Sunday, December 19, 2021

Is achieving human-level intelligence (Humint) still relevant for AI?

I have always been drawn to topics related to intelligence in a machine or computer. More so after reading The Emperor's New Mind: Concerning Computers, Minds and The Laws of Physics, a 1989 book by the mathematical physicist Sir Roger Penrose. a British mathematician, mathematical physicist, philosopher of science, and Nobel Laureate in Physics.

IMHO we still have a long way to go to have AI in a machine or computer that is at the level of human-level intelligence. Or perhaps the 'thinking' in a machine is so different than in a human that only certain aspects of the thinking in human intelligence are relevant for AI.

Deep-learning versus human intelligence

What does it mean to understand

Roger Penrose: Human thinking can never be emulated by a machine

   



Wednesday, October 29, 2014

Meteor 1.0 has finally arrived

The wait is over! Yeah...
The freaking awesome Meteor open source platform for building web and mobile apps in pure JavaScript finally reached 1.0.

https://www.meteor.com/


Thursday, May 16, 2013

Great Pacific Garbage Patch

The world needs to come together to tackle this growing problem. We owe it to ourselves and the next generation.

http://education.nationalgeographic.com/education/encyclopedia/great-pacific-garbage-patch/?ar_a=1

Friday, April 19, 2013

Hybrid and cross-pollination rule

In software development, just as in nature, hybrid and cross-pollination rule.

Friday, January 28, 2011

F# versus Scala

Both languanges IMHO provide similar functional programming features. And both are not pure functional programming language as they also support imperative and object-oriented style of programming.
Since I am not an expert with either programming language, I'm not qualified to provide detailed comparison of their features.
My expertise is more in OO software development, primarily in C# and prior to that in Java. I discovered functional programming quite by accident through a Microsoft pdc2008 presentation: An Introduction to Mcrosoft F#, by Luca Bolognese (back then the Principal Languages PM at Microsoft). That was my aha moment for functional programming, in particular F#. I have been hooked since and continue my exploration of this exciting landscape.
Coming from OO world when we often have to struggle with side effects, functional programming paradigm which avoids side effects of state and mutable data is a breadth of fresh air.
Given the plethora of functional programming languages out there, I would like to offer my two cents when the choice is between F# and Scala.
In my opinion, F# feels more functional than OO because its expression is closer to a mathematical expression. Scala on the other hand, feels more OO than functional.

For example, take this simple expression to increment an integer.
In F#:
let increment x = x + 1

In Scala:
def increment(x:Int) = x + 1

Due to type inference support in F#, there is no need to provide the type of the parameter x. In this case, the F# expression is more succinct.

Take another example, this time the expression involves Lambda expression.
In F#:
let list = List.map (fun i -> i + 1) [1;2;3]

In Scala:
val list = List(1,2,3).map( x => x + 1 )

In this case, the Scala expression is probably easier to understand than the F# one because it is more OO.

I personally prefer the approach that F# takes because the expression is more succinct. Although I must admit at times it takes some getting used to, especially when coming from OO background as I am.
But either language is a good choice to ease your way into functional programming from OO programming.

Thursday, April 22, 2010

F# has finally arrived

The long anticipated .NET 4.0, Visual studio 2010 and F# are finally here.

http://blogs.msdn.com/dsyme/archive/2010/04/12/f-2-0-released-as-part-of-visual-studio-2010.aspx

Wednesday, May 6, 2009