Common Sense A.I. might be coming sooner than you think
So, here's a photo of a bunch of books that just happen to be sitting on my work desk, looking all majestic and intellectual.
This is the sort of thing that a lot of people over the past few weeks have started to strategically place behind themselves for those all important conference calls, where other people will be looking and trying to eye up, "so what books do you have behind you?....oh, maybe I should read some of those and then I'll be as smart as you".
That could be why these books are not behind me, they are infront and to the side, I can see them but you cannot. What is behind me? hang on, let me turn around and look......a MASSIVE 3D printer with a Playstation 2 Time Crisis gun on the bed (absolutely no idea, maybe I found it in a cupboard and it was making it's way off the "games room", but I got distracted), some photo's of my MG F over the years taken by professional photographers as I was entering "event/shows" and a couple of fully working but not used Atari 520ST FM computers. Wow, not sure exactly what message people will take from that, but, y'know I like to keep them guessing.
Anyway, back to the books... I can actually confess that I have read "most" of them, quite a few are those dry academic / theoretical one's that professor type people like to read and as you know, I'm not really one of those, I like to get the general idea and then go make something with that knowledge and learn from making mistakes / errors (aren't you glad I don't do Nuclear stuff!)
As you can see, there are a lot of "proper" A.I. books there, and some random books that I bought way back in the early 2000's, such as, "The private life of the brain", "How to build a mind", "Cambrian Intelligence", "Life and how to make it", "Digital Biology" and "The next Common Sense".
Now, anyone, just reading those titles' may have wondered what was on "my" mind, back in those days. Well, I had a vision, I could "see" where things needed to go and I was trying to line up my own brain to be able to go in that direction. However, the technology was lacking. Yes, I was working for some very smart and advanced technology companies at the time, but, I was not meeting with like minded people. If I started talking about the world I could see and wanted to help build (maybe not be part of? 'cos, well, it's not my thing, so maybe this would be a Mars mission or something?), then people would just glaze over or pat me on the shoulder and tell me to "lay off the [insert whatever here]", I never really found "my kind".
I have been close though, I did meet up with the very nice people from Soul Machines - I still have the beautifully hand crafted blue business card of the boss blue-tacked to my wall, over there (not within web camera view). They were visionaries, they had access to the technology, they had money, but they had no real direction (at that time), I even pondered going to work for them, just to get close to the technology and the kool-aid. The moment passed.
I even ventured into trying to apply my vision into something real myself, but as we know, doing a very demanding job that takes up pretty much all my time, mostly has me working away 4-5 days a week, working evenings in hotels, I was never going to find the time to actually physically BUILD the robots and CODE the robots (hence the 3D printer). I even tried to convince my wife to install ROS onto her laptop so she could help me. She has no enthusiasm for the IT industry any more, she just knits, or sews, or does whatever she does, enjoys life playing with Libby, our rag-doll cat.
As you can see by this article, I did do some work with ROS:
http://tonyisageek.blogspot.com/2017/12/using-ros-to-control-gpio-pins-on.html
http://tonyisageek.blogspot.com/2017/12/m1ll13-robot-step-21.html
http://tonyisageek.blogspot.com/2018/03/m1ll13-robot-step-22.html
but that was back in 2017/2018! I've not had a chance to look into it since. Yes, I can put all sorts of reasons as to why. But ultimately, it was because it wasn't a high enough priority goal.
Last night, around 23:00, I decided to spend a few ££££ and refresh myself with ROS once more:
Okay, I'm not going to build a Terminator T100 (or whatever they are called), but I am going to teach myself the basics of the message queue mechanism that the ROS system is built upon. (I do remember that much from before), but this time I am going to be a little un-conventional, I'm going to be looking at putting the Robot Operating System into my Radio Controlled Boat and Tank!
Yes, due to the wonderful COVID-19 lockdown period of time, I have used the evenings and some weekend time (as I am actually AT home!), to build a small boating lake (well, 12-15ft x 8-10ft for the marina part), along with a 4ft wide "river" connecting down the garden to the proper pond thats about the same size as the marina, except it's about 5-6ft deep for the fish. I don't see it taking too long to finish that off, I do need to design a small motorised ramp, or, a Falkirk wheel of some sort - but that's a no-brainer, that's just digging stuff and getting pond liner worked out.
Then, I shall have ready my autonomous boats. Okay, so it'll just be an intelligent Train set, in theory, but I then want to take it beyond the boats just using sensors to do reactions. I want to then have the boats detect the sunshine, to then "decide" that they want to go for a drive out and about, rather than sitting on the shelf waiting for me to take them to the boating lake.
If you imagine, the boat(s) and tank(s) are just IoT devices on the Edge of the network... NOW, IT ALL STARTS TO MAKE SENSE!!! Ah, ha. a bit like back in 2016 when I was building "Smarter Buildings", this is the same sort of thing, but this time, I'm going to be capturing small packets of data from sensors, do some local processing and send larger datasets back to a central system for storage and then it can be analysed and smart stuff can be determined from the data. This smart stuff, can then be sent back to the device, so next time around, it can benefit from this rather than having to go back to the central system each time (a bit like a Cylon).
Then, I'll work on some other logic that calls out to the mysterious A.I. stuff (black box algorithms) that will allow the boats to make informed decisions, all on their own.
Then, of course, once I've got that perfected (it'll likely be around July/August 2020), I'll move onto the Tank(s). Well, that's going to be more fun...they can be protectors of the garden, they can patrol and learn about their environment. Okay, so basically, all of this is to eventually, get back to my in-famous desire to build a GardenBot! One that knows how to do gardening, learns about the growths and changes to the plants / flowers and behaves accordingly. Learns when too much water has been used, when to "thin out" the plants, when to add more water, when to cut out dead flowers, etc... but that'll be a 2021 goal target (well, it's been a goal for over a decade, but at least I'm doing something about it now!)
https://www.wired.com/story/watsons-creator-teach-ai-new-trick-common-sense/
This all ties in nicely with this article that was just published. The brains behind Watson (yeah, that thing that paid my bills for a few years), has decided to read some of the same books that I've had on the shelf and due to him not really needing any money, he can "play at R&D" to his hearts content. Am I jealous? sure. Do I want to call him a git, under my breathe, sure. Should I ping him an email and ask if we can work on something together? errr...maybe not just yet. Let me get my autonomous boats and tanks working and "thinking" first.
https://www.wired.com/story/watsons-creator-teach-ai-new-trick-common-sense/
This is the sort of thing that a lot of people over the past few weeks have started to strategically place behind themselves for those all important conference calls, where other people will be looking and trying to eye up, "so what books do you have behind you?....oh, maybe I should read some of those and then I'll be as smart as you".
That could be why these books are not behind me, they are infront and to the side, I can see them but you cannot. What is behind me? hang on, let me turn around and look......a MASSIVE 3D printer with a Playstation 2 Time Crisis gun on the bed (absolutely no idea, maybe I found it in a cupboard and it was making it's way off the "games room", but I got distracted), some photo's of my MG F over the years taken by professional photographers as I was entering "event/shows" and a couple of fully working but not used Atari 520ST FM computers. Wow, not sure exactly what message people will take from that, but, y'know I like to keep them guessing.
Anyway, back to the books... I can actually confess that I have read "most" of them, quite a few are those dry academic / theoretical one's that professor type people like to read and as you know, I'm not really one of those, I like to get the general idea and then go make something with that knowledge and learn from making mistakes / errors (aren't you glad I don't do Nuclear stuff!)
As you can see, there are a lot of "proper" A.I. books there, and some random books that I bought way back in the early 2000's, such as, "The private life of the brain", "How to build a mind", "Cambrian Intelligence", "Life and how to make it", "Digital Biology" and "The next Common Sense".
Now, anyone, just reading those titles' may have wondered what was on "my" mind, back in those days. Well, I had a vision, I could "see" where things needed to go and I was trying to line up my own brain to be able to go in that direction. However, the technology was lacking. Yes, I was working for some very smart and advanced technology companies at the time, but, I was not meeting with like minded people. If I started talking about the world I could see and wanted to help build (maybe not be part of? 'cos, well, it's not my thing, so maybe this would be a Mars mission or something?), then people would just glaze over or pat me on the shoulder and tell me to "lay off the [insert whatever here]", I never really found "my kind".
I have been close though, I did meet up with the very nice people from Soul Machines - I still have the beautifully hand crafted blue business card of the boss blue-tacked to my wall, over there (not within web camera view). They were visionaries, they had access to the technology, they had money, but they had no real direction (at that time), I even pondered going to work for them, just to get close to the technology and the kool-aid. The moment passed.
I even ventured into trying to apply my vision into something real myself, but as we know, doing a very demanding job that takes up pretty much all my time, mostly has me working away 4-5 days a week, working evenings in hotels, I was never going to find the time to actually physically BUILD the robots and CODE the robots (hence the 3D printer). I even tried to convince my wife to install ROS onto her laptop so she could help me. She has no enthusiasm for the IT industry any more, she just knits, or sews, or does whatever she does, enjoys life playing with Libby, our rag-doll cat.
As you can see by this article, I did do some work with ROS:
http://tonyisageek.blogspot.com/2017/12/using-ros-to-control-gpio-pins-on.html
http://tonyisageek.blogspot.com/2017/12/m1ll13-robot-step-21.html
http://tonyisageek.blogspot.com/2018/03/m1ll13-robot-step-22.html
but that was back in 2017/2018! I've not had a chance to look into it since. Yes, I can put all sorts of reasons as to why. But ultimately, it was because it wasn't a high enough priority goal.
Last night, around 23:00, I decided to spend a few ££££ and refresh myself with ROS once more:
Okay, I'm not going to build a Terminator T100 (or whatever they are called), but I am going to teach myself the basics of the message queue mechanism that the ROS system is built upon. (I do remember that much from before), but this time I am going to be a little un-conventional, I'm going to be looking at putting the Robot Operating System into my Radio Controlled Boat and Tank!
Yes, due to the wonderful COVID-19 lockdown period of time, I have used the evenings and some weekend time (as I am actually AT home!), to build a small boating lake (well, 12-15ft x 8-10ft for the marina part), along with a 4ft wide "river" connecting down the garden to the proper pond thats about the same size as the marina, except it's about 5-6ft deep for the fish. I don't see it taking too long to finish that off, I do need to design a small motorised ramp, or, a Falkirk wheel of some sort - but that's a no-brainer, that's just digging stuff and getting pond liner worked out.
Then, I shall have ready my autonomous boats. Okay, so it'll just be an intelligent Train set, in theory, but I then want to take it beyond the boats just using sensors to do reactions. I want to then have the boats detect the sunshine, to then "decide" that they want to go for a drive out and about, rather than sitting on the shelf waiting for me to take them to the boating lake.
If you imagine, the boat(s) and tank(s) are just IoT devices on the Edge of the network... NOW, IT ALL STARTS TO MAKE SENSE!!! Ah, ha. a bit like back in 2016 when I was building "Smarter Buildings", this is the same sort of thing, but this time, I'm going to be capturing small packets of data from sensors, do some local processing and send larger datasets back to a central system for storage and then it can be analysed and smart stuff can be determined from the data. This smart stuff, can then be sent back to the device, so next time around, it can benefit from this rather than having to go back to the central system each time (a bit like a Cylon).
Then, I'll work on some other logic that calls out to the mysterious A.I. stuff (black box algorithms) that will allow the boats to make informed decisions, all on their own.
Then, of course, once I've got that perfected (it'll likely be around July/August 2020), I'll move onto the Tank(s). Well, that's going to be more fun...they can be protectors of the garden, they can patrol and learn about their environment. Okay, so basically, all of this is to eventually, get back to my in-famous desire to build a GardenBot! One that knows how to do gardening, learns about the growths and changes to the plants / flowers and behaves accordingly. Learns when too much water has been used, when to "thin out" the plants, when to add more water, when to cut out dead flowers, etc... but that'll be a 2021 goal target (well, it's been a goal for over a decade, but at least I'm doing something about it now!)
https://www.wired.com/story/watsons-creator-teach-ai-new-trick-common-sense/
This all ties in nicely with this article that was just published. The brains behind Watson (yeah, that thing that paid my bills for a few years), has decided to read some of the same books that I've had on the shelf and due to him not really needing any money, he can "play at R&D" to his hearts content. Am I jealous? sure. Do I want to call him a git, under my breathe, sure. Should I ping him an email and ask if we can work on something together? errr...maybe not just yet. Let me get my autonomous boats and tanks working and "thinking" first.
https://www.wired.com/story/watsons-creator-teach-ai-new-trick-common-sense/
Comments
Post a Comment