Scribd
Upload
65 views2 pages

The Wonderful and Terrifying Implications of Computers That Can Learn

The document discusses the capabilities of computers that can learn through machine learning and deep learning techniques. It describes how computers can now learn, listen, see, and write through analyzing large amounts of data. Both opportunities and challenges of this development are mentioned. Computers may be able to perform many service sector jobs involving reading, writing, speaking, listening and looking at things. This could significantly impact human labor and the economy. Separating labor from earnings and moving to a craft-based economy may help address the challenges.

Uploaded by

Blom my
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
65 views2 pages

The Wonderful and Terrifying Implications of Computers That Can Learn

The document discusses the capabilities of computers that can learn through machine learning and deep learning techniques. It describes how computers can now learn, listen, see, and write through analyzing large amounts of data. Both opportunities and challenges of this development are mentioned. Computers may be able to perform many service sector jobs involving reading, writing, speaking, listening and looking at things. This could significantly impact human labor and the economy. Separating labor from earnings and moving to a craft-based economy may help address the challenges.

Uploaded by

Blom my
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

MIS | Tutorial 2 L.W.

Video TEDxBrussels | Jeremy Howard:


The Wonderful and Terrifying Implications
Of Computers That Can Learn
Computers can…

LEARN
 Arthur Samuel, father of machine learning
 One of the first examples: Google
 Commercial successes: Amazon, Netflix
 Algorithms that learn by themselves instead of being programmed
 Other examples: Self-driving cars
 Deep-learning
o Insprired by how the human brain works
o The more data you give it, the better it gets
o Wikipedia: Deep learning (also known as deep structured learning, hierarchical learning or
deep machine learning) is a branch of machine learning based on a set of algorithms that
attempt to model high level abstractions in data.
 An observation (e.g., an image) can be represented in many ways such as a vector
of intensity values per pixel, or in a more abstract way as a set of edges, regions of
particular shape, etc. Some representations are better than others at simplifying
the learning task (e.g., face recognition or facial expression recognition)
 One of the promises of deep learning is replacing handcrafted features with
efficient algorithms for unsupervised or semi-supervised feature learning and
hierarchical feature extraction.

LISTEN
 Speaking Chinese:
 Transcription of English, the translation to Chinese in text, Construction of voice

SEE
 2011: German Traffic Sign Recognition Benchmark
 Better than algorithm and better than people
 2012: Deep learning algorithms of Google learning with Youtube videos
 Now about 6% error of image recognition
 Chinese Baidoo: System finds similar images when uploading a photo (different from Google
where the search engines looks for text on websites to find photos)

READ
 ...
MIS | Tutorial 2 L.W.

WRITE
 Algorithm can describe pictures without having seen these situations

Great opportunities:
 Create algorithms of medical use without background in medicine
 Help prognosis to diagnose cancer /predicting survival rates
 Identify cancer areas in tissues under microscope
 Human and computer working together, where human “teaches” computer things
o Example: Classifying images
o “Learn things that would have taken 7 years by several people in 15 minutes by one
person”
 Fix lack of medical expertise in developing world (World Economic Forum)

Challenges:
 The service sector (80% ww):
Reading & Writing, Speaking &
Listening, Looking at things,
Integrating Knowledge

 Computers can do these jobs


 Hard to estimate what the
“new” jobs for humans will be
 Machine Learning Revolution
will develop exponentially, in a
far more intensive way than
the industrial revolution ever
did
 “The better computers get at
intellectual activities, the more
they can build better
computers to be better at
intellectual capabilities”
 Impact already to be seen these days:
Capital productivity has increased over the last 25 years, whereas labour productivity has rather
been flat

 What doesn’t help:


o Better education
o More incentives to work
 What does help:
o Separate labor from earnings
o Craft-based economy

You might also like