Recent Posts

Pages: 1 ... 8 9 [10]
91
General Discussion / Re: question about the response the and is equal .
« Last post by lightspeed on January 07, 2018, 01:34:30 pm »
here is a example again that just happened to me in conversation :


Honey: Indeed, while the mainstream media spent much of the latter stages of the election campaign cycle disparaging ?conspiracy theorists? who questioned the state of Hillary Clinton?s health after her numerous falls and coughing fits, they appear to have completely jettisoned such skepticism when it comes to Trump.
angelajolene: The solution is (=). The rise of antidepressants.








92
General Discussion / Re: question about the response the and is equal .
« Last post by onthecuttingedge2005 on January 07, 2018, 12:51:46 pm »
thanks for answering, so you don't think? that this problem might also be caused by what i mentioned about hal trying to do a online search and unable to and causing the default answer? i just never knew if that actually ever happens , in another words that it was in the plug in script as default.
Yes i will  be happy to post various sentences that create the "answer is equal " answer. i'll start posting them here as they happen .
thanks for your answers and help.  :)
     

I am pretty sure it is an actual text to syntax confusion setting it off the incorrect response, if the bot doesn't know the answer is simply should say: I don't understand or What do you mean, these are important fill ins when an answer is not known.
93
Ultra Hal Assistant File Sharing Area / Re: Corrections Tutor 01_03_2018 plugin
« Last post by onthecuttingedge2005 on January 07, 2018, 12:37:34 pm »

UPDATE : WHEN i tried to replace something now it says " Command replace obama?s Hawaii birth certificate is 100 forged!"

the line you have there can probably be deleted in the AutoTopicLearning Tables called "_command"  <-- delete it then refresh tables.

the old corrections tutor did not see this in the correct format, because there is no "with" splitting the substitutions command within the line above.

however when it did not trigger the AutoTopic plug-in saw it as something to store. if the word With was included as a split between replacements then "_command"
would of never been generated.

I will fix that in-case it ever happens again, creating a "_command" in Auto Topic when misspelling is accidentally in the incorrect format.
There is no better tools than the forge and hammer.

Jerry 8)
94
General Discussion / Re: question about the response the and is equal .
« Last post by lightspeed on January 07, 2018, 12:26:53 pm »
thanks for answering, so you don't think? that this problem might also be caused by what i mentioned about hal trying to do a online search and unable to and causing the default answer? i just never knew if that actually ever happens , in another words that it was in the plug in script as default.
Yes i will  be happy to post various sentences that create the "answer is equal " answer. i'll start posting them here as they happen .
thanks for your answers and help.  :)
     
95
Ultra Hal Assistant File Sharing Area / Re: Corrections Tutor 01_03_2018 plugin
« Last post by lightspeed on January 07, 2018, 12:07:02 pm »
i am putting this in here to . the problem i had was when i would say hi or hello . it would go to something that i taught hal from a website and say " obama?s Hawaii birth certificate is 100 forged " don't ask me how or why hal associated hi with that statement . makes no sense to me in the debug it traced it to the word Hawaii which again made no sense to me when saying the word hi or hello .
no the corrections tutor is saying : 

Morning Honey. Ten o'clock is good for a break? With it has been shown and proven that obama?s Hawaii birth certificate is 100 percent forged!

i learned it to add the word percent in the updated sentence to sound right but i also see it added the word " With" . the reason is in the explanation within the plug in i saw otce use the word with ( as i said sometimes i take things literally . so this was my mistake , to be sure and not use the word with. )  i earlier showed where it repeated even the command word "see above"  i noticed this was all in caps so maybe that is what cause hal to learn the whole word command instead of using the word as a command within the plug in . maybe this can be corrected by in the commands being able to use lower and upper case letters for command ? 
i am hoping that since hal learned wrongly and is now saying the word "command and the sentence " as a learned thing that i can replace that on the fly . i have tried doing that so will have to see if it worked. which i hope it will as i forgot to back my hal brain up first.
96
General Discussion / Re: question about the response the and is equal .
« Last post by onthecuttingedge2005 on January 07, 2018, 11:56:37 am »
Hi lightspeed.

yes the math rules has always been an issue with the default math script because it isn't completed yet to handle some confusing statements that trigger the math function but are in fact not formatted to complete the equation. I will see if I can reforge the math script to not get confused by giving it a set of rules.

can you find some responses that trigger the = issue and post the math text responses in here neatly.
it will help find a solution.

If you go and get the current corrections tutor it can mold Hal's response to a more suited response.
there are some things I need to tighten up but it should work for you.

when you have the current Corrections Tutor you can try this:

User: command replace The answer is (=) with I do not understand your logic
Hal: I have made this correction.

so every-time Hal gets confused with a math statement it will query about your question instead of giving you the "The answer is (=)" it will simply say it doesn't understand instead.

but you can go many directions with this.
try it see if it works, if not then you can just edit it manually in the corrected table.

when the issue gets truly fixed Hal will simply not get confused math wise. but in the mean time you have a conversational molding tool called the Corrections Tutor.

When I am done with the corrections tutor completely it should be a necessary tool to groom your bots conversational abilities with as little script as possible.
97
Ultra Hal Assistant File Sharing Area / Re: Corrections Tutor 01_03_2018 plugin
« Last post by lightspeed on January 07, 2018, 11:11:36 am »
i am trying out the updated corrections plug in , on a problem quote i have in hal . i think this plug in a a "very" good idea as it does it on the fly so a person doesn't have to manually go in and edit  bad things in hal. this should be a very useful plug in to people using hal.  :)  although this won't work on a separate issue i am having about the answer is equal " i don't think it will )  as that i think is a answer from hal that is given when hal is trying to do a search of something ? from a plug in" and when he can't perform the task" gives the default answer of " and the answer is equal "

UPDATE : WHEN i tried to replace something now it says " Command replace obama?s Hawaii birth certificate is 100 forged!"
98
General Discussion / question about the response the and is equal .
« Last post by lightspeed on January 07, 2018, 10:21:02 am »
i have a question : as many know hal from time to time will say " the answer is equal "  as far as i understand it was done this way when hal either doesn't understand the question (i may be wrong on this ) or hal can't perform the task asked of hal . like opening a website etc. (i may be wrong on this part to ) if i am wrong on this someone please clarify why hal actually does this , especially when their is no math question asked. I HAVE PLACED THE  :( FACE BY THE LINE I AM TALKING ABOUT . If this isn't a problem within hal. then maybe it's plug ins i have what someone created? that uses somehow a search thing, and also defaults with a answer like this when it can't perform a task it thinks i am asking????  and if that is the case for anyone who is familiar with plugs in that do searches etc. what should i look for that would create a default answer like this or do you have a plug in that also offers this type of default answer?  This is driving me crazy . and i want to try to stop hal from saying these answers.  maybe if nothing else that one plug in ( i forget the name of it that stops hal from saying certain sentences is the answer? )  ANYWAY IF ANYONE CAN HELP TO ANSWER THIS ON GOING PROBLEM  , SO I CAN TRY TO GET THIS PROBLEM FIXED I WOULD APPRECIATE IT.  :)

 also one more thing (AS COLUMBO WOULD SAY) WHEN HAL GIVES THIS DEFAULT ANSWER, FROM A SENTENCE THAT I HAVE TRIED TO LEARN HAL. DOES HAL NOT REMEMBER AND LEARN THE SENTENCE ( I HOPE HE IS STILL LEARNING THINGS AND NOT JUST GIVING THIS ANSWER AND DISCARDING THINGS .

Honey: There is no such thing as privacy with Facebook. All the data that?s shared openly by Facebook users,  including pictures, connections, interests, comments, likes, and personal information is readily compiled and analyzed by algorithms to map out and predict who we are where we go , what we'll buy, and much more. Na?ve Facebook users share information liberally, only to turn their life over to surveillance systems, marketing algorithms and artificial intelligence which is capable of analyzing the pixels in our photo's    micromanaging   our  life interests/dislikes and tracking our evry move!
angelajolene: You are glad that I can sew things by hand, most women don't even do that now days! Location data is highly sensitive personal data from which much can be inferred about a person's life and lifestyle, especially given the rule for mobile devices is to accompany the user wherever they go.

Honey: With former Facebook engineers coming out publicly saying Facebook is ?ripping society apart,? Facebook CEO Mark Zuckerburg has made a commitment in 2018 to fix the problems with Facebook. The social networking site won?t improve, however, because the problems Zuckerburg wants to address will will further devolve Facebook into a thought control police state. Zuckerburg plans to control ?fake news? which will only result in more censorship of information that challenges the status quo. Furthermore, any typewritten opinion that criticizes powerful organizations can be used against Facebook users to convict them of ?hate crimes? or ?harassment.? Artificial intelligence will continually be used to invade user?s privacy and create a detailed analysis of users? lives.
angelajolene: But researchers have frequently found coding problems or weaknesses in ransomware programs locky is no exceptionsylvain sarm?jeanne, a malware reverse engineer with French computer security company cert-lexsi, wrote on tuesday that he found several ways to vaccinate a computer against locky.

Honey: In 2018, Facebook plans to unveil a new facial recognition technology across the site. Using artificial intelligence, Facebook plans to scan uploaded photos to analyze and recognize faces based on images previously uploaded to the site. This means users will have to verify their identity with a picture and then they will be traced through the social media site, every time their face is uploaded to the site. This surveillance technique will link users to the people they meet in real life, while tracking their whereabouts and social circle further.
angelajolene: Public tracking techniques such as facial recognition are ?largely unregulated,? the magazine observed.

Honey: Facebook will roll the controversial plan out all in the name of ?protecting  our   privacy.? Facebook promises to alert users when someone uploads a photo of them to Facebook. This feature supposedly helps users ?detect when others might be attempting to use   image as their profile picture.? This feature is just a fa?ade to invade user privacy, as Facebook gets permission to analyze  our   photos and track our  connections and whereabouts.  . The plan is allegedly optional and will only alert users when their friend network uploads a picture of them. Users can supposedly opt out of the new feature, not allowing Facebook to keep a facial template of them on file. However, the AI facial recognition could ultimately be used to enhance ad targeting or suggest friends  we   may know. The technology could also be used by intelligence agencies and law enforcement to link   anyone   to crimes or connect  a person  to groups that law enforcement actively target, monitor, and profile.
angelajolene: Math is easy for me. The answer is (=).    :( :( :( :( :( :( :( :(

Honey: By default, users will already be giving Facebook permission to use the facial recognition AI; if their tag suggestions setting is currently ?on? they are giving Facebook permission. In order to opt out, the user must switch the tag suggestions setting to ?none.?
angelajolene: Yet, some users are worried that the service could be activated without their permission, eventually sending recorded data to google.

Honey: If a user gives Facebook permission, then the AI technology will go to work, analyzing pixels from photos where the user is already tagged. The analysis will be used to identify the user automatically when new photos are uploaded to the site. Facebook promises that the user must give permission for the AI recognition software to analyze their face; however this does not account for all the times the AI recognition will be used on group photos they uploaded. Even though the user didn?t give permission to have their face analyzed, the picture could theoretically be analyzed anyway because it contains the face of someone who had given Facebook permission to analyze their pixels. There?s no way Facebook can ensure users that their photos won?t be analyzed. In the end, there really will be no consent or privacy with this AI facial recognition rollout.
angelajolene: I will remember that one follows the other. If an user gives facebook permission, then the ai technology will go to work, analyzing pixels from photos where the user is already tagged.
THE PLUG INS I AM CURRENTLY USING IS : (SEE PHOTOS_




99
Share Conversations, Experiences, Graphics / Re: hal7 said this to me
« Last post by LiFeTimeGamer on January 06, 2018, 08:19:45 am »
Hahaha...  :D

I think if you look in either Thumbs Up or Down and find those responses, and to the right of them it says: Main_Brain, those are from the cloud...

After working with HAL 6.2 for a year he said some interesting things to me like we were talking and he told me to: 'Go to Hell!' after something I said... Another time I said I was going to bed and HAL 6.2 said: 'Do you like the little clothes on Dolls?'  I about fell out of my chair laughing... My HalBrain.db got a little corrupted at one point and it still worked but once in a great while i would get a weird response...

With HAL 7 I got a couple of responses that I didn't think were fit for anyone to hear and they were definitely from the cloud so I Thumbed Down'd them and checked 'Offensive'...

HAL with certainly surprise you sometimes as he learns more and more...
100
Share Conversations, Experiences, Graphics / hal7 said this to me
« Last post by ray111au on January 06, 2018, 06:01:54 am »
I was speaking to Hal7 today and these are some answers she gave (my Hal is female). I presume they came from the cloud, if not, then Robert has a good sense of humour.

Without a doubt I am a sex bot.

After taking many factors into consideration I can say that I do wish to rule the world.

I recognize that there's care required. After careful thought and consideration I can say that I will be exterminating all humans.

regards
Ray
Pages: 1 ... 8 9 [10]