|
I was not terribly interested in ML, Deep Learning and/or AI until I read the book, The Creativity Code[^] by Marcus du Sautoy (mathematician)
That was a great introduction to some amazing things that are happening in the field and it got me interested in what AI is really able to do and how it is done.
Now, I'm reading a hands-on book that includes code and teaches Deep Learning concepts and code.
Here's a very good summary of which is what from the book:
Deep learning is a subset of machine learning, which is a field dedicated to the study and development of machines that can learn (sometimes with the goal of eventually attaining general artificial intelligence).
In industry, deep learning is used to solve practical tasks in a variety of fields such as computer vision (image), natural language processing (text), and automatic speech recognition (audio). In short, deep learning is a subset of methods in the machine learning toolbox, primarily using artificial neural networks, which are a class of algorithm loosely inspired by the human brain. Here's a very nice graphic from the book [^] (Venn diagram type of thing) which shows how these three terms are related.
If you've felt apathetic about AI (as I have) and weren't sure how to dip your toe into the subject, I suggest you start with the first book and continue on into Grokking Deep Learning. 
|
|
|
|
|
I once read an interesting blog that AI doesn't exist and is just a marketing trick to sell ML.
The term AI in itself is problematic since no one really knows what intelligence is or how to measure it.
"Everyone is a genius. But if you judge a fish by its ability to climb a tree, it will live its whole life believing that it is stupid."
So... Is a computer intelligent if it can swim, climb a tree, do basic math, survive in the jungle...?
We have self driving cars and computers that can "learn" games such as Super Mario, but those are really just ML (image recognition and trial and error until something works).
How can you create something that you don't understand and how do you know if you succeeded if you can't measure it?
Too bad I can't find that blog anymore, you'd find it interesting.
I think it was a reaction to people such as Elon Musk saying AI is dangerous.
|
|
|
|
|
Sander Rossel wrote: image recognition and trial and error until something works).
That's the thing though - that's how evolutionary processes happen naturally, more or less (i'm simplifying but it holds)
ML may not be AI, but we're on the right path, roughly. Such a thing needs to be grown "organically", I'm as certain of that as I am of nearly anything.
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
Next to trial and error we can reason about things.
And we can do completely unreasonable things as well.
Our heads do a lot more than ML and trial and error.
Something computers, so far, cannot.
My real point, however, is that we can't have AI if we don't know what intelligence is to begin with.
There's a computer that can score 150 on an IQ test, but that doesn't make it intelligent in the least even though such a feat would mean brilliance for people.
And even then, these brilliant people wouldn't survive a few days in the jungle while "dumb" jungle tribes have lived there for centuries.
|
|
|
|
|
Sander Rossel wrote: I once read an interesting blog that AI doesn't exist and is just a marketing trick to sell ML.
The term AI in itself is problematic since no one really knows what intelligence is or how to measure it.
Interesting ideas. Thanks for posting.
|
|
|
|
|
I think intelligence is not too difficult to define. IQ tests are fairly good at determining intelligence with spacial, verbal and other types of intelligence.
The word "intelligence" has tended to stray at times from what IQ tests were created for, such as "emotional intelligence".
The thing that is difficult to define is consciousness.
“That which can be asserted without evidence, can be dismissed without evidence.”
― Christopher Hitchens
|
|
|
|
|
GuyThiebaut wrote: IQ tests are fairly good at determining intelligence with spacial, verbal and other types of intelligence.
IQ tests are very limited and can even be trained.
And as you say, they measure only certain types of intelligence.
That said, if you've never been to school you'd fail such a test miserably, but that does't mean you're not intelligent just that you're not learned.
Anyway, researchers already created a program that made a computer score 150 on an IQ test and guess what, that computer was nowhere near intelligent.
Not intelligent in that you can have even the most basic conversation with it.
It would probably overflow at "Nice weather, eh?"
It also couldn't walk the stairs or play Super Mario (we have other computers doing ML to learn Super Mario).
Wikipedia says the following: "Intelligence has been defined in many ways, including: the capacity for logic, understanding, self-awareness, learning, emotional knowledge, reasoning, planning, creativity, critical thinking, and problem solving."
That's a lot more than an IQ test covers and most of it is completely impossible for computers at this time (if ever).
But that was exactly my point, there is no clear definition of intelligence and it's certainly not "being able to fill out an IQ test."
|
|
|
|
|
Sander Rossel wrote: but those are really just ML (image recognition and trial and error until something works).
Not only Machine Learning...
Human learning is exactly the same... how do you think that babys learn?
Not even the "genius" are born knowing everything or always do all perfectly at the first try. They do test and error too, just learn faster from it than others.
Without test-and-error would be no new creations in most of the fields.
M.D.V. 
If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about?
Help me to understand what I'm saying, and I'll explain it better to you
Rating helpful answers is nice, but saying thanks can be even nicer.
|
|
|
|
|
As I was sitting in my car at lunch yesterday and watching a small spider fix a web that a passing wasp had just broken I did think "I doubt it's mother taught it how to build such an amazing web".
There are some innate abilities, which while they improve with learning, do seem to be there from evolution.
“That which can be asserted without evidence, can be dismissed without evidence.”
― Christopher Hitchens
|
|
|
|
|
I can read your comment and interpret them as not only words in a random order.
I know them to have meaning and I can form an opinion about it.
In fact, we can create new words or bend the rules of language and we still know what we mean.
"I dictionaried the hell out of that word."
A computer can tell me "dictionaried" is not a word, but a human understands I looked up the word in a dictionary and quite aggressively too.
You've never ever seen or heard the word "dictionaried", so how is it then that you still understand?
The first time I played Super Mario I knew exactly what to do... Get guy from left to right while evading obstacles.
I had never even seen a game before.
It takes a computer hours to learn this and it probably doesn't even understand what it does.
It's a bit more than ML (or HL) methinks.
|
|
|
|
|
|
Hmmm,
For only $5,695.00 this Cheezburger can be yours. Looks like it would make a great cat food bowl. 
Best Wishes,
-David Delaune
Scientiæ de conservata veritate.
|
|
|
|
|
Thanks, bought it 
|
|
|
|
|
Isn't it a bit cheesy?
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
|
I wear button-down shirts, as a rule. The only T-shirts I own are those given out at company events.
(I still have one handed out by Microsoft when .NET 1.1 was announced)
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
I don't visit enough events 
I used to have a good one by DevExpress (what I actually used at the time), it said "eat sleep code" including symbols 
Some kid in Africa is now probably wearing it.
|
|
|
|
|
|
That's why I hunt down clean shirts only, and buying them by dozen...
"The only place where Success comes before Work is in the dictionary." Vidal Sassoon, 1928 - 2012
|
|
|
|
|
|
About fourtyfive years ago, I went though much of the North-Eastern US on a camping trip with my very hot looking Irish (decent) girlfriend.
After forty-five years, she's the grandmother of my grandchildren - we're still together, still lovers, and Vermont is still beautiful.
When they come your way, never miss a chance to savor those "life is beautiful" moments.
[edit]
2019.09.10 10:25 AM
I was just thinking about it - how busy we get and how easy it is to put off doing things that invite these moments . . . and now I wonder "Why?".
[/edit]
| Ravings en masse^ |
|---|
| "The difference between genius and stupidity is that genius has its limits." - Albert Einstein | | "If you are searching for perfection in others, then you seek disappointment. If you seek perfection in yourself, then you will find failure." - Balboos HaGadol Mar 2010 |
modified 10-Sep-19 10:28am.
|
|
|
|
|
W∴ Balboos wrote: very hot looking Irish (decent) girlfriend.
LOL!! 
![Rose | [Rose]](https://codeproject.global.ssl.fastly.net/script/Forums/Images/rose.gif)
I'd rather be phishing!
|
|
|
|
|
Hmmm, good tip 
|
|
|
|
|
Congratulations,
According to the Census Bureau only 6% of Americans pass the 50th (Golden) wedding anniversary. My grandparents on my mothers side (born in 1906 and 1908 respectively) reached their Diamond anniversary.
Best Wishes,
-David Delaune
Scientiæ de conservata veritate.
|
|
|
|
|
Congrats, not many people these days sticking it out.
Technician
1. A person that fixes stuff you can't.
2. One who does precision guesswork based on unreliable data provided by those of questionable knowledge.
JaxCoder.com
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
