Uploaded by Jason Pierce

S19005148 - Jason Broderick-Pierce - Assessment 3

advertisement
The Technological Singularity
1
The Technological Singularity
Jason M T Broderick-Pierce, Member, IET.
Abstract—The technical singularity is a topic of hot debate and
is an ever-increasing likelihood within the next few decades.
Despite the popularity of the topic, there still appears to be no
widely accepted definition for it. This report looks at a small cross
section of academic work and attempts to identify commonality
between the definitions offered for the technical singularity. The
‘technical singularity’ is found to have four main features, all of
which are put together quite effectively in the current Wikipedia
definition. There is also some discussion on the technology
required to achieve and the wider impacts of the technical
singularity, should it occur. It seems that the technology on the
machine side is well underway and on track to deliver our vision
as early as 2030. If we are to follow the transhumanist route then
more research needs to be done on the human brain. It seems that
the widespread disruption to humanity as a whole is inevitable so
we should be preparing for it. But if nothing happens, we just
carry on as normal!
I. INTRODUCTION
T
HE phrase ‘technological singularity’, first coined in a setting
outside of mathematics over 70 years ago, still has no widely
agreed definition. But there is one fact that is certainly agreed
upon, that when it comes it will cause massive change.
Intelligent machines have left the realms of science fiction and
entered our everyday lives. We may not have truly intelligent
machines just yet but we are certainly expecting them.
This is a short report and as a result my discussion needs to be
focused. We will attempt to broadly define the ‘technological
singularity’ as a phrase and also have a look at the technology
trends that seem to be drawing the prospect of the technical
singularity ever closer. This is concluded with a short
discussion on the wider societal impacts of such an event
II. DISCUSSION
A. Defining the ‘Technological Singularity’
Upon inspection of available resources including websites,
books, magazine articles and research papers, it became
obvious that no clear, agreed definition of the technological
singularity exists. Originally a mathematical term to describe a
point at which an object can no longer be defined or no longer
follows a predictable set of rules, the term is now used to
describe a future point in time that is just as impossible to
This paragraph of the first footnote will contain the date on which you
submitted your paper for review. It will also contain support information,
including sponsor and financial support acknowledgment. For example, “This
work was supported in part by the U.S. Department of Commerce under Grant
BS123456.”
The next few paragraphs should contain the authors’ current affiliations,
including current address and e-mail. For example, F. A. Author is with the
National Institute of Standards and Technology, Boulder, CO 80305 USA (email: author@ boulder.nist.gov).
S. B. Author, Jr., was with Rice University, Houston, TX 77005 USA. He is
now with the Department of Physics, Colorado State University, Fort Collins,
CO 80523 USA (e-mail: author@lamar.colostate.edu).
T. C. Author is with the Electrical Engineering Department, University of
Colorado, Boulder, CO 80309 USA, on leave from the National Research
Institute for Metals, Tsukuba, Japan (e-mail: author@nrim.go.jp).
The Technological Singularity
predict.
The term as applied to technological advancement is thought
to have gained popularity after John Von Neumann used the
term ‘singularity’ in a conversation with Stanislav Ulam. Ulam
later quoted this conversation in an obituary to Von Neumann
that was published in an American Mathematical Society
Bulletin [1]. Vernor Vinge is then thought to have been
responsible for mainstream introduction of the term
‘technological singularity’ to the technical world when he used
the term in his book, Marooned in Realtime (1986)[2], and also
later in his paper ‘The Coming Technical Singularity[3]. At this
point the terms ‘singularity’ and ‘technological singularity’
seem to be used interchangeably and equate to essentially the
same thing.
Prominent futurologist, Ray Kurzweil appears to have
brought the term into popular use in the current era of science
fiction and the scientific community seems to have followed
suit.
To find a more uniform definition for the ‘technological
singularity’ we searched a collection of research material[4][17], magazine articles and books[4]-[17], and found the
authors most explicit definition of the technological singularity
or singularity in each. We then compared these definitions and
looked for the similarities or differences between them.
Although only a very small study, the results were pretty
clear. The ‘technological singularity’ has four main features.
First, artificial intelligence reaches human like levels of
intelligence. Second, the AI take over control of their own
design and manufacture, at which point humans effectively lose
control. Thirdly, the machines continue to evolve at an
exponential rate. The fourth feature is that it will change
humanity dramatically and irreversibly.
When used in the same context, singularity and technological
singularity seem to mean essentially the same thing and many
authors use them interchangeably.
B. Driving Technologies
Two main factors influence the approach of the technological
singularity. Firstly micro-processor speed. The ability of a
machine to think at the complexity of a human appears to be
closely related to the speed of the CPU controlling that
machine. Our current fastest Supercomputer is Fugaku[18], in
Japan. This has a speed of 415.53PLFOPS or Petaflops
(1015)[19]. For a machine to achieve human like intelligence
is believed to require roughly an octillion flops (1027), so there
may seem to be a way to go, but at the current rate of
improvement in technology it is predicted to obtain parity with
the human brain as early as 2030[14].
Secondly, it seems quite widely agreed that the technological
singularity will lead to some form of transhumanism[17] or
‘post biology’[7] world. We still have very little understanding
of the human brain at a level of being able to insert technology
into it. However, as Reeke[17] argues, ‘the ultimate goals of
AI and neuroscience are quite similar but that they have
become obscured’. It seems that there is a disconnect between
the two fields and that there perhaps ought to be more
collaboration between them. Ultimately it is likely that the
2
understanding of the human brain is going to be very much the
limiting factor in any transhumanist, post biological future.
But as technology evolves at an ever-increasing rate, so does
our imaging technology and other technology involved in
understanding the workings of the human brain. Neuralink[20]
was set up in 2016 by Tesla boss Elon Musk. Its aim is to
create ‘an integrated brain-machine interface platforms with
thousands of channels’.
C. Wider Impact
From a review of the documents referenced it seems that
whilst there may be some disagreement over what the
technological singularity actually is or when it will come,
there is very little disagreement on the expected outcome. All
texts essentially say the same thing, that humanity will never
be the same. It is not necessarily bad, depending how you look
at it, just very different.
Clearly, an event such as the technological singularity will
bring about huge change for humanity. The wider social,
ethical and moral impacts of such an event for example would
be enormous and at this stage, besides a few of the
futurologists such as Kurzweil, few are prepared to conjecture
just how enormous the change.
Obviously, jobs are a good starting point when attempting to
appreciate the far-reaching implications of the technological
singularity. Without even venturing too far back in our own
history it is possible to see the massive impact automation has
had on levels of jobs across all sectors. But this change has
been slow, because the rate of technological change has been
comparatively slow. The industries have had time to take up
the job losses and create new jobs. But as we near the elbow
of an exponential trend, the rate of improvement in the
technology will explode. Potentially millions of jobs could be
lost overnight as automation begins to accelerate. Eventually,
all jobs will likely be done by machines, leaving humans with
nothing to do but lean back and enjoy the fruits of our labor.
Sure, the journey there will be a little uncomfortable but
eventually we can all live a utopian dream, can’t we? Well
technically yes, but there are numerous other issues to
consider, most beyond the remit of this short paper. However,
perhaps the most interesting discussion is that of social
structure. Humans have a basic instinct to survive and along
with that goes a competitive nature. It is this nature that drives
capitalism. There will always be a winner and a loser, the rich
and the poor. But get rid of all jobs and even if there is still
some form of social ladder still in place, the masses now
require a way to live. Is the answer a minimum living wage?
More fundamentally, is there any evidence to suggest that we
could as a species make such a rapid change in our
foundational systems or are we doomed to wipe ourselves out
in the process? These are all points that are worth noting and
would make interesting discussion but are all substantial
topics so will not be covered here.
III. CONCLUSION
From assessing a small selection of papers and other media it
was pretty clear that ‘technological singularity’ was widely
The Technological Singularity
accepted to have 4 main features as discussed earlier. Like
grout[3], for risk of objection, we are inclined to suggest that
the Wikipedia[24] definition actually works quite well, for all
parties. It is a little wordy in places and academics may want
to trim it down but on the whole it fits the purpose. So ‘the
technological singularity is a hypothetical point in time at
which technological growth becomes uncontrollable and
irreversible, resulting in unforeseeable changes to human
civilization.[2][3] According to the most popular version of
the singularity hypothesis, called intelligence explosion, an
upgradable intelligent agent will eventually enter a
"runaway reaction" of self-improvement cycles, each new
and more intelligent generation appearing more and more
rapidly, causing an "explosion" in intelligence and resulting
in a powerful superintelligence that qualitatively far
surpasses all human intelligence.’.
As an alternative to the ‘technological singularity’, I. J.
Good[15] talked of an ‘intelligence explosion’ which has
a very similar definition. What is impressive with this
report is that it was written in 1965, long before the
concept of a computer was anything like what it is now.
And Vinge[22] in his report on ‘The coming
technological singularity’ makes the concise yet chilling
opening statement, “Within thirty years, we will have the
technological means to create superhuman intelligence.
Shortly after, the human era will be ended.”. Many
prominent futurologists such as Kurzweil[14] predict a
similar trajectory in the technology even based on
current data. So it seems sensible to accept that if it
happens, the singularity will change humanity at the
most fundamental levels and that change will be
irreversible.
Is it likely to happen? I would agree strongly with
Vinge’s statement[22] that ‘if the technological
singularity can happen, it will. Even if all the
governments of the world were to understand the
“threat” and be in deadly fear of it, progress toward the
goal would continue.’ And again, I would have to
strongly agree with that statement. Powerful AI is the
new most potentially powerful weapon of war currently
available to a human. Every global superpower has one
and we would be naïve to believe that their development
will stop over security and other concerns. We have the
nuclear arms race to use as a reference for this claim.
However, in terms of a more solid indication that the
technological singularity will occur, we believe that
Kurzweil[14] makes a convincing case for the ‘law of
accelerating returns’. Essentially technology and even biology
have always, and continue to follow a well defined trend.
They all follow an exponential curve in terms of the rate of
advancement. Exponential trends start out with small
incremental steps but very quickly after the elbow of the curve
the rate of change is huge, and we are on that elbow now.
Kurzweils[14] extrapolation of historical data leads him to
predict the technological singularity will occur between 2030
3
and 2045. At this moment in time that kind of timeline would
seem perfectly feasible.
So where is this going to leave humanity?
Transhumanism seems to be the only feasible outlook?
Once humans lose control of the machines and the
machines enter the process of rapid self-improvement
they would be expected to rapidly approach
superintelligence. However, as this happens, humans
stand still and continue to evolve at painfully low speed.
How quickly do humans outlive any useful purpose and
become surplus to requirement?
If you can’t beat them, join them seems to be the
consensus on humanities long term survival strategy.
Entrepreneur Elon Musk[20, 21] has bet big on this
transhumanist future and in 2016 he started ‘Neuralink’,
a company that researches and produces digital implants
that interface directly with the brain. Once again, I
would agree with the transhuman future theory. It really
does seem that to find any sensible and sustainable way
forward in the post singularity world we would have to
consider merging with the machines.
And if we don’t reach the singularity? Well that’s just fine.
Humanity will surely just continue at a steady pace and the
law of accelerating returns will continue to serve us at a more
manageable speed.
The Technological Singularity
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
S.
Ulam,
“John
Von
Neumann,
19031957,"Bull.Amer.Math.Soc., vol 64, pp. 5, 1958.
V. Vinge, “Marooned in Realtime,” 1st ed, New York, NY,
USA: Bluejay Books Inc., 1986, [Online]. Available:
https://doi.org/10.1090/S0002-9904-1958-10189-5.
V. Grout, “The Singularity Isn’t Simple! (However We Look
at It) A Random Walk between Science Fiction and Science
Fact,” Information, vol. 9, no. 4, p. 99, Apr. 2018.
A. Potapov, “Technological Singularity: What Do We Really
Know?,” Information, vol. 9, no. 4, p. 82, Apr. 2018.
R. Yampolskiy, “The Singularity May Be Near,”
Information, vol. 9, no. 8, p. 190, Jul. 2018.
лю. L. Lazareva. “On the Threshold of Technological
Singularity: Human Readiness To The New Stage of
Evolution.” Antropologìnì Vimìri Fìlosofs'kih Doslìden', no.
14, 2018, pp. 119–131, 2018.
P. Wang, K. Liu, and Q. Dougherty, “Conceptions of
Artificial Intelligence and Singularity,” Information, vol. 9,
no. 4, p. 79, Apr. 2018.
R. Logan, “Can Computers Become Conscious, an Essential
Condition for the Singularity?,” Information, vol. 8, no. 4, p.
161, Dec. 2017.
D. E. Schrader and D. Ghosh, "Proactively Protecting
Against the Singularity: Ethical Decision Making in AI," in
IEEE Security & Privacy, vol. 16, no. 3, pp. 56-63, May/June
2018.
C. Last, “Cosmic Evolutionary Philosophy and a Dialectical
Approach to Technological Singularity,” Information, vol. 9,
no. 4, p. 78, Apr. 2018.
C. Last, “Cosmic Evolutionary Philosophy and a Dialectical
Approach to Technological Singularity,” Information, vol. 9,
no. 4, p. 78, Apr. 2018.
Riccardo Campa, Konrad Szocik and Martin Braddock,
“Why space colonization will be fully automated,” Tech.
Forec. And Soc. Change, vol 143, p.168, June 2019.
A. Braga and R. Logan, “The Emperor of Strong AI Has No
Clothes: Limits to Artificial Intelligence,” Information, vol.
8, no. 4, p. 156, Nov. 2017.
4
[14] R. Kurzweil, “The Singularity Is near : When Humans
Transcend Biology,” New York, New York USA: Viking,
2005.
[15] I. J. Good, “Speculations Concerning the First
Ultraintelligent Machine,” Advances in Computers, vol. 6,
p33, 1965.
[16] J. Cordeiro, “The Boundaries of the Human: From
Humanism to Transhumanism”, World Futures Review, 6(3),
pp. 231–239, 2014.
[17] G. N. Reeke and G. M. Edelman. "Real Brains and Artificial
Intelligence." Daedalus 117, no. 1, 1988: 143-73.
www.jstor.org/stable/20025142
[18] Fugaku
(Supercomputer),
Available
online:
https://en.wikipedia.org/wiki/Fugaku_(supercomputer)
[19] Fujitsu
on
Fugaku,
Available
online:
https://www.fujitsu.com/global/about/resources/news/pressreleases/2020/0622-01.html
[20] Neuralink, Available online: https://www.neuralink.com/
[21] E. Musk, “An integrated brain-machine interface platform
with thousands of channels”, Neuralink, July 2019.
[22] V. Vinge, "The Coming Technological Singularity: How to
Survive in the Post-human Era." NASA. Lewis Research
Center, Vision 21: Interdisciplinary Science and Engineering
in the Era of Cyberspace, p11-22, 1993.
[23] Elon
Musk
on
AI,
Available
online:
https://www.youtube.com/watch?v=B-Osn1gMNtw
[24] Technological
Singularity,
Available
online:
https://en.wikipedia.org/wiki/Technological_singularity
[25] I.
J.
Good,
Available
online:
https://en.wikipedia.org/wiki/I._J._Good
[26] Ray
Kurzweil,
Available
online:
https://en.wikipedia.org/wiki/Ray_Kurzweil
[27] Singularity
Symposium,
Available
online:
http://www.singularitysymposium.com/definition-ofsingularity.html
Download