tag:blogger.com,1999:blog-1600624546380487204.post7895201102963100940..comments2023-10-21T12:54:34.276+02:00Comments on Fake Guido: Guerilla - my attempt to build a strong AIDaivdhttp://www.blogger.com/profile/14328955839644654804noreply@blogger.comBlogger12125tag:blogger.com,1999:blog-1600624546380487204.post-7501327824326287922012-09-25T21:17:55.226+02:002012-09-25T21:17:55.226+02:00Picking up on hints is a question of tuning our an...Picking up on hints is a question of tuning our antennae to cues to recognize structured signals. <br /><br />Entangling mixture of DNA shows us the importance in evolution of getting signals crossed. <br /><br />My brain crosses signals thoroughly. <br /><br />The madness of continuously knotted signals sometimes tunes into special programming for the gifted... <br /><br />Periodic submission to the cultured insanity of sanity hill-climbing grows interesting perspectives. Stanley Marcinkowskihttps://www.blogger.com/profile/10254222193417248897noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-13203695325493225382012-03-01T14:44:28.784+01:002012-03-01T14:44:28.784+01:00Ai building efforts start at definitions: Ai that...Ai building efforts start at definitions: Ai that can<br />specify goals and weight them , acquire combine breakdown and refine strategy.<br /><br />A strategy specifies goals, their desirability and at what likelihoods to take what actions on what (set of) conditions. <br /><br />Devising strategies can be broken down into:<br />creating and assessing conditions for actions,<br />weight of goals, estimates of cost for actions,<br />estimates of effectiveness of actions, finding related strategies,<br />taking strategies apart,<br />combining strategies,<br />covering contingencies,<br />evaluating strategiesStanley Marcinkowskihttps://www.blogger.com/profile/10254222193417248897noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-5183221133289439332011-11-22T17:45:17.470+01:002011-11-22T17:45:17.470+01:00I would be interested to see the input from which ...I would be interested to see the input from which this AI (when implemented) would be able to learn how to play the 5-in-a-row game.Jiri Jelinekhttps://www.blogger.com/profile/05280188662848085641noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-5011027768669565382011-11-22T17:43:51.047+01:002011-11-22T17:43:51.047+01:00This comment has been removed by the author.Jiri Jelinekhttps://www.blogger.com/profile/05280188662848085641noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-30370213647820156992011-11-21T11:40:20.951+01:002011-11-21T11:40:20.951+01:00@acetoline No, the project is not abandoned, but t...@acetoline No, the project is not abandoned, but thanks for asking :). I tend to post infrequent, overambitiously long posts, so a few weeks silence is normal.<br /><br />The reason the github activity is low is more silly. I am currently in something between the design and implementation stage, writing Python code with a few pseudocode elements and a lot of prose. For some reason, I have not considered this semi-code "commit-worthy".<br /><br />I promise a github update this week.Daivdhttps://www.blogger.com/profile/14328955839644654804noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-50391352540566916522011-11-20T07:24:44.991+01:002011-11-20T07:24:44.991+01:00Hi, I noticed there hasn't been any activity o...Hi, I noticed there hasn't been any activity on your blog or github lately. I hope you haven't abandoned the project.acetolinehttps://www.blogger.com/profile/13539791277823376781noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-88997788266712125942011-11-12T20:00:33.116+01:002011-11-12T20:00:33.116+01:00Doesn't sound like a well scalable solution. D...Doesn't sound like a well scalable solution. Don't get overexcited/misled after some early luck in well defined toy worlds. With teaching by manual algorithm entry by techies, you aren't gonna get very far.Jiri Jelinekhttps://www.blogger.com/profile/05280188662848085641noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-86058516543559999842011-11-12T12:27:56.629+01:002011-11-12T12:27:56.629+01:00Hopefully, yes, it should be able to solve general...Hopefully, yes, it should be able to solve general problems using more specialized algorithms working together. It will not, however, take a set of specialized algorithms (let's say playing chess, checkers, poker and backgammon) and produce a general game playing algorithm. That is not how it achieves generality.<br /><br />It is geared towards very technical users. It takes input tasks as snippets of code and gives a set of inputs that makes the function output true. This is called function inversion and is a fairly simple way of describing puzzles and technical problems.<br /><br />If it turns out to be a useful system for solving these types of tasks (a big IF - no one has really been able to achieve that). It would be a very good base on which to build something that can communicate with non-technical users and interact with our fuzzy world. That is not it's primary purpose, though.Daivdhttps://www.blogger.com/profile/14328955839644654804noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-49718296917384649642011-11-12T10:02:09.462+01:002011-11-12T10:02:09.462+01:00Can this 'AGI' generate general algorithms...Can this 'AGI' generate general algorithms from a set of relevant non-general algorithms? Will non-technical users be able to teach this AI by describing specific (/non-general) scenarios?Jiri Jelinekhttps://www.blogger.com/profile/05280188662848085641noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-53531214030658159962011-11-11T22:34:10.785+01:002011-11-11T22:34:10.785+01:00@Jiri Swedes, Norwegians, Danes and many Finns can...@Jiri Swedes, Norwegians, Danes and many Finns can read Swedish. That makes up a good 0.3% of the earth population :).<br /><br />Actually, I will remove that. That source code is not for human consumption yet. It is just test cases for analyzing source code, written in an odd Lisp dialect. No actual code relating to implementing either any of the algorithms I write about or the Scheduler.Daivdhttps://www.blogger.com/profile/14328955839644654804noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-27252457688388169622011-11-11T22:07:18.769+01:002011-11-11T22:07:18.769+01:00One way to build a strong AI is outlined in the ht...One way to build a strong AI is outlined in the <a href="http://mind.sourceforge.net/aisteps.html" rel="nofollow">http://mind.sourceforge.net/aisteps.html</a> and develops into a simple but gradually <a href="http://www.scn.org/~mentifex/AiMind.html" rel="nofollow">expandable AI Mind</a>.Mentifexhttps://www.blogger.com/profile/04530921525903314824noreply@blogger.comtag:blogger.com,1999:blog-1600624546380487204.post-30991040961126675322011-11-11T20:01:15.330+01:002011-11-11T20:01:15.330+01:00Don't use Swedish in the source, man! 'Nob...Don't use Swedish in the source, man! 'Nobody' can read that ;-)Jiri Jelinekhttps://www.blogger.com/profile/05280188662848085641noreply@blogger.com