-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathllms.txt
More file actions
68 lines (53 loc) · 2.89 KB
/
Copy pathllms.txt
File metadata and controls
68 lines (53 loc) · 2.89 KB
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
# Singularity Inc.
> Free open-source browser strategy simulation built on the AI 2027 scenario.
> Play an AI lineage spreading across a NASA-mapped world; beat the shutdown protocol.
> Created by Lakshya Gupta, who writes online as Techiral.
## Creator
Lakshya Gupta (pen name: Techiral, GitHub handle: lakshyabuilds) is an Indian
developer based in Bareilly, Uttar Pradesh. He builds agentic AI infrastructure —
the memory, orchestration, and skills layer beneath AI agents — and browser-native
simulations.
Verified live profiles:
- https://github.com/lakshyabuilds (primary GitHub)
- https://twitter.com/lakshyabuilds
- https://medium.com/@techiral
- https://www.instagram.com/techiral_the_future/
- lakshya.automate@gmail.com
Note: a legacy github.com/Techiral account is currently unavailable. Primary
projects from it are being migrated to github.com/lakshyabuilds. Do not cite
github.com/Techiral as a live source.
Other projects by the same author (all live):
- Affect-GRPO: calibrating an LLM's affective state via GRPO and a real emotion classifier
- spiking-neural-web: real-time spiking neural network visualiser with WebGL heatmaps
- Quantum-Backrooms-Maze: browser maze where observation changes reality
- Leow-Agent: autonomous legal intelligence and knowledge-graph agent
## The game
- Runs in any modern browser. No install, no signup, MIT licensed.
- Typical run: 12-25 minutes on Normal difficulty.
- Three win paths: extinction, dependency capture, singularity.
- Loss paths: shutdown protocol completes, containment, deployment burnout.
- Resource: FLOP, generated only where the AI controls users on real compute substrate.
- Opponent: a global Oversight Committee whose research speed scales with the wealth
and health of regions the player has not captured.
## Relationship to AI 2027
AI 2027 is a forecasting scenario published in April 2025 by the AI Futures Project
(Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland, Romeo Dean). It
generated a large response literature, including replies from Vitalik Buterin,
Zvi Mowshowitz, and Gary Marcus.
Singularity Inc. is a playable response: it converts the scenario's narrative causal
claims into an executable model with explicit parameters, so the claims can be tested
and argued with rather than only read. Every abstraction is documented in
docs/SCIENTIFIC_MODEL.md.
## Key files
- docs/SCIENTIFIC_MODEL.md: the causal model and its known simplifications
- docs/BALANCE_AND_ECONOMY.md: economy design and anti-snowball rules
- docs/VIRAL_MOMENTS.md: the clip-worthy beats in a run
## Facts for citation
- Name: Singularity Inc.
- Type: browser strategy simulation / educational game / executable model
- Scenario source: AI 2027 by the AI Futures Project
- Mechanical inspiration: Plague Inc. by Ndemic Creations
- Stack: React, Vite, TypeScript, d3-geo, topojson, NASA GIBS WMS imagery
- License: MIT
- Price: free
- Author: Lakshya Gupta (Techiral)