Artificial Intelligence (AI) is quickly moving from a science-fiction concept to reality where machines now have the capability to perform tasks commonly associated with humans. We are starting to see our society transformed because of AI, so having a better understanding of what it is and what it is capable of doing is essential. AI helps power Amazon’s Alexa personal assistant, Google’s Deep Dream neural network, various marketing initiatives, health applications, the aviation industry, and much more. In this keynote,
- Discover what Artificial Intelligence (AI) is and how it is becoming a “machine trait.”
- Gain an appreciation of AI pioneers like John McCarthy, Alan Turing, Marvin Minsky et al.
- Learn how AI works and explore some applications that could play a role in your library.
- Reflect on the future of AI and the implications for libraries and society in general.
Presented at the Leading Edge Libraries Conference, hosted by the Special Library Association Florida and Caribbean Chapter. Orlando, Florida. September 21, 2018. Keynote address.
13. AI Applications:
• Game playing - To beat a world champion by brute force and known reliable heuristics requires
being able to look at 200 million positions per second.
• Speech recognition - it is possible to instruct some computers using speech, most users have
gone back to the keyboard and the mouse as still more convenient.
• Understanding natural language - The computer has to be provided with an understanding of the
domain the text is about, and this is presently possible only for very limited domains.
• Computer vision - full computer vision requires partial three-dimensional information that is not
just a set of two-dimensional views. At present there are only limited ways of representing three-
dimensional information directly, and they are not as good as what humans evidently use.
• Expert systems - A ``knowledge engineer'' interviews experts in a certain domain and tries to
embody their knowledge in a computer program for carrying out some task.
• Heuristic classification - most feasible kinds of expert system is to put some information in one of
a fixed set of categories using several sources of information.
Source: https://goo.gl/g1q7vg
14. “If we want our machines to possess anything
approximating human intelligence, maybe we should
think about giving them a childhood, too.”
- Alison Gopnik
Source: https://goo.gl/6GVKkp
16. Professor John McCarthy, 1958 - 2011
Image source: http://jmc.stanford.edu
“I don't see that human intelligence is something
that humans can never understand."
Founded the field of Artificial Intelligence
17. Alan Turing, (1912–1954)
Educator, Mathematician
“A computer would deserve to be called
intelligent if it could deceive a human into
believing that it was human.”
21. “Inspirational” posters powered by Inspirobot.me
“Forever generating unique inspirational quotes for the endless
enrichment of pointless human existence.”
22. Microsoft deletes 'teen girl' AI after it became a
Hitler-loving sex robot within 24 hours
Source: https://goo.gl/qydQj1
23. Black Mirror is a great TV show that
highlights our “collective unease about
our modern world.”
Technology is the villain!
When a screen is off it is like a “black
mirror”, so perhaps it is a warning?
Funny thing: we are watching the show
on our screens!
24. • This was my idea when I was
much younger, so whatever!
• Bumblebeespaces uses smart
robots and Artificial Intelligence
to "increase usable space and
reduce your footprint by
storing, managing, and
retrieving objects on-demand
in your ceiling." YOUR AI-BUTLER
25. United States Tennis Association (USTA) has partnered with IBM to use Watson to help
review hours of match footage and automatically identify and index key points and stats.
Source: https://goo.gl/bHNoHj
They created a “virtual concierge” (an AI-powered chatbot) that answers questions about
scoring, schedules, transportation, dining options and more.
29. "Deconstructing Symmetric Encryption" was written by an artificial
intelligence and published instantly under my name.
Read at https://goo.gl/vh8vFP
31. Natural Language Processing
• Car User Interface (CUI?)
• Voice biometrics
• How do we interact with connected devices?
• How WILL we interact with connected devices?
• App based? Too many apps for one device (e.g., Philips Hue)
• Sentient?
• Hello Natural Language!
• Internet of Things (IoT) will continue to capture, track, and
share data. What will AI do with it?
40. Examples of object detection using TensorFlow
Slide courtesy of Owen Cegielski, Social Studies Teacher at STEM School Highlands Ranch
41. Example of how data from
Tensorflow could be
viewed geospatially in
Google Earth Engine.
Note: This example was
developed by another
corporation, but it’s similar
in concept, in which
students can map human
activity on the ground.
Slide courtesy of Owen Cegielski, Social Studies Teacher at STEM School Highlands Ranch
42. OpenCV was built to provide a common
infrastructure for computer vision applications
and to accelerate the use of machine
perception in commercial products.
Open Source Computer Vision Library is
an open source computer vision and
machine learning software library.
https://opencv.org
43. Painting by Harold Cohen's AARON sold for $2,000!
Source: https://goo.gl/UUpvwM
AI Artist
53. Revolving Discovery, which sees retailers
developing in-store programming and
thematic displays to give customers
something new each time they visit.
Libraries??!!
Source: https://goo.gl/W9kZ2b
Retail OS (or maybe Library OS) would use an underlying layer
of technology (perhaps AI and biometrics) to enhance the in-
store (or in-library) experience.
57. New Google AI Can Have Real Life Conversations With Strangers
Video: https://youtu.be/lXUQ-DdSDoE
58. “A super intelligent AI will be extremely good at accomplishing its
goals, and if those goals aren't aligned with ours, we're in trouble.
- Stephen Hawking
59. “… to fund war gaming, experimentation and the demonstration of new technologies aimed at ensuring
a continued military edge over China and Russia.” The military is looking to develop more advanced
weapons technologies that will include autonomous weapons and deep learning machines.”
Source: https://goo.gl/goNL1E
60. Keep your eye on Google:
• Google Neural Machine Translation - Bridging the gap between human
and machine translations.
• Google Deep Dream – Inside an AI brain. A computer vision program that
uses a neural network to find and enhance patterns in images.
• Google Reply - an AI-based messaging technology for Android.
• Google Now - predictive cards with information and daily updates in the
Google app for Android and iOS.
• Google Brain - deep learning artificial intelligence that combines open-
ended machine learning research with system engineering and Google-
scale computing resources.
61.
62. Most machine learning researchers expect machines
will be able to create top quality music by 2036.
More studies: https://goo.gl/X6pK2F
Predictions of Human-Level AI Timelines
72. You will program your AI as a game simulation, similar to a “Choose your Own
Adventure!” Requirements:
1) Explore use of the software at: https://api.ai/. A good written guide can be
found here. Please view this STEM student-made video tutorial here which
includes additional code to help your AI have a temperament, an AI with
attitude! You may also need to download the Google Home App on Google Play
or Itunes. Once you program your AI with the API.AI software, you’ll need export
and upload it into Google Home via the APP. Extra Note: Google Home
simulator. Other software option: TensorFlow for more advanced “Machine
Learning.”
The WWI AI Lesson:
Kaiser Wilhelm II brought back to life 100 years later!
Owen Cegielski, Social Studies Teacher at STEM School Highlands Ranch
73. ● 2) Based on your collected information above, program a dialogue in AI, in which you defend your country’s position. Your AI will
be the historical leader of your country who will interact with a yet to be determined delegate team of another country. So, for
example, your AI will be activated by “OK, Tsar Nicholas II.” Your AI should be prepared to present biographical detail of your
country’s leader.
● 3) Next, program your country’s initial position in a short speech, providing some historical background leading up to the July
Crisis. Begin your dialogue initially hostile and aggressive, defending why you want to declare war. Remember, you will also need
to predict what questions your AI will be asked and feature an appropriate response.
● 4) Program different paths and outcomes! If delegates (game players) from another country ask you the “right” questions and
provide appropriate responses, you will begin to consider various paths of peace. If opposing delegates ask you the “wrong”
questions, you will move towards the path of declaring war! So, your AI game will feature two game outcomes: war or peace!
More advanced coders can also program “termperaments” (see guides below for further instructions!)
● 5) Your AI simulation should last approximately 5-10 minutes for the gameplayer, present plenty of historically accurate
biographical info and historical background, and feature multiple paths and outcomes, leading to peace or war depending upon the
player’s actions.
Step 3: Once all groups have finished programming their AI’s, we will rotate delegation groups, facing the different AI simulations from all
countries represented. If enough groups succeed with convincing the AI’s to pursue a path to peace, then World War I will be averted. If
not, the War begins!
The WWI AI Lesson: The Process