If games/media are designed systems that can be analyzed through HPS, what about systems that are emergent?
emergence
highways example already used. Termites - structure from simple rules
connection with enchantment
an 'encounter' with order from randomness
leads to unexpected & expected transformations
Oregon Trail: 1971
Not Emergent
Modern parser based IF, often written with Inform10: can look like this:
"Cave Entrance"
The Cobble Crawl is a room. "You are crawling over cobbles in a low passage. There is a dim light at the east end of the passage."
A wicker cage is here. "There is a small wicker cage discarded nearby."
The Debris Room is west of the Crawl. "You are in a debris room filled with stuff washed in from the surface. A low wide passage with cobbles becomes plugged with mud and debris here, but an awkward canyon leads upward and west. A note on the wall says, 'Magic word XYZZY'."
The black rod is here. "A three foot black rod with a rusty star on one end lies nearby."
Above the Debris Room is the Sloping E/W Canyon. West of the Canyon is the Orange River Chamber.
ELIZA: 1966, by Joseph Weizenbaum... for him, it showed the limits of human-computer interaction: and that it was undesireable
outcomes largely ordained; there might be many of them, but they are all a function of how you traverse the 'map' of connected 'rooms' - ie, a network
Code was deterministic in the sense it searched for keywords and made substitutions BUT any particular 'play' had impossible to predict outcome
Emergence of D&D in the 1970s
systems for generating 'dungeons', characters, objects,
dungeon master or storyteller reacts dynamically to players within these worlds/constraints
Early computing hardware: serious memory & computing limits.
Rogue et al 1980
Eliza Effect
human motivations, aspirations, reasons ascribed to machines
note also the gendered aspect
https://upload.wikimedia.org/wikipedia/en/2/26/TheOregonTrail1971Gameplay.png
https://archive.org/details/OregonTrailMainframe
note small file size! played at first via teletype
perceptron 1943. Warren McCulloch, Walter Pitts invent the algorithm of the perceptron
https://www.simplilearn.com/tutorials/deep-learning-tutorial/perceptron
perceptron seemed a dead end until we figured out how to add more layers, how to feed forward and backpropogate, and how to feed it not just images...
add many more layers;
add many more inputs;
you get a modern neural network
And here's how the perceptron does the math
https://miro.medium.com/max/1002/1*ztXU57QEETPHGXczHrSWSA.gif
And you get AI writing stories, or being chatbots. AI Dungeon trained on old parser based stories; chatbots trained on phonecall transcripts
https://huggingface.co/spaces/awacke1/CB-GR-Chatbot-Blenderbot
key themes:
emergence
eliza effect
non-linear systems
No Man's Sky
formulae to cover evolution of species
fractals to generate forms
formulae to explain connections between air temeperature, pressure, terrain etc
VAST universe, all of it generated from random chaos EXCEPT the 'source seed' which comes with the game; hence every player explores the same universe (MC has a different seed for each game)
here's some python, similar in spirit (https://python.plainenglish.io/create-a-random-dungeon-with-python-f17118c1eebd):
def init_rooms():
"""Initializes the rooms in the dungeon."""
total_rooms = randrange(min_rooms,max_rooms)
for i in range(max_iters):
for r in range(total_rooms):
if len(rooms) >= max_rooms:
break x = randrange(0,map_width)
y = randrange(0,map_height) width = randrange(min_room_size,max_room_size)
height = randrange(min_room_size,max_room_size)
room = Room(x,y,width,height) if check_for_overlap(room, rooms):
pass
else:
rooms.append(room) for room in rooms:
for y in range(room.y, room.y+room.height):
for x in range(room.x, room.x+room.width):
map[x,y] = 1these two trends not connected yet, but give it time!
Watabou - ProcGen Arcana
systems for ttrpg
...just need a story teller to connect them
take incredibly lifelike simulations, generated by the machine
take incredibly lifelike HUMANS, generated by the machine
Put them into historical situations
Where is the 'history' here?
How do we deal with the dangers of the Eliza effect?
for next day: we raise the dead with https://minimaxir.com/2019/09/howto-gpt2/
to prepare, find the writing of a historical individual on Gutenberg Project
compile their writing into a single text document
DO NOT copy the copyright notices etc
aim for < 10 mb in size
We're creating the source for an experiment in raising the dead
Frank Rosenblatt builds the Perceptron Mark 1958
Machine meant for image recognition. Would learn shapes as a statistical distribution. Even then, it was realized that the framework could learn other things too.
depends on feedback loop [key part of emergence in complex systems, eh?]
https://www.glass-bead.org/article/machines-that-morph-logic/?lang=enview
You may add notes by double clicking.
Single click on a note to turn on linking; click on link icon, then click on note to link to. Click the canvas again to turn off linking.