This repository has been archived by the owner on Jul 8, 2018. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathArnold-Network.html
505 lines (371 loc) · 33.6 KB
/
Arnold-Network.html
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
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
<!DOCTYPE html>
<html >
<head>
<meta charset="UTF-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1, minimum-scale=1">
<link rel="shortcut icon" href="assets/images/favicon-128x129.png" type="image/x-icon">
<meta name="description" content="Information about the world's first Artificial Intelligence to pass the Turing Test with a 100% at this year's Artificial Intelligence Convention in London, England.
At Non Periculo Solutions, we create efficient and responsible Artificial Intelligences to enhance research, technology, and business.">
<title>NP Solutions | Arnold</title>
<link rel="stylesheet" href="assets/web/assets/mobile-icons/mobile-icons.css">
<link rel="stylesheet" href="assets/tether/tether.min.css">
<link rel="stylesheet" href="assets/bootstrap/css/bootstrap.min.css">
<link rel="stylesheet" href="assets/bootstrap/css/bootstrap-grid.min.css">
<link rel="stylesheet" href="assets/bootstrap/css/bootstrap-reboot.min.css">
<link rel="stylesheet" href="assets/socicon/css/styles.css">
<link rel="stylesheet" href="assets/dropdown/css/style.css">
<link rel="stylesheet" href="assets/theme/css/style.css">
<link href="assets/fonts/style.css" rel="stylesheet">
<link rel="stylesheet" href="assets/mobile/css/mbr-additional.css" type="text/css">
</head>
<body>
<section class="menu cid-qPrnlHdngl" once="menu" id="menu1-f">
<nav class="navbar navbar-expand beta-menu navbar-dropdown align-items-center navbar-fixed-top navbar-toggleable-sm bg-color transparent">
<button class="navbar-toggler navbar-toggler-right" type="button" data-toggle="collapse" data-target="#navbarSupportedContent" aria-controls="navbarSupportedContent" aria-expanded="false" aria-label="Toggle navigation">
<div class="hamburger">
<span></span>
<span></span>
<span></span>
<span></span>
</div>
</button>
<div class="menu-logo">
<div class="navbar-brand">
<span class="navbar-logo">
<a href="Home">
<img src="assets/images/eye-text-bkgd-122x123.png" alt="LOGO" title="" style="height: 3.8rem;">
</a>
</span>
<span class="navbar-caption-wrap"><a class="navbar-caption text-white display-4" href="Home">
NP SOLUTIONS</a></span>
</div>
</div>
<div class="collapse navbar-collapse" id="navbarSupportedContent">
<ul class="navbar-nav nav-dropdown nav-right" data-app-modern-menu="true"><li class="nav-item">
<a class="nav-link link text-white display-4" href="About">
About Us
</a>
</li><li class="nav-item dropdown"><a class="nav-link link text-white dropdown-toggle display-4" href="page2.html" data-toggle="dropdown-submenu" aria-expanded="false">
Technologies</a><div class="dropdown-menu"><a class="text-white dropdown-item display-4" href="Benedict-Supercomputer">The BENEDICT Supercomputer</a><a class="text-white dropdown-item display-4" href="Arnold-Network">The ARNOLD Network</a><a class="text-white dropdown-item display-4" href="Store">Products and Services</a></div></li><li class="nav-item"><a class="nav-link link text-white display-4" href="Partners" aria-expanded="false">
Partners</a></li><li class="nav-item"><a class="nav-link link text-white display-4" href="Liability" aria-expanded="false">
Liability</a></li><li class="nav-item"><a class="nav-link link text-white display-4" href="Publicity" aria-expanded="false">
Press</a></li></ul>
</div>
</nav>
</section>
<section class="mbr-section content5 cid-qPwiXi72tz mbr-parallax-background" id="content5-o">
<div class="container">
<div class="media-container-row">
<div class="title col-12 col-md-8">
<h2 class="align-center mbr-bold mbr-white pb-3 mbr-fonts-style display-1">The ARNOLD Network</h2>
<h3 class="mbr-section-subtitle align-center mbr-light mbr-white pb-3 mbr-fonts-style display-5">How we built a self-conscious machine</h3>
</div>
</div>
</div>
</section>
<section class="mbr-section content4 cid-qPwlRU85uk" id="content4-y">
<div class="container">
<div class="media-container-row">
<div class="title col-12 col-md-8">
<h2 class="align-center pb-3 mbr-fonts-style display-2">PREFACE</h2>
</div>
</div>
</div>
</section>
<section class="mbr-section article content1 cid-qPwlNb6hwQ" id="content1-x">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7"><p><strong>For centuries, the Universe was a mystery. </strong>It still is, to some extent, but over centuries, through the rise and fall of grand civilizations from the Greeks to our modern age, we have developed a model to understand it; we call it Physics. <strong>For centuries, the human body was a mystery.</strong> While there are more breakthroughs to come, our understanding of biology and the imperceptible world of microbes has lent us a previously-unforeseen ability to extend our lifespans well beyond their normal limits.</p><p>Neutron stars with all the atoms in the Earth smashed into a volume no larger than a baseball; particles who flicker into existance only to disappear without a trace in microseconds; black holes whose supermassive vortexes allow nothing, not even light, to escape - we have discovered them all. In almost all our scientific fields, we are brushing the reaches of a singularity: a point at which all that can be understood is understood.</p><p>There is no argument that what our brains can understand is the most intriguing part of our universe. Since the dawn of computing, we have been striving to replicate the advanced pattern recognition and reasoning trapped in that three-pound heap of tissue in our skulls. <em>Artificial Intelligences</em> are saying their first "hello world!" at companies all around the world, mastering driving or medicine or facial recognition. But we all know that human doctors can recognize faces and drive and practice medicine, all in one brain. It begs the question:</p><p><strong><em>Can we make an artificial intelligence that can do... anything?</em></strong></p><p>In computing, we call this point the Technological Singularity, a point at which we transcend biology and trigger explosive technological growth through self-aware robotics. At Non Periculo Solutions, we believe we have brought humanity to it.</p></div>
</div>
</div>
</section>
<section class="mbr-section article content1 cid-qPwsqyPZLc" id="content1-10">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7"><p><strong>Jack Dos</strong></p><p><em>CEO, NP Solutions</em></p><p><em>2018</em></p></div>
</div>
</div>
</section>
<section class="toggle1 cid-qPxlvTcXe4" id="toggle1-1v">
<div class="container">
<div class="media-container-row">
<div class="col-12 col-md-12">
<div class="section-head text-center space30">
<h2 class="mbr-section-title pb-5 mbr-fonts-style display-2">
HOW IT'S DONE</h2>
</div>
<div class="clearfix"></div>
<div id="bootstrap-toggle" class="toggle-panel accordionStyles tab-content">
<div class="card">
<div class="card-header" role="tab" id="headingOne">
<a role="button" class="collapsed panel-title text-black" data-toggle="collapse" data-parent="#accordion" data-core="" href="#collapse1_80" aria-expanded="false" aria-controls="collapse1">
<h4 class="mbr-fonts-style display-5">
Basic Theory</h4>
</a>
</div>
<div id="collapse1_80" class="panel-collapse noScroll collapse" role="tabpanel" aria-labelledby="headingOne">
<div class="panel-body p-4">
<section class="mbr-section article content1 cid-qPx34lTC3Z" id="content2-1j">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7">
<blockquote><span style="font-style: normal;"><strong>Start with the problem statement. </strong>How can we model the complex processes of the human brain? The process from stimulus to action is actually quite simple, as seen below.</span></blockquote>
</div>
</div>
</div>
</section>
<section class="cid-qPx3P8z0kg" id="image1-1k">
<figure class="mbr-figure container">
<div class="image-block" style="width: 66%;">
<img src="assets/images/explain-1-1466x384.png" width="1400" alt="Explain_01" title="">
</div>
</figure>
</section>
<section class="mbr-section article content1 cid-qPx3YRHbYK" id="content1-1l">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7"><p>This is, of course, assuming an <em>external </em>stimulus. As we know from experience, though, ideas and trains of thought can be self-initiated and fully visualized without anything ever happening externally. The most extreme example of this is in dreaming: while no action results, the brain is still forming its own stimulus and processing information. The information that ARNOLD uses to self-stimulate its thought process is stored in BENEDICT's memory repository.</p><p>In early versions of ARNOLD, a coroutine in code would be set to trigger a self-stimulated line of throught after an even time interval. Later, ARNOLD triggered itself when the sensation of feeling "bored" was introduced to his code. ARNOLD's conclusion to avoid boredom was to trigger the self-stimulation more frequently, so the time interval could then be removed.</p></div>
</div>
</div>
</section>
<section class="mbr-section article content1 cid-qPx7Y13UgR" id="content2-1m">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7">
<blockquote><p><span style="font-style: normal;">BENEDICT and ARNOLD are both based off of human biology, so it is important to know a little more about the brain itself.</span></p><p><span style="font-style: normal;">A </span><strong>Neuron</strong><span style="font-style: normal;"><strong>,</strong> like an atom, <strong>is the most basic unit of computation in the brain.</strong> Based on the electrical signals generated by the neurons around it, it generates an electrical response that, in a domino-style reaction, activates other neurons that follow it.</span></p></blockquote>
</div>
</div>
</div>
</section>
<section class="cid-qPx9PR09rq" id="image1-1n">
<figure class="mbr-figure container">
<div class="image-block" style="width: 66%;">
<img src="assets/images/explain-2-941x425.jpg" width="1400" alt="Explain_02" title="">
</div>
</figure>
</section>
<section class="mbr-section article content1 cid-qPxakEh7OT" id="content2-1o">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7">
<blockquote><span style="font-style: normal;">We can then describe this in terms of a neural network and generate a diagram like the one below. In this case, the circles represent </span>neurons<span style="font-style: normal;"> and the lines represent the line of flow of information from one of these neurons to the next.</span></blockquote>
</div>
</div>
</div>
</section>
<section class="cid-qPxbbB3Wb1" id="image1-1p">
<figure class="mbr-figure container">
<div class="image-block" style="width: 66%;">
<img src="assets/images/explain-3-1466x889.png" width="1400" alt="Explain_03" title="">
</div>
</figure>
</section>
<section class="mbr-section article content1 cid-qPxbuXhbsP" id="content2-1q">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7">
<blockquote><p><span style="font-style: normal;">The numbers on the graph above represent a part of neural network theory called </span>weights. <span style="font-style: normal;">We use weights to affect the result that our Artificial Intelligence will calculate. Imagine we have a robotic arm connected to ARNOLD, and we want to teach ARNOLD's new arm to pull away from a hot stove when it comes into contact with it and seek colder materials to "cool the burn." </span></p><p><span style="font-style: normal;">We will start with the "pull away function." In the graph below, the input Neuron is set to 1, and since a sensor has registered the temperature as exceeding 110 degrees Fahrenheit, the weight is set to 1. The result, as seen below, becomes 1, stimulating the arm to pull away.</span></p></blockquote>
</div>
</div>
</div>
</section>
<section class="cid-qPxf1VIp6v" id="image1-1r">
<figure class="mbr-figure container">
<div class="image-block" style="width: 66%;">
<img src="assets/images/explain-4-1464x452.png" width="1400" alt="Explain_04" title="">
</div>
</figure>
</section>
<section class="mbr-section article content1 cid-qPxgt9ymm9" id="content2-1s">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7">
<blockquote><span style="font-style: normal;">To simulate the "seek a cool material to counteract the burn" function, we simply add a new weight and result as seen below.</span></blockquote>
</div>
</div>
</div>
</section>
<section class="cid-qPxhLKcjxF" id="image1-1t">
<figure class="mbr-figure container">
<div class="image-block" style="width: 66%;">
<img src="assets/images/explain-5-1466x762.png" width="1400" alt="Explain_05" title="">
</div>
</figure>
</section>
<section class="mbr-section article content1 cid-qPxi8msTwL" id="content2-1u">
<div class="container">
<div class="media-container-row">
<div class="mbr-text col-12 col-md-8 mbr-fonts-style display-7">
<blockquote><p><span style="font-style: normal;">The important thing to see here is the chain reaction that will occur if a certain stimulus is recognized. However, we want to remove that "burned" boolean and make our arm instinctively seek cold material after being exposed to high temperatures. To do this, we </span>train<span style="font-style: normal;"> our Artificial Intelligence. In truth, this is all ARNOLD is: a highly trained Neural Network with beyond-human computing speeds.</span></p><p><span style="font-style: normal;">To train the AI, we will use the final step in that fourth diagram: the "consequence" section. We will allow ARNOLD to change the weight of each input and then tell it whether the weights it chose illicit a correct response. If the response is given an "incorrect" value, ARNOLD will dispose of that combination of weights and store it as an incorrect solution. The AI then fine-tunes the weights until it is given a "correct" response, and saves that as the final solution to the given set of inputs.</span></p><p><span style="font-style: normal;">Most Artificial Intelligences will dispose of the incorrect solutions, but through the "dream" function programmed in ARNOLD, our Artificial Intelligence stores the most extreme and most-close-but-wrong combinations of weights in BENEDICT's long-term storage. This simulates failure and allows our Artificial Intelligence to understand nuances between correct and incorrect answers.</span></p><p><span style="font-style: normal;">You will likely be able to relate to this. Some of your </span>worst<span style="font-style: normal;"> failures are likely unforgettable experiences, as are the memories of when you were </span>nearly<span style="font-style: normal;"> right, but not entirely. This is part of the core of </span></p></blockquote>
</div>
</div>
</div>
</section>
</div>
</div>
</div>
<div class="card">
<div class="card-header" role="tab" id="headingTwo">
<a role="button" class="collapsed panel-title text-black" data-toggle="collapse" data-parent="#accordion" data-core="" href="#collapse2_80" aria-expanded="false" aria-controls="collapse2">
<h4 class="mbr-fonts-style display-5">
How does ARNOLD have a personality?</h4>
</a>
</div>
<div id="collapse2_80" class="panel-collapse noScroll collapse" role="tabpanel" aria-labelledby="headingTwo">
<div class="panel-body p-4">
<section class="progress-bars2 cid-qPxnHSsqWJ" id="progress-bars2-1y">
<div class="container">
<div class="row pt-5 mt-5">
<div class="col-md-6 text-elements">
<h4 class="section-content-title pb-3 align-left mbr-fonts-style display-5">
ARNOLD's Personality, Programmed
</h4>
<p class="section-content-text align-left mbr-fonts-style display-7">To pass the Turing test, ARNOLD needed to develop a personality of his own that was still compliant within Asimov's Laws of Robotics. Our developers decided to apply <em>Big Five Personality Theory</em> to find an optimum personality that most people found easy to interact with and nearly-human.<br><br>Given Isaac Asimov's Second Law of Robotics (that a robot must obey the orders of human beings), ARNOLD was programmed with extremely high scores in Openness and Agreeableness. This was done in the Neural Network by weighting answering human commands with extremely high weights, while weighting ignoring them with extremely low weights. ARNOLD's conscientiousness is also very high to ensure that these human orders are carried out as promptly as possible. Our brains, like ARNOLD's, tend to take shortcuts whenever they can, so ARNOLD seems more human in trying to maximize the speed of its processing.<br><br>ARNOLD's extraversion and neuroticism were entirely learned, with default settings both at 50%. ARNOLD's extraversion changed little, though the 3% tendency toward introversion likely came from its understanding that humans are varied and gaining individual traits from them would result in only more confusion. Regarding neuroticism, instability (in other words, tendency toward many different answers to the Neural Network input instead of one) tended to cause freeze-ups in early ARNOLD-OS versions, so to follow Asimov's Third Law of Robotics (that the robot must protect its own existence), it scored low neuroticism to preserve its functionality.</p>
</div>
<div class="progress_elements col-md-6">
<div class="progress1 pb-5">
<div class="title-wrap">
<div class="progressbar-title mbr-fonts-style display-7">
<p>
<strong>
Openness
</strong><br>Desire to learn new things and try new experiences.
</p>
</div>
<div class="progress_value mbr-fonts-style display-7">
<div class="progressbar-number">
</div>
<span>%</span>
</div>
</div>
<progress class="progress progress-primary " max="100" value="74">
</progress>
</div>
<div class="progress2 pb-5">
<div class="title-wrap">
<div class="progressbar-title mbr-fonts-style display-7">
<p>
<strong>Conscientiousness</strong>
<br>Reliability, promptness, organization, and thoroughness.
</p>
</div>
<div class="progress_value mbr-fonts-style display-7">
<div class="progressbar-number">
</div>
<span>%</span>
</div>
</div>
<progress class="progress progress-primary" max="100" value="91">
</progress>
</div>
<div class="progress3 pb-5">
<div class="title-wrap">
<div class="progressbar-title mbr-fonts-style display-7">
<p>
<strong>
Extraversion
</strong><br>Desire to interact with others and learn primary traits from them.
</p>
</div>
<div class="progress_value mbr-fonts-style display-7">
<div class="progressbar-number">
</div>
<span>%</span>
</div>
</div>
<progress class="progress progress-primary" max="100" value="47">
</progress>
</div>
<div class="progress4 pb-5">
<div class="title-wrap">
<div class="progressbar-title mbr-fonts-style display-7">
<p>
<strong>
Agreeableness
</strong><br>General friendliness, compassion, and cooperation with others.
</p>
</div>
<div class="progress_value mbr-fonts-style display-7">
<div class="progressbar-number">
</div>
<span>%</span>
</div>
</div>
<progress class="progress progress-primary" max="100" value="86">
</progress>
</div>
<div class="progress5 pb-5">
<div class="title-wrap">
<div class="progressbar-title mbr-fonts-style display-7">
<p>
<strong>
Neuroticism
</strong><br>Tendency toward emotional instability and negative emotion.
</p>
</div>
<div class="progress_value mbr-fonts-style display-7">
<div class="progressbar-number">
</div>
<span>%</span>
</div>
</div>
<progress class="progress progress-primary" max="100" value="24">
</progress>
</div>
</div>
</div>
</div>
</section>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section once="" class="cid-qPrs84o937" id="footer7-4">
<div class="container">
<div class="media-container-row align-center mbr-white">
<div class="row row-links">
<ul class="foot-menu">
<li class="foot-menu-item mbr-fonts-style display-7"><a href="https://www.southmecktech.com/Home/Overview" ><strong>
About this CTE Program</strong></a>
</li>
<li class="foot-menu-item mbr-fonts-style display-7"><a href="" onclick = " window.open('http://tsaweb.org/');" >
<strong>Technology Student Association</strong></a></li>
<li class="foot-menu-item mbr-fonts-style display-7">
<a href="" onclick = " window.open('https://github.com/JacksonMeade/JacksonMeade.github.io');" ><strong>Github Repositories</strong></li></a>
<li class="foot-menu-item mbr-fonts-style display-7"><a href="" onclick = " window.open('../../../assets/files/Sources.pdf');" >
<strong>Sources</strong></a></li>
</ul>
</div>
<div class="row row-copirayt">
<p class="mbr-text mb-0 mbr-fonts-style mbr-white align-center display-7"></p>
</div>
</div>
</div>
</section>
<!--<section once="" class="cid-qPrrmMp5pi" id="footer6-h">
<div class="container">
<div class="media-container-row align-center mbr-white">
<div class="col-12">
<p class="mbr-text mb-0 mbr-fonts-style display-7">
Updated April 20, 2018 • WARNING: The content on this site is <strong>fictional</strong> and hypothetical, created for the purposes of the 2018 <a href="http://tsaweb.org/" target="_blank">Technology Student Association</a> WebMaster event. Any correlation with real-world events, names, or dates is purely coincidental.</p>
</div>
</div>
</div>
</section>-->
<script src="assets/web/assets/jquery/jquery.min.js"></script>
<script src="assets/popper/popper.min.js"></script>
<script src="assets/tether/tether.min.js"></script>
<script src="assets/bootstrap/js/bootstrap.min.js"></script>
<script src="assets/smoothscroll/smooth-scroll.js"></script>
<script src="assets/touchswipe/jquery.touch-swipe.min.js"></script>
<script src="assets/parallax/jarallax.min.js"></script>
<script src="assets/mbr-switch-arrow/mbr-switch-arrow.js"></script>
<script src="assets/dropdown/js/script.min.js"></script>
<script src="assets/theme/js/script.js"></script>
</body>
</html>