Sunday, March 27, 2016

Issues of AlphaGo's victory

In my last post, "Victory for AI", I wrote about the computer program AlphaGo and its victory in a game of Go with Lee Se-dol of South Korea, the world champion of the game. So many news sites, blogs and websites have written about this news, so there's plenty for ESP teachers to choose from to use for discussion, technical description, or focus on artificial intelligence.

One of the main areas focused on is the question of the impact this development could have - will AI eventually become smarter than humans and take over the world. Many articles see this as a possibility, while others feel this is an unlikely scenario.

An article in the Sunday Times (Great Britain) presents some of the issues, and ends with the writer's opinion. The article is, "I for one welcome the rise of the robots. They can do the work while I play."

Link: http://www.thesundaytimes.co.uk/sto/comment/columns/dominiclawson/article1679638.ece

Some of the points mentioned:
  • "why Go has been seen as such a good testing ground for the holy grail of modern computing, artificial intelligence"
  • "Hassabis, however, seems most gripped by what it could do to solve humanity's most pressing anxieties."
  • "Will these programs develop to the stage when most, if not all, of our professions are made redundant?"
  • "Others worry most about the military applications of artificial intelligence."
  • "And what if such technology got into the wrong hands?"
  • The author quotes Stephen Hawking: "The primitive forms of artificial intelligence we already have have proved very useful. But I think the development of full artificial intelligence could spell the end of the human race."
  • And the author's opinion: "And would it matter if the existing skills of the rest of us were largely rendered redundant by his work? If they were, it would mean that AI had given the world a superabundance of what it needs to live comfortably, at a very low cost."


For a focus on some of the ethical issues involved, I checked one of my favorite blogs, Engineering Ethics. I wrote about the blog in my post of January 14, 2015 ("Relevant links"), and there is a link to it under Relevant Links on this blog. The post is titled, "AlphaGo Deeats Human Go Champion: Go Figure."

Here is the link to this specific post:  http://engineeringethicsblog.blogspot.co.at/2016/03/alphago-defeats-human-go-champion-go.html

The blogger refers to the article in Wired magazine that I mentioned in my last post. He is also of the opinion that intelligent machines are unlikely to become more powerful than humans. At the end of his post he comments:

"And as long as enough people remember than humans are not machines, and machines are made by, and should be controlled by, humans, I think we don't have to lose a lot of sleep about machines taking over the world. What we should watch are the humans running the machines." (emphasis is mine)

I think this point is an important one, and it has come up often in discussions about issues in engineering ethics that I've had with my students.

I also looked at the blog Robotics and AI - also on my Relevant Links - since this is an area the blog particularly focuses on. Their article is "Machine vs. Man. What happened when AI played against a World champion."

Link to post:  http://www.roboticsandai.com/2016/03/machine-vs-man-what-happened-when-ai-played-against-a-world-champion/

This post gives a general overview of the news and, not surprisingly, concludes with:

"one thing is clear ... machines with intuitive capabilities are a not too distant reality."

Whether this will be a good thing or not -- well, that's an interesting topic to discuss with our students.

Wednesday, March 23, 2016

Victory for AI

Demis Hassabis (photo by David Ellis for Google)
The AlphaGo computer program, developed by Demis Hassabis for the Google-owned company DeepMind, won a series of matches (9-15 March) of the game Go against Lee Se-dol, a champion of the game.

To a non-techie like me, this doesn't seem like such earth-shattering news. But it is being hailed as a great victory for Artificial Intelligence. In fact, in looking for articles as material to use with my students, I've been spoilt for choice.

First it's necessary to understand the complexity of the game Go. In an article from The Telegraph, it is described thus:

"Go takes a lifetime to master and unlike chess, a computer cannot play by simply assessing all possible moves but must rely on something akin to intuition."

Link to article, "Google's DeepMind AI makes history by defeating Go champion Lee Se-dol:
http://www.telegraph.co.uk/technology/2016/03/09/googles-deepmind-beats-go-champion-in-historic-moment-for-ai/

I also looked at an article on The Verge website, since it's oriented towards more technical readers. The article's title, "Google's DeepMind defeats legendary Go player Lee Se-dol in historic victory," seemed rather similar to the one from The Telegraph, since they both refer to the victory as an historic moment.

Link: http://www.theverge.com/2016/3/9/11184362/google-alphago-go-deepmind-result

In addition to the detailed information about the game, about the computer program and about why this is significant for artificial intelligence, I decided to focus on the language in both articles that referred to the significance of this development.

Here are some examples.

From The Telegraph article:
  • makes history
  • has beaten ... for the first time in history
  • fiendishly complex
  • a "suberb" game that would be studied for years to come
  • the breakthrough
  • a watershed moment for artificial intelligence
  • a milestone potentially more significant than IBM ...
  • Go takes a lifetime to master
  • (a computer) must rely on something akin to intuition
  • this amazing program
From The Verge article:
  • historic victory
  • a huge milestone
  • legendary Go player
  • historic matches
  • the first time
  • such a perfect manner
  • amazing game
  • pushed AlphaGo to its limits
  • one of the great challenges faced by AI
  • made it tough for computers to crack
  • the most advanced effort yet
  • a complex network of deep neural networks and machine
  • a colossal moment for AI

In Wired magazine as well, an article uses much of the same vocabulary to describe the significance of this victory for artificial intelligence.

The article is "In Two Moves AlphaGo and Lee Seedol Redefined the Future."

Link: http://www.wired.com/2016/03/two-moves-alphago-lee-sedol-redefined-future/

Some vocabulary from the beginning of the article:
  • redefined the future
  • the enormously powerful and rather mysterious talents of modern artificial intelligence
  • machines are now capable of moments of genius
  • the historic match
  • the machine claimed victory
  • the victory is notable
  • the technologies at the heart of AlphaGo are the future
  • poised to reinvent everything

In fact, all of the articles I've looked at so far refer to the match as "historic" or "making history." This will be interesting to compare with other historic developments in artificial intelligence - and what the possible impacts could be. This is an aspect I will look at in my next post.

Sunday, March 13, 2016

Inventor of email has died

Ray Tomlinson
In my post of February 6, 2016, titled Pioneer in AI has died, I wrote about Marvin Minsky, the founding father of artificial intelligence. Unfortunately, another pioneer has recently died -- one whose invention has had an impact on the lives of everyone who is reading this post, as well as most other people in the world.

On March 5, 2016 Ray Tomlinson, the man who invented email, died aged 74.

In the numerous articles about him that have been published recently, the other aspect of his invention that has been focused on is his choice of the @ symbol for use in email addresses. For choosing articles, information and ideas to share with our students, it is difficult to decide whether to focus on his particular achievements, the impact those achievements have had, the characteristics he had that made him an innovator, or his particular choice of this symbol, which has become the sign of the internet age. Whichever you choose, there are articles to focus on.

Here is just a small sampling of the articles I have already gathered to work with in my lessons:

1) Email inventor Ray Tomlinson dies at 74 from the BBC news website
This is a fairly short, straightforward news item with the basic facts about Ray Tomlinson and his invention of email. What I found particularly interesting about this article is that at the end there are the following questions for readers:

Do you have a funny, interesting, or shocking story about an email you have sent or received? Do you regret sending an email, or did you send an email that changed your life? We want to hear your email stories. Contact us on haveyoursay@bbc.co.uk.

This could be a very good writing activity for students; perhaps they could first discuss ideas in the classroom, and colleagues could ask questions to elicit more details or further information. After the lesson they could write their story as a homework assignment - or be inspired to send their story to BBC news for possible publication (quite motivating!).

2) The Man Who Made You Put Away Your Pen from National Public Radio
This is not only an article, it is also the text of an interview with Ray Tomlinson on the National Public Radio program. The recording of this interview is also at this site (4:03). In the interview, Tomlinson explains how and why he developed email, as well as why he chose the @ symbol to separate the name of the sender from the name of the machine.

This is a good basis from which students could discuss how ideas are transformed into reality, and how necessity is truly "the mother of invention."

3) How One Man Saved the '@' Symbol from Time magazine website
For further information about the choice of the @ symbol for email - thus making it one of the most recognized symbols on the internet - this article focuses on why this symbol in particular was chosen (linking back to the NPR article).

This is an aspect of this innovation that might not be as well known to our engineering students, but is in any case very relevant in discussing the impact of email.

4) merchant@florence wrote it first 500 years ago from The Guardian
As a follow-up to the discussion about the choice of @ for email is the information in this article about when this symbol first appeared, and how it was used.


According to Giorgio Stabile, a professor of the history of science at La Sapienza University in Rome, the symbol was used as long ago as May 4, 1536 as an indication of a measure of weight or volume. This is certainly a fascinating aspect of impact analysis!

In addition to these articles - and so many others - there are links to Ray Tomlinson's account of how and why he invented what is now known as email, and why he chose the @ symbol for addresses. This information could be used for writing summaries and practicing relative clauses (see an explanation of my ideas in the posts Pioneer in AI has died, February 6, 2016 and in Billions in Change, February 10, 2016).

Saturday, March 5, 2016

Interaction through a computer screen

inFORM Shape Display
Recently a group of my students were working on projects about recent innovations in various fields. In looking for information, we came across a short news item from Brainiac, described as the "daily blog of the (Boston) Globe's Sunday Ideas section, covering news and delights from the worlds of art, science, literature, design, and more."

This article was MIT Makes it Possible to Reach Through Your Computer Screen, and the accompanying video was not only fascinating, but provided a lively discussion.

link: http://www.boston.com/bostonglobe/ideas/brainiac/2013/11/children_new_to.html

The article is fairly short (224 words) and written in a simple, conversational style. In any case, it has some useful collocations with verbs. For example:
  • reach through
  • get involved
  • physically interact with
  • track hand movements
  • dissimilar to
  • emerge from
  • model (their) work
  • come out of

The video (3:41) shows how the interaction works. A man sits in front of a computer screen and manipulates various objects on the other side of the screen. There is no talking, so it is easy to follow. There is, however, rather annoying music - but this can be turned off.

Students can create their own explanations of what is happening as or after they watch the video. More advanced students can transform the explanation for different audiences (e.g., technical, non-technical, pupils, customers, advertisement, etc.).

There is a link to the original source - an article from the Tangible Media Group at MIT's Media Lab (which also has the video). The title is "inFORM - Interacting with a Dynamic Shape Display."

link: http://tangible.media.mit.edu/project/inform/

This article is much more technical, but also short (254 words). It could be used to compare the language with the article from Brainiac - how the information is written in different levels of formality.

The article mentions a number of possible future uses of such technology:
  • Urban planners and architects can view 3D designs physically and better understand, share and discuss their designs.
  • 3D modelers and designers can prototype their 3D designs physically without 3D printing (at a low resolution).
  • Cross sections through volumetric data such as medical imaging CT scans can be viewed in 3D physically and interacted with.
  • Medical or surgical simulations might eventually be possible.

Students can not only discuss the implications of these developments, but can also brainstorm further ideas of their own.


Again, the language in the text offers useful collocations as well as a variety of adverbs:

  • currently exploring a number of application domains
  • working on
  • view 3D designs physically
  • understand, share, discuss (their) designs
  • collaborating with
  • prototype (their) 3D designs physically
  • Finally
  • scans can be viewed in 3D physically
  • interacted wtih
  • explore medical or surgical simulations
  • intrigued by
  • remotely manipulating objects
  • primarily focused on
  • emphasis on
  • dynamically changing
  • facilitate by providing
  • restrict by guiding
  • manipulate by actuating

All in all, this certainly looks like a fascinating development, and my students enjoyed the video and the ideas it inspired.