Artificial

intelligence
 Question # 1
What are examples of AI?
https://assistant.google.com/
https://ada.com/
https://quickdraw.withgoogle.com/
https://open.spotify.com/
What other examples of AI do you know?
 Question # 2
What is artificial intelligence?
 Artificial intelligence (AI) is an intelligent assistant.

https://www.goodreads.com/review/show/4276261415
Page 22.
 AI has the ability to
 anticipate needs of people and
 make recommendations to people.

https://hbr.org/2020/03/ai-is-changing-work-and-leaders-need-to-adapt
 Most artificial intelligence examples such as self driving cars rely
heavily on deep learning and natural language processing.

Using these technologies, computers can be trained to accomplish

specific tasks by processing large amounts of data
and recognizing patterns in the data.

https://www.sas.com/en_us/insights/analytics/what-is-artificial-intelligence.html
 Machine learning is part of / a subset of artificial intelligence.

Machine learning involves computational methods that use past

information / data, which has been labeled, to
improve performance or make accurate predictions.

https://www.goodreads.com/review/show/4276261415
Pages 52 and 62.
 Deep learning is part of / a subset of machine learning.

In deep learning, task are broken down and

distributed onto machine learning algorithms
that are organized in consecutive layers.

https://www.goodreads.com/review/show/4276261415
Page 62.
 A small, representative set of data is better than a large,
unrepresentative set of data.

Example: If 10,000 images of faces represent 8 out of 10

skin tones, AI cannot recognize the 2 skin tones it has not
been exposed to at all.

https://www.goodreads.com/review/show/4276261415
Page 106.
 AI is an old field, and few ideas are truly new.

The current advances in deep learning have their roots in neural

network research from the 1980s and 1990s, and that research in
turn was based on ideas and experiments from all the way back
in the 1940s.

http://togelius.blogspot.com/2017/07/some-advice-for-journalists-writing.html
 Question # 3
What is the purpose of
artificial intelligence?
 What is the
purpose of
artificial
intelligence?

https://www.weforum.org/agenda
/2018/11/chatbots-without-a-
cause-why-conversational-ai-wont-
work-without-purpose
 Artificial intelligence (AI) is enabling vertical farming in which plants
can be grown in city warehouses in 10 meter stacks of trays.

Such vertical farms require less water and fertilizer than

traditional farms.

https://www.goodreads.com/review/show/2570746538
 Question # 4
What are machines really good at?
 Machines can answer specific questions.

https://www.goodreads.com/review/show/4276261415
Page 22.
 Computers such as calculators can calculate numbers
much faster than humans.

https://www.goodreads.com/review/show/4276261415
Page 6.
 Machines are good at
 doing repetitive, routine work and
 analyzing large data sets.

https://www.goodreads.com/review/show/2570746538
Page 5.
 A goal of the Chinese government is to be able to use
its cameras to identify people within 3 seconds
with 90% accuracy.

https://www.goodreads.com/review/show/4276261415
Page 119.
 AI techniques from deep learning, image classification and object
recognition can be used to find cancer on magnetic resonance
imagining (MRIs) with the same accuracy as highly trained
radiologists.

https://www.sas.com/en_us/insights/analytics/what-is-artificial-intelligence.html
 AI is particularly good at analyzing images.

For example, you can imagine taking a mobile phone and

uploading an image and training an AI system to say, “Is this
likely to be skin cancer or not?”

https://www.mckinsey.com/featured-insights/artificial-intelligence/ai-ethics-podcast
 AI can be used to monitor buildings and machines
- forecasting when and what to repair before breakdowns.

https://www.goodreads.com/review/show/4276261415
Page 80.
 Precision agriculture requires a vast network of Internet of
Things sensors.

This information might include images by satellites

or drones, sensors in the field and sensors on
farm equipment.

https://www.goodreads.com/review/show/2570746538
Page 35.
 Machines can compose text that is almost
indistinguishable from human writing.

https://www.goodreads.com/review/show/4276261415
Page 5.
 AI can transform the field of law. An example: Negotiating and
finalizing a contract includes that lawyers on each side manually
review, edit and exchange documents. This process can
take a long time, for example because contracts are long. There is an
opportunity to automate this process.

AI technology will broaden a lawyer’s role from a narrow focus on risk

mitigation to more strategic engagement on company initiatives.

https://www.forbes.com/sites/robtoews/2019/12/19/ai-will-transform-the-field-of-law/#5f3901447f01
 AI will disrupt every phase of the scientific process:
1. Ask questions and make observations.
2. Devise hypotheses.
3. Design experiment and test predictions based on hypotheses.
4. Run tests and collect data.
5. Develop generalized theories.

https://www.goodreads.com/review/show/2570746538
Page 69.
 Question # 5
How can people help machines?
 Humans can understand the purpose of doing something.

https://www.goodreads.com/review/show/4276261415
Page 6.
 Humans can train machines to perform tasks.

https://www.goodreads.com/review/show/2570746538
Page 107.
 Empathy trainers can teach AI systems
how to display compassion.

https://www.goodreads.com/review/show/2570746538
Page 117.
 Personality trainers can help AI systems
become more humanlike.

https://www.goodreads.com/review/show/2570746538
Page 118.
 Worldview and localization trainers help chatbots become
more sensitive to different ways of thinking and
communicating across the globe.

https://www.goodreads.com/review/show/2570746538
Page 119.
 AI safety engineers
 anticipate unintended consequences of AI
systems and
 address any harmful occurrences with the
appropriate strategy.

https://www.goodreads.com/review/show/2570746538
Page 129.
 Humans can explain machine outcomes.

https://www.goodreads.com/review/show/2570746538
Page 107.
 There is a need for people to learn more about how to
 think creatively, for example rethinking how people work,
 make the right judgments and
 communicate effectively.

https://hbr.org/2020/03/ai-is-changing-work-and-leaders-need-to-adapt
What else can people do to help machines?
Other sources of inspiration
https://deepdreamgenerator.com/
https://www.instagram.com/botsofnewyork/
https://openai.com/blog/openai-api/
https://thispersondoesnotexist.com/