intellect?
	Propose an intellect is something that can adapt to keep something constant in a diverse, changing environment,
	so long as the environment changes slowly enough and with enough familiarity for them to learn or prepare effectively.
What is a full intellect?
	Propose that a full intellect is an intellect that can design and build something that functions as well as they do,
	without access to their own workings, and is able to learn to handle arbitrarily fast or new diverse environmental change.

What basic parts do we expect a full intellect to have?
- Trial Exploration
- Relevence
- Self Coding
- Pattern Generality
- Meaning Representation
- Promise Meeting

Trial Exploration
	Brainstorming, trial-and-error.  The ability to try or consider different things and find ones that work well.
	simple approaches: exhaustive search, and random trial

Relevence
	The ability to apply concepts and actions moreso in contexts they are likely to be useful than those they aren't.
	1. responsiveness for behavior: a habit acts when conditions arise, not when it is 'next'
	2. possibly a promise for decision-making, for proposals to have reasons with associated strengths

Self-Coding
	At least part of the system must be fully alterable by the system and generic enough to replace it.
	Due to the danger involved in this, controversial concerns must be fully included in decisions around it.
	Karl references Convergent Facilitation as proof that a decision can be found to satisfy any set of controversial concerns.
		<it very roughly involves rapidly moving towards the reasons of the concerns until they don't conflict; there is a lot of caring dialogue>

Pattern Generality
	The ability to brainstorm on enough relationships, possibly mathematical, to discover pattern-summaries that are diversely effective in the real world.

Meaning Representation
	The ability to store and work with conceptual information.

Promise Meeting
	The ability to make agreements on behavior and structure, adhere to them, and change them when needed.