1. Search Marketers Guide to Our New Robot Overlords
@KelvinNewman - State of Search, 2016
2. @kelvinnewman So, who is this guy? @kelvinnewman
3. Founder of BrightonSEO @kelvinnewman [email protected]
http://www.slideshare.net/kelvinnewman Kelvin Newman
4. @kelvinnewman Im from the United Kingdom @kelvinnewman
5. @kelvinnewman In the seaside resort of Brighton
@kelvinnewman
6. @kelvinnewman Actually; I live in Worthing
7. @kelvinnewman Which is famous for one thing @kelvinnewman
Birdman Competition where people see how far they can jump of the
pier
8. @kelvinnewman Room above the pub I started organising a meet
up in the upstairs room of this pub @kelvinnewman
9. @kelvinnewman And 3500 people came along to the last one
@kelvinnewman
10. @kelvinnewman Because I organise an SEO conference I spend
more time than is healthy thinking about what SEO is really about *
this isnt actually me thinking about the future of search, Im
reading my emails, but it makes me look moody and enigmatic
11. @kelvinnewman Consequently, Im really interested in how
technology works
12. @kelvinnewman But Im most interested in the people who use
technology @kelvinnewman
13. @kelvinnewman And the impact that technology has on people.
@kelvinnewman
14. @kelvinnewman I also spend a lot of time assembling stupid
stuff I bought of eBay
15. @kelvinnewman But I got told thatd be a boring topic for a
talk though
16. @kelvinnewman So what is this presentation about?
@kelvinnewman
17. @kelvinnewman Well, this presentation isnt about
@kelvinnewman
18. @kelvinnewman If you spend fifteen minutes doing this youll
get seventy billion links Its more about getting you to think
differently about the future. And the impact technology has on
people.
19. @kelvinnewman It will introduce you to our new marketing
robot overlords
20. @kelvinnewman Most of the SEO talks about AI, Machine
Learning, robots, etc. focus on the search engine,
21. @kelvinnewman Not the search engine optimizer
22. @kelvinnewman I want to talk about both.
23. @kelvinnewman When we talk about robot overlords you
probably think of things like
24. @kelvinnewman@kelvinnewman
25. @kelvinnewman@kelvinnewman
26. @kelvinnewman
27. @kelvinnewman Or the scariest robot of all, one that will
send shivers down your spine and give you nightmares.
28. @kelvinnewman
29. @kelvinnewman But the ones that should the ones that should
really scare us look very different. @kelvinnewman
30. @kelvinnewman@kelvinnewman
31. @kelvinnewman Or dont even really look like robots at all
@kelvinnewman
32. @kelvinnewman So why should I be scared @kelvinnewman
33. @kelvinnewman Mark Carney Governor of the Bank of England
Hes not a robot you should be scared of. But his organisation is
behind a report that scared me.
34. @kelvinnewman Taking the probabilities of automation, and
multiplying them by the numbers employed, gives a broad brush
estimate of the number of jobs potentially automatable. For the UK,
that would suggest up to 15 million jobs could be at risk of
automation. In the US, the corresponding gure would be 80 million
jobs. Andy Haldane Chief Economist at the Bank of England
35. @kelvinnewman Is 15 Million a lot of Brits though?
@kelvinnewman
36. @kelvinnewman Hands up whos birthday is between Jan 1st and
May 2nd? @kelvinnewman
37. @kelvinnewman Sorry the robots taken your job.
@kelvinnewman
38. @kelvinnewman Thats about a third of the population whos
job is at risk due to robots
39. @kelvinnewman oops I made a mistake 15 million is a third
of the population, but actually about half of the working
population
40. @kelvinnewman Thats 30 times as many U.S jobs as lost in
the 2008 financial crash @kelvinnewman
41. @kelvinnewman
42. @kelvinnewman@kelvinnewman OK, breathe, I dont flip burgers
in McDonalds so Im alright?!?
43. @kelvinnewman
http://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf
Perception & Manipulation Creative Intelligence Social
Intelligence Frey and Osborne have examined how susceptible jobs
are to computerisation
44. @kelvinnewman Three types of task Perception &
Manipulation Robots are still unable to match the depth and breadth
of human perception. Tasks that relate to an unstructured work
environment can make jobs less susceptible to computerisation.
Creative Intelligence The psychological processes underlying human
creativity are difcult to specify. The principal obstacle to
computerising creativity is stating our creative values sufciently
clearly that they can be encoded in an program. Social Intelligence
Human social intelligence is important in a wide range of work
tasks, such as those involving negotiation, persuasion and care.
The realtime recognition of natural human emotion remains a
challenging problem.
45. @kelvinnewman Things people are good at Tasks requiring
perception in unstructured environments. Creative tasks which rely
both on novelty and value. Non-routine social tasks which rely on
hard to articulate common sense Things machines are good at Routine
tasks in structured environments. Tasks which require novelty and
experimentation. Analysis at a speed and scale.
55. @kelvinnewman Answer the Public scrapes and probes Google
Suggest answerthepublic.com
56. @kelvinnewman These sort of things arent sexy But are
likely to have the biggest most immediate impact on your life.
@kelvinnewman
57. @kelvinnewman Machine learning is sexy And Articial
Intelligence, Deep Learning and other sort of related things along
those lines. @kelvinnewman
58. @kelvinnewman Artificial Intelligence, Machine Learning and
Deep Learning Whats the difference between
https://blogs.nvidia.com/blog/2016/07/29/whats-difference-articial-intelligence-machine-learning-deep-learning-ai/
59. @kelvinnewman I studied media studies at Uni not computer
science or anything like it Now feels like a good time to share
@kelvinnewman
60. @kelvinnewman Computer systems able to perform tasks
normally requiring human intelligence, such as visual perception,
speech recognition, decision-making, and translation between
languages. What is artificial intelligence? @kelvinnewman
61. @kelvinnewman The wide availability of GPUs that make
parallel processing ever faster, cheaper, and more powerful. All
aided by practically innite storage and a ood of data sources. Why
AI has exploded? @kelvinnewman
62. @kelvinnewman Theres three levels of articial intelligence
in rising levels of nightmarish-ness Difference between narrow,
general and super intelligence
63. @kelvinnewman AI that specializes in one area. Theres AI
that can beat the world chess champion in chess, but thats the only
thing it does. Ask it to gure out a better way to store data on a
hard drive, and itll look at you blankly. Artificial Narrow
Intelligence (ANI)
64. @kelvinnewman A computer that is as smart as a human across
the boarda machine that can perform any intellectual task that a
human being can. Creating AGI is a much harder task than creating
ANI, and were yet to do it. Artificial General Intelligence
(AGI)
65. @kelvinnewman Mustafa Suleyman Our mission is to solve
general intelligence
66. @kelvinnewman Articial Superintelligence ranges from a
computer thats just a little smarter than a human to one thats
trillions of times smarteracross the board. Artificial
Superintelligence (ASI)
67. @kelvinnewman ASI is likely to either mean we become
immortal or face extinction @kelvinnewman
68. @kelvinnewman The 24k words on AI on this blog will
literally blow your mind http://waitbutwhy.com/ 2015/01/articial-
intelligence- revolution-1.html
69. @kelvinnewman But enough of the sci-fi dystopia stufff
70. @kelvinnewman Machine Learning Machine Learning at its most
basic is the practice of using algorithms to parse data, learn from
it, and then make a determination or prediction about something in
the world. So rather than hand-coding software routines with a
specic set of instructions to accomplish a particular task, the
machine is trained using large amounts of data and algorithms that
give it the ability to learn how to perform the task.
71. @kelvinnewman Deep Learning A branch of machine learning
based on a set of algorithms that attempt to model high level
abstractions in data by using a deep graph with multiple processing
layers, composed of multiple linear and non-linear
transformations.
72. @kelvinnewman
73. https://www.flickr.com/photos/61408718@N02/6333712753/ When
Robots impressed me
74. @kelvinnewman That previous slide was made by slidebot.io
25,352,683 images combine with 10,000+ design rules and 237 unique
layout styles to create one powerful presentation.
75. @kelvinnewman
76. @kelvinnewman The rest of the slides I made myself. But
could the machine of done better?
77. @kelvinnewman
78. @kelvinnewman
79. @kelvinnewman Spotifys key staff share how it works
https://www.youtube.com/watch?
v=qhYlmnqSDV0&index=7&list=PLSRBXQp3FvKJa4J5ZlKsupg189WJ4YhYc
80. @kelvinnewman Beatles articial intelligence Flow Machine -
Daddys Car http://www.ow-machines.com/ai-makes-pop-music/ A
Beatles-esque song written by artificial intelligence
@kelvinnewman
81. @kelvinnewman To me it sounds more like the Beach Boys than
the Beatles, but still hugely impressive
82. @kelvinnewman https://github.com/david-gpu/srez A Neural
Networks that can reconstruct Pixelated faces
83. @kelvinnewman An Artificial Intelligence that finds art
like current news photos http://recognition.tate.org.uk/
84. @kelvinnewman Newspaper sports reports are often generated
by machines Friona fell 10-8 to Boys Ranch in five innings on
Monday at Friona despite racking up seven hits and eight runs.
Friona was led by a flawless day at the dish by Hunter Sundre, who
went 2-2 against Boys Ranch pitching. Sundre singled in the third
inning and tripled in the fourth inning Friona piled up the steals,
swiping eight bags in all
https://www.wired.com/2012/04/can-an-algorithm-write-a-better-news-story-than-a-human-reporter/
85. @kelvinnewman The software also generate natural language
Google Analytics Reports.
https://www.narrativescience.com/quill-engage
86. @kelvinnewman IBM Watson and cancer treatments
http://www.nytimes.com/2016/10/17/technology/ibm-is-
counting-on-its-bet-on-watson-and-paying-big-money-for-
it.html?_r=0
87. @kelvinnewman At the University of North Carolina School of
Medicine, Watson was tested on 1,000 cancer diagnoses made by human
experts. In 99 percent of them, Watson recommended the same
treatment as the oncologists. In 30 percent of the cases, Watson
also found a treatment option the human doctors missed. Some
treatments were based on research papers that the doctors had not
read more than 160,000 cancer research papers are published a
year.
88. @kelvinnewman Eliza simulated conversation by using a
'pattern matching' and substitution methodology that gave users an
illusion of understanding on the part of the program, but had no
built in framework for contextualizing events.
https://en.wikipedia.org/wiki/ELIZA Eliza the 1960s rubbish robot
people loved
89. @kelvinnewman
https://theringer.com/meet-the-man-who-created-a-virtually-immortal-joey-
tribbiani-21b910295b1f#.ykixtfnhz A video chat bot that generates
new lines for Joey from friends
90. @kelvinnewman When the robots disappointed me
91. When 30 of the USs biggest companies lost 9% of their value
for no apparently caused by algorithmic trading May 2010 Flash
Crash I did a whole deck about the the implications of the ashcrash
on how we think about SEO http://bit.ly/ashcrashlessons
92.
@kelvinnewmanhttps://blog.arik.io/facebook-m-the-anti-turing-test-74c5af19987c#.mq0ix93r2
Facebooks Fake Robots Their AI Assistants sometimes are actually
people
93. @kelvinnewman
http://arstechnica.co.uk/business/2016/09/at-the-bleeding-edge-of-ai-quantum-grocery-picking-and-transfer-learning/
Everything but packing the bag Ocado have automated nearly every
aspect of their business apart from packing the shopping bags
@kelvinnewman
94. @kelvinnewman Theres software used across the country to
predict future criminals. And its racist.
https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-
sentencing
95. @kelvinnewman Should we resist and fight back?
96. @kelvinnewman https://en.m.wikipedia.org/wiki/
William_Lee_(inventor) Elizabeth 1st refused patent for knitting
machine out of fear it would cause mass unemployment
97. @kelvinnewman http://www.slate.com/articles/business/
the_dismal_science/1997/01/ the_accidental_theorist.html Technology
tends to ultimately displace rather than completely replace
jobs.
98. @kelvinnewman Anything that is in the world when youre born
is normal and ordinary and is just a natural part of the way the
world works. Douglas Adams @kelvinnewman
99. @kelvinnewman Anything thats invented between when youre
fifteen and thirty-five is new and exciting and revolutionary and
you can probably get a career in it. Douglas Adams
@kelvinnewman
100. @kelvinnewman Anything invented after youre thirty-five is
against the natural order of things. Douglas Adams
@kelvinnewman
101. @kelvinnewman Im 33, so I hoping the rate of change slows
down a bit
102. @kelvinnewman So what are the actions? @kelvinnewman
103. Things you can do when you get back to the office
@kelvinnewman
104. @kelvinnewman Panic! and drink
105. @kelvinnewman Buy shares in Uber, Amazon and Google
@kelvinnewman
106. @kelvinnewman OK, heres some serious advice
@kelvinnewman
107. @kelvinnewman Never do a job a machine could do quicker Or
better than you @kelvinnewman
108. @kelvinnewman Examine the rough edges and the exceptions
to the rules. @kelvinnewman
109. @kelvinnewman
110. @kelvinnewman
111. @kelvinnewman
112. @kelvinnewman
113. @kelvinnewman Stay ten minutes ahead of Google
@kelvinnewman
114. @kelvinnewman Use the tech before it becomes part of the
core algo @kelvinnewman
115. @kelvinnewman Google are taking machine learning for image
recognition seriously
116. @kelvinnewman
117. @kelvinnewman image recognition open source GoogleBrain
have released their image captioning system available as an open
source model in TensorFlow.
https://research.googleblog.com/2016/09/show-and-tell-image-captioning-open.html
118. @kelvinnewman
https://research.googleblog.com/2016/09/introducing-open-images-dataset.html
Released a huge learning set of images - a dataset consisting of ~9
million URLs to images that have been annotated with labels
spanning over 6000 categories
119. @kelvinnewman Is it part of the core algorithm
120. @kelvinnewman Probably not
121. @kelvinnewman Will it be?
122. @kelvinnewman Certainly
123. @kelvinnewman Is the lag an opportunity?
124. @kelvinnewman Certainly
125. @kelvinnewman How much could you improve you alt text on
your website if you ran your images through this captioning
system?
126. @kelvinnewman Use some off the shelf options
127. @kelvinnewman Altify Uses deep learning to caption images
in an HTML file and fills out its alternative text attributes with
the related caption https://github.com/ParhamP/altify
@kelvinnewman
128. @kelvinnewman https://algorithmia.com/ A marketplace for
algorithms
129. @kelvinnewman
130. @kelvinnewman
131. @kelvinnewman Tutorials on use Machine Learning on over 1M
hotel reviews finds interesting insights
https://blog.monkeylearn.com/machine-learning-1m-hotel-reviews-nds-interesting-insights
132. @kelvinnewman Leave you with a final thought
@kelvinnewman
133. @kelvinnewman Your job is to be Luke Skywalker
Inexplicably understand the robots when nobody else can
134. @kelvinnewman@kelvinnewman
135. Thanks for listening Kelvin Newman - Founder of
BrightonSEO @kelvinnewman [email protected]
http://www.slideshare.net/kelvinnewman