3
$\begingroup$

The term edutainment was introduced with the advent of the television and summarizes a corpus of books and TV-shows in which academic topics are presented with lower access barrier. What edutainment blogs have in common is, that the reader doesn't have to think very much but can consume the information as a background task for personal pastime.

In the domain of Artificial Intelligence, it's hard to identify a sort of media which is bridging the gap between expert audience and amateurs who have a need for easy-to-consume information. Only information about classical computer science for example programming languages like Python or operating systems like Windows 10 are teached in a pleasant way. The attempts from the past to explain neural networks and Game AI to a wider audience were not very successful. The existing blogs which are addressed to newbies can't be recommended anymore.

Let us take a look into a sad attempt to explain Artificial Intelligence with a consumer product, it's called Lego Mindstorms and many universities are using this toy together with books and challenges in the curriculum. But something is wrong with robotics as a vehicle to teach Artificial Intelligence. Because Mindstorms works always with a hands-on-mentality. The students are asked to unbox the toy, program it by themself and solve a line following problem by typing in Java code. For most students this is within their reach but it's the opposite from a laid back mentality, in which the student doesn't has to do anything but press only the play button and consumes what other have prepared in advance.

Why is it so hard to write edutainment information about Artificial Intelligence? Are some good example available in which complicated subjects are introduced with a low entry barrier?

$\endgroup$
  • $\begingroup$ Do you contemplate also "old fashioned" AI? This is an interesting point of view for communicating basic AI algorithms. $\endgroup$ – David Sep 16 at 18:49
  • $\begingroup$ Logic production systems like this one, are working on a symbolic level. They bring a system from an initial state into a goal state. In a toy example, the prisoner is captured in a room and has to find the way out. It's possible to invent a story around it and introduce basic AI principles in a colloquial way. $\endgroup$ – Manuel Rodriguez Sep 16 at 19:49
  • $\begingroup$ I'll take that as a yes and I will think about an appropriate answer. $\endgroup$ – David Sep 17 at 0:03
  • $\begingroup$ Question, although not a duplicate, is heavily related to ai.stackexchange.com/questions/13376/… $\endgroup$ – mcRobusta Sep 17 at 16:30
  • $\begingroup$ What level of detail/understanding do you expect to get from pressing the 'play button'? Knowing that will help to give an answer. $\endgroup$ – Dunk Sep 17 at 17:04
0
$\begingroup$

Here's the cinch. Current deep learning comes down to some very hard mathematics.

This isn't new: game theory, an earlier form of AI, is a potent mix of statistics, matrices, set theory and functional analysis that's tough to get your head around. Supervised learning with matrix multiplication, Jacobian values and eigenvectors need a solid grasp on AP-level mathematics to even think about approaching. It's difficult to introduce this gently. TensorFlow purposely treats its open source project like a black box in its documentation because what's going on under the hood is tricky to get to grasps with.

However, Towards Data Science is a great blog site to start. I'm self taught in AI (and I now work in the biz!) and this was a really great place to go. Places like StackOverflow are also indispensible because you can ask lower-level questions that are just assumed knowledge in things like academic papers.

As long as you remember it's all just maths in there, you should be able to get to grips with it. Definitely start off reading about statistics and appreciating the mathematics of it all- it'll make topics in the 'expert audience' much more approachable.

It's great to see you're enthusiastic about learning the 'why' instead of just how to run a TensorFlow model- I hope it helps!

$\endgroup$
  • 2
    $\begingroup$ Blogs like “towards data science” are a good example in how to create a boring edutainment media. On the first look the subject is explained more easily, but even readers of academic papers have problems to understand the written information. The normal audience doesn't like it. There most be a pathway available which teaches the AI subject more eloquent to the public. $\endgroup$ – Manuel Rodriguez Sep 16 at 20:09
  • $\begingroup$ @ManuelRodriguez There is- university. $\endgroup$ – mcRobusta Sep 16 at 21:38
1
$\begingroup$

Direct Answer to Your Question and What This Answer is About:--

People have certain expectations from what they consider to be video games.

People have certain expectations from what they consider to be training tools about artificial intelligence.

These two do not always align. I'll try to find games that meet both expectations.


Body:--

" ... In the domain of Artificial Intelligence, it's hard to identify a sort of media which is bridging the gap between expert audience and amateurs who have a need for easy-to-consume information. ...

... Why is it so hard to write edutainment information about Artificial Intelligence? Are some good example available in which complicated subjects are introduced with a low entry barrier? ... " ~ mcRobusta, Opening Poster

That is the area that my answer will focus on.

(https://en.wikipedia.org/wiki/Educational_entertainment)

The issue is that A.I. is a deeply complex field, involving hard mathematics and study. Most resources I could find were mid-tier difficulty and required some background in programming to get into. It is much easier to find video games that teach general programming than it is to find video games about artificial intelligence.

Another significant issue is that if you are effectively learning a skill, this is becomes a training tool rather than a video game that is recreational. Thus walking the line between video game and training tool is extremely tricky. People have certain expectations from what they consider to be video games.

Artificial intelligence is hard. Game design is hard. Combining these two is even harder because they are both hard fields. Complexity can multiply complexity.

(https://en.wikipedia.org/wiki/Educational_entertainment#Criticism)

I have attempted to mine tools and games that are both valid video games and valid training tools.


Relevant MOOCs (Absolute Beginner):--

Certain MooCs fill in this gap. Other MooCs cover advanced topics or mid-tier topics.

Take a gander through this link below:--

(https://www.coursera.org/learn/ai-for-everyone?)


Relevant Video Games (Mid Tier):--

Some video games that are related to A.I., aimed at children:--

  1. (https://www.sciencedaily.com/releases/2009/02/090203081622.htm)

This is not precisely about A.I., but I have decided to this MMORPG include due to its interesting nature:--

  1. (https://store.steampowered.com/app/464350/Screeps/)

WarriorJS is another relevant title that is an MMORPG:--

  1. (https://www.warriorjs.com/)

The link below is also worth noting because of its innovative nature and ability to apply A.I. to teach children via VR, but it does not teach about A.I. in its own right:--

(https://www.doubleyoukids.com/)


Relevant Toys (Mid Tier):--

The children are the future.

" ... More and more kids are learning to code at an early age, so what’s better than playing with coding robots that teach the basics of programming?

Robots are the type of toy that brings children together to play. Mums, dads, brothers and sisters can all have fun playing with cool coding toys.

Like many of the educational toys on the market, robots offer a fun and engaging way to get hands-on with learning. Interactive toys encourage children to get involved, solve problems, and make quick decisions, developing their skills through play. ... "

— "20 Best Coding Toys & Coding Robots for Kids." CodeMom.AI - Coding for Kids & Artificial Intelligence. Web. < http://www.codemom.ai/toys/coding-robots-for-kids/ >.

There is a small list of relevant toys, here:--

(https://www.codemom.ai/toys/coding-robots-for-kids/)


Relevant Side Notes and Miscellaneous:--

  • My background is largely in wargaming (or at least what is left of it), such as titles produced by Matrix Games, Paradox Interactive, or Relic Entertainment. I have written about Company of Heroes 2, elsewhere. I have worked in regular eSports (League of Legends, Dota 2; etc.) as well. I mention this so that it may give you a better understanding of where I am coming from and reveal any biases I have.

  • I am not a child psychologist or professional educator, at least at the time of writing.

  • I recommend researching intellectual giftedness, particularly in children: < https://en.wikipedia.org/wiki/Intellectual_giftedness >.

  • I am a strong supporter of the Finnish education system: < https://en.wikipedia.org/wiki/Education_in_Finland >. I recommend studying this system. This system has been largely lauded by many educational sources (see the Wiki link) as being successful. Talking about this paradigm may help people understand my approach better.


Sources, References, and Further Reading:--

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.