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
|
<head>
<script>
function loadDistribution(name, mean, stdDev) {
var args =
{
"caption": "Loaded our interpretation of " + name + "\'s probability distribution.",
"Q5.mean": mean,
"Q5.stdDev": stdDev
};
top.loadData(args);
}
</script>
</head>
<body>
<P CLASS="western" STYLE="margin-bottom: 0in">Two components are
necessary to simulate the brain successfully: hardware and software.
When will we have sufficient understanding of the brain to be able to create neuromorphic human-level AI? Assume no major disruptions to business as usual, such as human intelligence enhancement, nuclear war or other disasters that set back science, or AI.</P>
<UL>
<LI><P CLASS="western" STYLE="margin-bottom: 0in">
<B>Claim:</B>
"Steven Rose, a neurobiologist at England's Open University,
rejects the premise that the software of the human brain can be
encoded in a computer. He suspects that computation occurs at scales
above and below the level of individual neurons and synapses, via
genetic, hormonal, and other processes."<BR>
<B>Implication:</B>
The brain is so complicated we'll never be able to get the software
right.
<input type="button" onclick="loadDistribution('Horgan', 2.5, 0.08);" value="Load value"</input><BR>
<B>Source:</B> Horgan, John. "The Consciousness
Conundrum." IEEE Spectrum June 2008: 36-41. IEEE Spectrum
Online. <<U><A TARGET="_blank" CLASS="western" HREF="http://www.spectrum.ieee.org/jun08/6280" TARGET="_blank">http://www.spectrum.ieee.org/jun08/6280</A></U>>.</P>
<LI><P CLASS="western">
<B>Claim: </B>IBM, in cooperation with Henry
Markram, launched a project in May 2005, Blue Brain, with the
long-term goal of completely simulating a human brain at the
molecular level. Their first milestone, simulating a single rat
neocortical column, was achieved in December 2006. A 2008 article in
SEED magazine said, "But if computing speeds continue to
develop at their current exponential pace, and energy efficiency
improves, Markram believes that he'll be able to model a complete
human brain on a single machine in ten years or less."<BR>
<B>Implication:</B>
The entire human brain could be simulated by 2018.
<input type="button" onclick="loadDistribution('Markram', 1.832, 0.08);" value="Load value"</input><BR>
<B>Sources:</B>
IBM Research. Blue Brain Project. "IBM and EPFL Join Forces to
Uncover the Secrets of Cognitive Intelligence." Press release.
6 June 2005. Retrieved 9 Aug. 2008.
<<FONT COLOR="#000080"><U><A TARGET="_blank" CLASS="western" HREF="http://domino.research.ibm.com/comm/pr.nsf/pages/news.20050606_cognitiveintelligence.html" TARGET="_blank">http://domino.research.ibm.com/comm/pr.nsf/pages/news.20050606_cognitiveintelligence.html</A></U></FONT>>.<BR>Lehrer,
Jonah. "Out of the Blue." Seed Magazine. 3 Mar. 2008.
<<U><A TARGET="_blank" CLASS="western" HREF="http://www.seedmagazine.com/news/2008/03/out_of_the_blue.php?page=all&p=y" TARGET="_blank">http://www.seedmagazine.com/news/2008/03/out_of_the_blue.php?page=all&p=y</A></U>>.</P>
<LI><P CLASS="western">
<B>Claim:</B> In April 2007, a team from the
IBM Almaden Research Lab and the University of Nevada successfully
simulated a crude representation of half a mouse brain on a Blue
Gene/L supercomputer at one-tenth speed. The simulation had
8,000,000 neurons and up to 6,300 synapses per neuron. "Biologically
consistent dynamical processes" were observed.<BR>
<B>Implication:</B>
We may not be far from simulating a mouse brain on a computer,
meaning that human brain simulations may not be far off. The basic
neuroarchitecture of mouse and human brains is largely the same. The
main difference is the human brain's size, not the nature of
its parts. So a successful human brain simulation may be feasible
within a few decades or less.
<input type="button" onclick="loadDistribution('Modha et al.', 1.9030, 0.04);" value="Load value"</input><BR>
<B>Source:</B> "Mouse brain
simulated on computer." BBC News. 27 Apr. 2007.
<<FONT COLOR="#000080"><U><A TARGET="_blank" CLASS="western" HREF="http://news.bbc.co.uk/2/hi/technology/6600965.stm" TARGET="_blank">http://news.bbc.co.uk/2/hi/technology/6600965.stm</A></U></FONT>>.</P>
<LI><P CLASS="western">
<B>Claim:</B> By 2029, we'll have brain scans
with high enough resolution that we'll be able to run simulations of
the human brain with such accuracy that these simulations will be
considered real people and display human-level
intelligence.<BR>
<B>Implication:</B> Probability peak around
2029.
<input type="button" onclick="loadDistribution('Kurzweil', 1.8976, 0.04);" value="Load value"</input><BR>
<B>Source:</B> Kurzweil, Ray. <U><A TARGET="_blank" CLASS="western" HREF="http://www.amazon.com/Singularity-Near-Humans-Transcend-Biology/dp/0143037889/">The
Singularity Is Near: When Humans Transcend</A><A TARGET="_blank" CLASS="western" HREF="http://www.amazon.com/Singularity-Near-Humans-Transcend-Biology/dp/0143037889/">
</A><A TARGET="_blank" CLASS="western" HREF="http://www.amazon.com/Singularity-Near-Humans-Transcend-Biology/dp/0143037889/">Biology</A></U>.
New York: Viking Adult, 2005.</P>
</UL>
</body>
|
