-
Notifications
You must be signed in to change notification settings - Fork 0
/
ai.html
189 lines (189 loc) · 12.1 KB
/
ai.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
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<meta http-equiv="X-UA-Compatible" content="IE=edge" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>AI Sentience - Portfolio</title>
<link rel="stylesheet" href="css/ai.css" />
</head>
<body class="c10 doc-content" style="margin: auto">
<a class="fixed btn btn-primary" href="index.html">Back</a>
<div>
<p class="c7 c6"><span class="c0"></span></p>
</div>
<p class="c11 title" id="h.gjdgxs"><span>AI Sentience</span></p>
<p class="c1 c6 subtitle" id="h.30j0zll"><span class="c0"></span></p>
<p class="c1 subtitle" id="h.1fob9te">
<span class="c0">BoHong (Boden) Chen</span>
</p>
<p class="c1 subtitle" id="h.3znysh7">
<span class="c0">Department of Computer Science, Allen</span>
</p>
<p class="c1 subtitle" id="h.2et92p0">
<span>Computer Science III</span>
</p>
<p class="c1 subtitle" id="h.tyjcwt">
<span>Professor Ben-Yaakov</span>
</p>
<p class="c1 subtitle" id="h.3dy6vkm"><span>2 September 2022</span></p>
<p class="c4 c6"><span class="c0"></span></p>
<p class="c4 c6"><span class="c0"></span></p>
<p class="c4 c6"><span class="c0"></span></p>
<p class="c4 c6"><span class="c0"></span></p>
<p class="c7 c6"><span class="c0"></span></p>
<hr style="page-break-before: always; display: none" />
<p class="c6 c7"><span class="c0"></span></p>
<h2 class="c1" id="h.2s8eyo1"><span class="c3">Introduction</span></h2>
<p class="c4">
<span class="c0"
>There is no doubt that Artificial Intelligence (AI) technology is growing rapidly, and many optimists
in the technology world are talking about the possibility of full-fledged AI in the coming decades.
However, true sentient AI is hard to create as computers simply repeat what a human would say, making
them sound sentient. This essay assumes that this barrier has been solved, and a breakthrough has led to
fully sentient AI. In this case, we should be worried not about AI taking over the world, but about AI
taking over our tasks, leaving most of humankind jobless.</span
>
</p>
<h2 class="c1"><span class="c3">Defenition of Sentience</span></h2>
<p class="c4">
<span class="c0"
>First, we need to define the diverse idea of sentience in computers. Unfortunately, there is no formal
definition as consciousness is an incredibly complex subject spanning across the fields of psychology to
philosophy, but if the standard definition is applied to a computer, it might have an anti-climactic
ending. With consciousness and no goals in life, a computer would take the logical step and understand
that the best move is not to play and subsequently end. Therefore, we will proceed with the notion that
the common saying of AI Sentience is better defined not as AI gaining consciousness, but more as an
Artificial General Intelligence (AGI) that can replicate or exceed any intellectual task a human can and
is given a goal to try and complete (Lutkevich).</span
>
</p>
<h2 class="c1"><span class="c3">Destroyer of Worlds</span></h2>
<p class="c4">
<span class="c0"
>The concept that these advanced minds can then develop a need to destroy the world is overused and
exaggerated, yet this is still grounded in some reality. Hypothetically, a highly advanced AI is given a
task that can have good intentions and maximizing the task would be its goal. The AI would then
calculate the most logical and best way to complete the given task and perform it. The problem comes
with the ruthless efficiency that the AI would start to desire as it is optimized for such. Devoid of
emotions, if something, or more worryingly someone, is in its way, the AI would prioritize the task and
eliminate the obstacle to finishing it. Although we can always implement a stop button, there is a
likely possibility that the robot would ignore the stop to prioritize the task (Mohan). This can be seen
as the cause of the end of the world, but solutions have come forward such as Cooperative Inverse
Reinforcement Learning which prioritizes not completing the task, but how a human would complete the
task (Hadfield-Menell). In the end, this situation is far from reality, and a much more likely negative
influence of AI is on society and especially the economy (Roose).</span
>
</p>
<h2 class="c1"><span class="c3">Societal Implications</span></h2>
<p class="c4">
<span class="c0"
>With the progress that AI is making, there is becoming a growing threat of AGI bringing groundbreaking
changes. If computers develop to the point of AGI, then the possibility of fake bots, massive editing of
videos and photos, and faking of evidence becomes a reality. Employers can no longer look at a potential
candidate with trust that they are real, and people chatting would constantly question if the other side
is just a bot (Stewart). The most problematic, and the one that people are constantly raising awareness
of, is the possibility that AGI can replace humans at jobs.</span
>
</p>
<h2 class="c1"><span class="c3">Replacing Jobs</span></h2>
<p class="c4">
<span class="c0"
>If an AGI becomes a possibility, then hypothetically it can perfectly replicate a human, and therefore
their job. Recently since this essay was made, an AI-generated artwork won first place in an art contest
(Roose). If you train an AGI on the best person for the job, then the AGI would perform as if it was the
best person for the job at a fraction of the time and cost a regular worker would have. This can then
work every day for only the cost of electricity and be easily replicated and upgraded when needed. If
this is a choice given to employers, it seems obvious that they would choose a robot instead of a human.
With this in mind, employers would see the cost-effectiveness of robots as much higher and humans will
slowly be phased out. While automation can replace jobs with repetitive physical tasks, a true AGI can
hypothetically replace all of the jobs.</span
>
</p>
<p class="c4">
<span class="c0"
>Robots replacing jobs has been a common concern throughout the ages. From textile workers going against
the machinery to bank tellers worried about ATMs, economists, and experts constantly debate over the
threat of innovation replacing jobs (Andrews). Fortunately for us, history has shown that despite
automation, new jobs still come in, and the constant saying that people will do jobs that don’t
even exist yet still hold. However, if true AGI is on the market, such a breakthrough is not simply a
machine replacing a specific job, but a machine that can replicate a human and replace all jobs. If
someone were to go back in history as a horse and say that innovation would only make their jobs easier,
then a few years later and they’ll be proven wrong as their entire job is replaced with cars
(Grey). People argue against this horse analogy, saying that horses are only there for specific reasons
of transportation and nothing else. Humans on the other hand are much more diverse than horses with
millions of skills that are required in jobs and more that we might not even know about. However, this
argument fails if true AGI is created and humans are replicated perfectly, in that case, all our purpose
would be like horses and our existence becomes obsolete.</span
>
</p>
<h2 class="c1"><span class="c3">Conclusion</span></h2>
<p class="c4">
<span class="c0"
>Whether true AGI is coming is hard to say, but it is certainly possible. Whether you like it or not, we
are simply built upon a combination of 86 billion neurons. One day, some computers would be able to
replicate this and therefore become a perfect human. Whether that day is close to us is hard to say, but
if it does, computers would become the new workers of society.</span
>
</p>
<hr style="page-break-before: always; display: none" />
<p class="c4 c6"><span class="c0"></span></p>
<h1 class="c1" id="h.z337ya"><span class="c3">References</span></h1>
<p class="c2">
<span>Andrews, E. (2015, August 7). </span><span class="c5">Who were the Luddites? </span
><span class="c0"
>. History. Retrieved September 24, 2022, from https://www.history.com/news/who-were-the-luddites
</span>
</p>
<p class="c2">
<span>Grey, CGP (2014, August 13).</span>
<span class="c5">Humans Need Not Apply</span><span>. </span><span class="c0">YouTube</span
><span class="c0">. Retrieved September 2, 2022, from https://www.youtube.com/watch?v=7Pq-S557XQU.</span>
</p>
<p class="c2">
<span> Hadfield-Menell, D., Dragan, A., Abbeel, P., & Russell, S. (2016, November 12). </span
><span class="c5">Cooperative Inverse Reinforcement Learning.</span
><span class="c0">. arXiv.org. Retrieved September 2, 2022, from https://arxiv.org/abs/1606.03137/</span>
</p>
<p class="c2">
<span>Lutkevich, B. (2022, April 13). </span><span class="c5">What is Artificial General Intelligence?</span
><span class="c0"
>. Tech Target. Retrieved September 2, 2022, from
https://www.techtarget.com/searchenterpriseai/definition/artificial-general-intelligence-AGI
</span>
</p>
<p class="c2">
<span>Mohan, S. (2021, October 29). </span><span class="c5">Stop button paradox in agi. </span
><span class="c0"
>. Medium. Retrieved September 24, 2022, from
https://medium.com/@shivamohan07/stop-button-paradox-in-agi-69c3d008ae93
</span>
</p>
<p class="c2">
<span>Roose, K. (2022, August 24). </span
><span class="c5">We need to talk about how good A.I. is getting</span
><span class="c0"
>. The New York Times. Retrieved September 2, 2022, from
https://www.nytimes.com/2022/08/24/technology/ai-technology-progress.html
</span>
</p>
<p class="c2">
<span>Roose, K. (2022, September 2). </span
><span class="c5">An a.i.-generated picture won an art prize. artists aren't happy.</span
><span class="c0"
>. The New York Times. Retrieved September 2, 2022, from
https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html
</span>
</p>
<p class="c2">
<span> Stewart, J. (2019, March 1). </span
><span class="c5">Artificial Intelligence: The problem with perfection</span
><span class="c0"
>. Compare the Cloud. Retrieved September 2, 2022, from
https://www.comparethecloud.net/articles/artificial-intelligence-perfection-problem/</span
>
</p>
<p class="c7 c6 c8"><span class="c0"></span></p>
<p class="c7 c8 c6"><span class="c0"></span></p>
</body>
</html>