After participating in the neural network hype in the 1980s (I spent 1 year on a DARPA committee for NN tools, and was the original author of the SAIC ANsim NN software product) I found Hawkin's book to be light technically, but I really enjoyed reading it.
His work might have been inspired by Kohonen's classic Springer-Verlag book "Self-Organization and Associative Memory".
I downloaded their software last night but have had little time doing anything but building and running two examples. When I get 20 hours to really kick the tires, I will blog about it on my AI blog http://artificial-intelligence-theory.blogspot.com /
I am hoping that NTA will really simulate temporal memory and spacial invariance that the neocortex apparently has.
A little off topic, but I love the way they package the NTA software: most of the low level code is C++, that builds into sharable libraries loaded and used in a Python wrapper. Neat stuff. The free license is only for non-commercial use, BTW.
In the 1980s I believed that "strong AI" was forthcoming, but now I have my doubts that is reflected in the difference of tone from the first Springer-Verlag AI book that I wrote to my current skepticism. One of my real passions has for decades been natural language processing (NLP) but even for that I am a convert to statistical NLP using either word frequencies or Markov models instead of older theories like conceptual dependency theory that tried to get closer to semantics.
Just a gut feeling but I don't think that we will develop real general purpose AIs without some type of hardware breakthrough like quantum computers.
Flex 2 applications look great although for now I slightly prefer the approach of OpenLazlo because OpenLazlo supports both Flash and DHTML on the back end.
Check out gliffy.com for just how good OpenLazlo (with Flash backend) applications can look, with good functionality.
Adobe's upcoming Apollo will probably build on Flex 2 (not sure off hand) and promises the ability to have one code base for both web and desktop applications.
I have been writing open source stuff for over a decade, and I will continue with occasional GPL projects. That said, without corporate sponsorship it is rough financially to spend too much time on open source projects. I am in the process of switching to a plan where for my three current non-consulting development projects (text mining, visualization, and document web portal) I always have a source code drop available with a free for non-commercial use license. Commercial licenses are available for a small (tiny compared to the development costs) fee. This seems like a reasonable approach to me, and potential commercial customers have access to full source code before they commit to spending money. Wil I sometimes get ripped off? Sure, but I prefer to take my chances and trust people.
I don't think that any company should use any important infrastructure software that they do not have source for. Open source like Linux, OpenOffice.org, Apache, etc. are best, but for some more niche infrastructure components that are not commercially sponsored, an approach like the one I am starting to use make sense: consumers are protected by having source code, and developers of niche projects have some chance of making money to support future development.
As a Mac user, I am used to paying a little more:-)
Seriously, the.Mac service for $99/year is a good deal: 1/2 gig storage, nice integration with OS X and Backup, and apparently fairly unlimited bandwidth (I use.Mac as a mirror for some of my downloads).
I think that the two licenses are for very different purposes. Personally, I use the GPL for my open source projects and I use CC for my open content writing projects. Sharing software and sharing writing are different, at least for me, so I prefer different licenses.
I want people to be able to modify my software and reuse it as they need to (I also offer more commercial friendly licensing when people ask for it). On the other hand, I don't want people to be able to modify my free web books (except for emailing me corrections and suggestions) so I choose the appropriate CC license for that.
I use CC licenses for my Java AI and Common Lisp primer "free web books" and was the 'featured commoner' a few years ago.
I am pleased that CC is not standing still on licenses. Although I have written 14 published books, there are a few strong reasons why I am transitioning to CC open content authoring; the the primary reason is that I tend to be interested in niche technical areas and conventional publishers in the past have pressured me to tailor my works to a larger market. I am in the slow process of "dual publishing" my CC licensed content: free PD downloads and lulu.com instant print books for a fee for the occasional reader who wants a physical book.
My original motive for doing CC open content was simply that I got tired of having teachers, etc. ask if they good copy a chapter or two of my published books for their students - and my having to turn down their requests because my publishers own my material. Other reasons for CC based open content are a wider readership and thus more frequent interesting connections with my readers.
Really, the only advantage of using publishers is making money:-)
I just built the Ruby 2.0 svn trunk under OS X (as per directions on the web, I used an application suffice of -yarv to not interfere wih my production Ruby build).
My irb-yarv will not work (for me), but I have tried running ruby-yarv against the examples for my in-progress "Ruby AI" book and did not see any problems.
Yarv will be cool! Especially when it supports Rails.
I would use Ruby more if the runtime performance were better. For speed and dynamic language niceness I use Franz Common Lisp for most of the work for my current customer (AI medical applicaion) and I still find myself using Java a lot for serverside deployment goodness:-)
I would have prefered Ruby, BTW, because it is easy to train people in Ruby development while Common Lisp is a harder learning curve.
Once Yarv is very stable (maybe it is already?), then I will use Ruby for a larger percent of my new projects.
Ruby on Rails is certainly a giant step forward in increasing programmer productivity. The new REST support is great, but I must say that the web based SOAP debugging interface made using SOAP with Rails painless.
A little off topic, but for my customer I integrate Rails with back end Common Lisp code - so far just using home grown REST support at both ends. Eventually, I would like to get the time to generate compiled Lisp libraries directly linked to Ruby when everything is runnng on a single server.
BTW, I created a simply Ruby and Ruby Rails news clipping web site, mostly for my own use, but it might be useful for you: http://rubyplanet.net/
I do almost all of my consulting work remotely since I live in the mountains in Arizona, 2.5 hours from the Phoenix airport. I do sometimes get to meet my customers face to face and from then on, work is a little smoother - the face time effect is real.
However, there are real cost and performance advantages to telecommuting: commute/travel time is freed up, long periods of quiet time for working on more difficult problems, energy savings, savings in office space, etc.
The question is, will your managers balance the benefits and draw backs for telecommuting in a fair way? Good luck:-)
Re:I use Common Lisp because of its 'white hot' sp
on
Lisp and Ruby
·
· Score: 1
Right you are - sometimes when very new objects are GCed, they can be collected faster than malloc'ed memory blocks. Objects that are in older generations can take much longer to GC -- I will update the book text -- thanks.
Re:I use Common Lisp because of its 'white hot' sp
on
Lisp and Ruby
·
· Score: 1
The Ruby AI book is still a work in progress, but I decided to finish the second edition of the Lisp book first because it is a smaller effort. I have also done about 60% of my professional programming in Common Lisp in the last year, so I have Lisp on my mind. I use Ruby about 25% of the time now, and Java for most of the rest.
Ruby does rock!
Re:I use Common Lisp because of its 'white hot' sp
on
Lisp and Ruby
·
· Score: 1
BTW, I started work on a second edition of "Loving Lisp, or the Savvy Programmer's Secret Weapon" two weeks ago - if you download the PDF for the current edition and like it, then check back in a couple of months.
I use Common Lisp because of its 'white hot' speed
on
Lisp and Ruby
·
· Score: 1
I love programming in Ruby and use it a lot for both small utilities and some Rails web apps.
The problem is that Ruby has very poor runtime performance. As a result, I often use Common Lisp (Franz for commercial development, but also LispWorks, and SBCL). What kind of runtime performance will this proposed Lisp (it is all just talk for now) have?
Common Lisp is a great language. Several years ago, with some great input from the Lisp community, I wrote a free 50 page tutorial "Loving Lisp, or the Savvy Programmer's Secret Weapon" that is a free download from the open content page on my web site.
As a USA taxpayer, I believe that this is an example of what our government should be doing.
Before the politically motivated "war on terror", I remember seeing news articles about our FBI working with foreign governments to break up foreign hacking rings. Since 911 I don't recall hearing about this anymore.
Our NSA has in the past also donated Linux security enhancements. Excellent! Protecting our national infrastructure.
A little off topic, but this issue has me fairly angry: our government should spend money on just a few things (education, critical infrastructure, defending our country from direct attacks - not Bush's radical foreign adventures, and a minimal social security style safety net). Huge tax breaks for oil companies, attacking foreign countries for the financial benefit of a few corporation, etc. just does not cut it as good government.
The cost of Bush nuking Iran (if that happens) will end up ruining our economy and our way of life - not to mention making us a pariah state.
I will read the paper when I get the proceedings for the International Joint Conference for Artificial Intelligence. From the article, this seems like a statistical natural language processing application: the examples looked like they collect statistics of associations for both single word and short word sequences.
BTW, associating, clustering, etc. documents using single word statistics is computationally cheap and easy - it is also associating short word sequences that makes this a difficult problem.
I may be missing something, but isn't searching text (all types of documents) much more important for **real work** that searching pictures and video?
On the other hand, I believe in the utility of next generation text search that clusters documents and allows search for words by word sense (search for "bank" in the sense noun, financial institution - and not return results for "by the bank of the river", "bank the airplane to he right", etc.). Also, support better search within search, etc. I am working on these technologies for my little http://knowledgebooks.com/ business.
I do get high value from online videos of technical talks and really enjoy listening or watching keynote talks, etc. However, I know what sites to go to for technical talk videos. If I need to search for photos (seldom), Flickr and Picasa/Google work fine. Fun sites like YouTube.com are easily searchable. Actually, a little OT, but: YouTube.com is useful in a practical sense: a good source of news clips covering wider points of view not usually seen on news media owned by our corporate overlords/masters.
Sure, provide Windows but make the Windows partition 50% of the disk.
Then it is easier to install Linux for dual boot. Dell spends very little on OEM Windows licenses so the cost to me of having Windows is very small. However, getting a PC with half the disk unused makes life simpler.
I simply forward my GMail to my local POP3 account and get a local copy using ThunderBird. Once a week, it takes me about 10 minutes to delete stuff that I don't want permanently archived. When I backup my stuff to DVD-Rs, I also include my ThunderBird inbox.
Google relies on replication of data for backup - statistically OK, but for a free service I don't mind also doing my own backups.
I have spent a fair amount of time playing with 2 versions GWT and one of the Eclipse GUI designer plugins: good stuff for sure.
Problem is that it is Java centric. Since I don't need to do too much AJAX, I decided that it would be easier for me to learn how to use the Dojo Javascript library. For me, this works for my Common Lisp and Java based web apps. Dojo also works well with Rails, but Rail's built in AJAX support is so good I have (so far) just used that.
The press in the US (and other western countries) is failing in its basic obligation of questioning our "elected" leadership. And it is not enough to ask the occasional tough question - when Bush (for example) rambles on about something besides answering the question, re-state the question and push back.
The "owners" control the news media and control what people hear and see. Perfect setup, from their point of view.
This whole "give up freedoms to protect against terrorism" is bullshit, but still effective for controlling the masses. The reason that this is working is the same reason that the Nazis in Germany gained control: the economy.
On at least a subconscious level, people in the US (and some other industrialized countries) understand that other parts of the world are catching up and that their material standard of living may soon take a hit. I think that playing to this fear is what makes this "give up freedoms to protect against terrorism" thing work, but I think that it is fear of the future economy, not of the occasional random terrorist attack.
Any solution? Well I think so: if enough people turn away from materialism and get back to the basic joys in life: family, friends, sports, education, job satisfaction, etc. How many people do you know who have bought a new car because their (proverbial) brother in law just got one, when their old car ran fine - just an example.
Instead of BUY, BUY, BUY, how about LIVE, LIVE, LIVE. Readjust priorities and learn to live in a world that may not live up to media-driven materialistic expectations. And, push back hard against elected representatives who want us to give up our freedoms.
I thought that it was very cool, but I wonder how confusing the Squeak "playground" might be without some hands on training/tutorial.
I only spent 20 minutes running the image, but one thing that I did not notice was region-specific documentation for water cleanliness training, etc. I thought that these would be customized for each country/region. Does anyone know about this?
I was telling a lot of non-nerd friends about this project at a big Thanksgiving party yesterday, and not only did everyone really like the idea but no one seemed to think that this would seriously take away money from food, fresh water, etc. programs.
Slashdot Mirror
After participating in the neural network hype in the 1980s (I spent 1 year on a DARPA committee for NN tools, and was the original author of the SAIC ANsim NN software product) I found Hawkin's book to be light technically, but I really enjoyed reading it.
m /
His work might have been inspired by Kohonen's classic Springer-Verlag book "Self-Organization and Associative Memory".
I downloaded their software last night but have had little time doing anything but building and running two examples. When I get 20 hours to really kick the tires, I will blog about it on my AI blog http://artificial-intelligence-theory.blogspot.co
I am hoping that NTA will really simulate temporal memory and spacial invariance that the neocortex apparently has.
A little off topic, but I love the way they package the NTA software: most of the low level code is C++, that builds into sharable libraries loaded and used in a Python wrapper. Neat stuff. The free license is only for non-commercial use, BTW.
In the 1980s I believed that "strong AI" was forthcoming, but now I have my doubts that is reflected in the difference of tone from the first Springer-Verlag AI book that I wrote to my current skepticism. One of my real passions has for decades been natural language processing (NLP) but even for that I am a convert to statistical NLP using either word frequencies or Markov models instead of older theories like conceptual dependency theory that tried to get closer to semantics.
Just a gut feeling but I don't think that we will develop real general purpose AIs without some type of hardware breakthrough like quantum computers.
Flex 2 applications look great although for now I slightly prefer the approach of OpenLazlo because OpenLazlo supports both Flash and DHTML on the back end.
Check out gliffy.com for just how good OpenLazlo (with Flash backend) applications can look, with good functionality.
Adobe's upcoming Apollo will probably build on Flex 2 (not sure off hand) and promises the ability to have one code base for both web and desktop applications.
I have been writing open source stuff for over a decade, and I will continue with occasional GPL projects. That said, without corporate sponsorship it is rough financially to spend too much time on open source projects. I am in the process of switching to a plan where for my three current non-consulting development projects (text mining, visualization, and document web portal) I always have a source code drop available with a free for non-commercial use license. Commercial licenses are available for a small (tiny compared to the development costs) fee. This seems like a reasonable approach to me, and potential commercial customers have access to full source code before they commit to spending money. Wil I sometimes get ripped off? Sure, but I prefer to take my chances and trust people.
I don't think that any company should use any important infrastructure software that they do not have source for. Open source like Linux, OpenOffice.org, Apache, etc. are best, but for some more niche infrastructure components that are not commercially sponsored, an approach like the one I am starting to use make sense: consumers are protected by having source code, and developers of niche projects have some chance of making money to support future development.
As a Mac user, I am used to paying a little more :-)
.Mac service for $99/year is a good deal: 1/2 gig storage, nice integration with OS X and Backup, and apparently fairly unlimited bandwidth (I use .Mac as a mirror for some of my downloads).
Seriously, the
a typo: I meant: "free PDF downloads"
I think that the two licenses are for very different purposes. Personally, I use the GPL for my open source projects and I use CC for my open content writing projects. Sharing software and sharing writing are different, at least for me, so I prefer different licenses.
:-)
I want people to be able to modify my software and reuse it as they need to (I also offer more commercial friendly licensing when people ask for it). On the other hand, I don't want people to be able to modify my free web books (except for emailing me corrections and suggestions) so I choose the appropriate CC license for that.
I am happy with both licenses
I use CC licenses for my Java AI and Common Lisp primer "free web books" and was the 'featured commoner' a few years ago.
:-)
I am pleased that CC is not standing still on licenses. Although I have written 14 published books, there are a few strong reasons why I am transitioning to CC open content authoring; the the primary reason is that I tend to be interested in niche technical areas and conventional publishers in the past have pressured me to tailor my works to a larger market. I am in the slow process of "dual publishing" my CC licensed content: free PD downloads and lulu.com instant print books for a fee for the occasional reader who wants a physical book.
My original motive for doing CC open content was simply that I got tired of having teachers, etc. ask if they good copy a chapter or two of my published books for their students - and my having to turn down their requests because my publishers own my material. Other reasons for CC based open content are a wider readership and thus more frequent interesting connections with my readers.
Really, the only advantage of using publishers is making money
I just built the Ruby 2.0 svn trunk under OS X (as per directions on the web, I used an application suffice of -yarv to not interfere wih my production Ruby build).
My irb-yarv will not work (for me), but I have tried running ruby-yarv against the examples for my in-progress "Ruby AI" book and did not see any problems.
Yarv will be cool! Especially when it supports Rails.
I would use Ruby more if the runtime performance were better. For speed and dynamic language niceness I use Franz Common Lisp for most of the work for my current customer (AI medical applicaion) and I still find myself using Java a lot for serverside deployment goodness :-)
I would have prefered Ruby, BTW, because it is easy to train people in Ruby development while Common Lisp is a harder learning curve.
Once Yarv is very stable (maybe it is already?), then I will use Ruby for a larger percent of my new projects.
Thanks to the Rails developers!
Ruby on Rails is certainly a giant step forward in increasing programmer productivity. The new REST support is great, but I must say that the web based SOAP debugging interface made using SOAP with Rails painless.
A little off topic, but for my customer I integrate Rails with back end Common Lisp code - so far just using home grown REST support at both ends. Eventually, I would like to get the time to generate compiled Lisp libraries directly linked to Ruby when everything is runnng on a single server.
BTW, I created a simply Ruby and Ruby Rails news clipping web site, mostly for my own use, but it might be useful for you: http://rubyplanet.net/
I do almost all of my consulting work remotely since I live in the mountains in Arizona, 2.5 hours from the Phoenix airport. I do sometimes get to meet my customers face to face and from then on, work is a little smoother - the face time effect is real.
:-)
However, there are real cost and performance advantages to telecommuting: commute/travel time is freed up, long periods of quiet time for working on more difficult problems, energy savings, savings in office space, etc.
The question is, will your managers balance the benefits and draw backs for telecommuting in a fair way? Good luck
Right you are - sometimes when very new objects are GCed, they can be collected faster than malloc'ed memory blocks. Objects that are in older generations can take much longer to GC -- I will update the book text -- thanks.
The Ruby AI book is still a work in progress, but I decided to finish the second edition of the Lisp book first because it is a smaller effort. I have also done about 60% of my professional programming in Common Lisp in the last year, so I have Lisp on my mind. I use Ruby about 25% of the time now, and Java for most of the rest.
Ruby does rock!
BTW, I started work on a second edition of "Loving Lisp, or the Savvy Programmer's Secret Weapon" two weeks ago - if you download the PDF for the current edition and like it, then check back in a couple of months.
I love programming in Ruby and use it a lot for both small utilities and some Rails web apps.
The problem is that Ruby has very poor runtime performance. As a result, I often use Common Lisp (Franz for commercial development, but also LispWorks, and SBCL). What kind of runtime performance will this proposed Lisp (it is all just talk for now) have?
Common Lisp is a great language. Several years ago, with some great input from the Lisp community, I wrote a free 50 page tutorial "Loving Lisp, or the Savvy Programmer's Secret Weapon" that is a free download from the open content page on my web site.
As a USA taxpayer, I believe that this is an example of what our government should be doing.
Before the politically motivated "war on terror", I remember seeing news articles about our FBI working with foreign governments to break up foreign hacking rings. Since 911 I don't recall hearing about this anymore.
Our NSA has in the past also donated Linux security enhancements. Excellent! Protecting our national infrastructure.
A little off topic, but this issue has me fairly angry: our government should spend money on just a few things (education, critical infrastructure, defending our country from direct attacks - not Bush's radical foreign adventures, and a minimal social security style safety net). Huge tax breaks for oil companies, attacking foreign countries for the financial benefit of a few corporation, etc. just does not cut it as good government.
The cost of Bush nuking Iran (if that happens) will end up ruining our economy and our way of life - not to mention making us a pariah state.
I will read the paper when I get the proceedings for the International Joint Conference for Artificial Intelligence. From the article, this seems like a statistical natural language processing application: the examples looked like they collect statistics of associations for both single word and short word sequences.
BTW, associating, clustering, etc. documents using single word statistics is computationally cheap and easy - it is also associating short word sequences that makes this a difficult problem.
I may be missing something, but isn't searching text (all types of documents) much more important for **real work** that searching pictures and video?
On the other hand, I believe in the utility of next generation text search that clusters documents and allows search for words by word sense (search for "bank" in the sense noun, financial institution - and not return results for "by the bank of the river", "bank the airplane to he right", etc.). Also, support better search within search, etc. I am working on these technologies for my little http://knowledgebooks.com/ business.
I do get high value from online videos of technical talks and really enjoy listening or watching keynote talks, etc. However, I know what sites to go to for technical talk videos. If I need to search for photos (seldom), Flickr and Picasa/Google work fine. Fun sites like YouTube.com are easily searchable. Actually, a little OT, but: YouTube.com is useful in a practical sense: a good source of news clips covering wider points of view not usually seen on news media owned by our corporate overlords/masters.
Sure, provide Windows but make the Windows partition 50% of the disk.
Then it is easier to install Linux for dual boot. Dell spends very little on OEM Windows licenses so the cost to me of having Windows is very small. However, getting a PC with half the disk unused makes life simpler.
I simply forward my GMail to my local POP3 account and get a local copy using ThunderBird. Once a week, it takes me about 10 minutes to delete stuff that I don't want permanently archived. When I backup my stuff to DVD-Rs, I also include my ThunderBird inbox.
Google relies on replication of data for backup - statistically OK, but for a free service I don't mind also doing my own backups.
I have spent a fair amount of time playing with 2 versions GWT and one of the Eclipse GUI designer plugins: good stuff for sure.
Problem is that it is Java centric. Since I don't need to do too much AJAX, I decided that it would be easier for me to learn how to use the Dojo Javascript library. For me, this works for my Common Lisp and Java based web apps. Dojo also works well with Rails, but Rail's built in AJAX support is so good I have (so far) just used that.
Right on - if I had mod points, you would get them.
The press in the US (and other western countries) is failing in its basic obligation of questioning our "elected" leadership. And it is not enough to ask the occasional tough question - when Bush (for example) rambles on about something besides answering the question, re-state the question and push back.
The "owners" control the news media and control what people hear and see. Perfect setup, from their point of view.
This whole "give up freedoms to protect against terrorism" is bullshit, but still effective for controlling the masses. The reason that this is working is the same reason that the Nazis in Germany gained control: the economy.
On at least a subconscious level, people in the US (and some other industrialized countries) understand that other parts of the world are catching up and that their material standard of living may soon take a hit. I think that playing to this fear is what makes this "give up freedoms to protect against terrorism" thing work, but I think that it is fear of the future economy, not of the occasional random terrorist attack.
Any solution? Well I think so: if enough people turn away from materialism and get back to the basic joys in life: family, friends, sports, education, job satisfaction, etc. How many people do you know who have bought a new car because their (proverbial) brother in law just got one, when their old car ran fine - just an example.
Instead of BUY, BUY, BUY, how about LIVE, LIVE, LIVE. Readjust priorities and learn to live in a world that may not live up to media-driven materialistic expectations. And, push back hard against elected representatives who want us to give up our freedoms.
I thought that it was very cool, but I wonder how confusing the Squeak "playground" might be without some hands on training/tutorial.
I only spent 20 minutes running the image, but one thing that I did not notice was region-specific documentation for water cleanliness training, etc. I thought that these would be customized for each country/region. Does anyone know about this?
I was telling a lot of non-nerd friends about this project at a big Thanksgiving party yesterday, and not only did everyone really like the idea but no one seemed to think that this would seriously take away money from food, fresh water, etc. programs.