[Tease AI Java] [NEED YOUR HELP] Modernizing Tease personalities

[ski23]

In the poll here(if you haven't voted yet please do), I can see that modernizing personalities/adding new modules is a high priority. I would like to build a new personality unlike any other from the ground up. With this new personality, I will attempt to solve as many problems as I can that cause existing personalities to get old. The end goal here is to be able to create 1 large single personality that most users could enjoy using frequently for a long time without it getting boring, repetitive, or predictable. I have wanted to do this for a long time and have lots of ideas on how to address many existing issues. However, to make this personality as good as possible, I need your help to identify issues with existing personalities and why you might have not liked a personality or stopped using it. I have put together this list of issues I commonly see with personalities. Please comment with any information you can about any other issues you may have or any information about something already on the list. Getting this information now is crucial so that I can design the personality with solving these issues in mind:

• Flow feels disjointed like you’re talking to someone different in each module with no memory of chat history.
• Modules get repeated and it feels repetitive and predictable.
• Images displayed are usually irrelevant and displayed at random times.
• There isn’t a whole lot of user interaction and when there is, it feels repetitive and predictable.
• There is no sense of realism about who the domme is or any information about him/her that makes him/her seem like a real person instead of just a bit sending commands.
• Hardly any games and the existing ones aren’t super well thought out.
• No good punishment system in any personality.
• No very good reward system in any personality.
• Not much customization about how long is spent edging/talking/stroking/cool down time/etc
• Normal Stroking parts can be boring
• If toys are used, don’t use too many for cleaning purposes.


Also, I have created this document for anyone interested (its pretty technical and might be boring to some) listing these problems and ideas for solutions I'm exploring for a new personality: https://docs.google.com/document/d/16Op ... sp=sharing

[avatarbr]

Good luck

I took a look at your doc, and this could ending super complex (and a lot of work for 1 person only)

• Flow feels disjointed like you’re talking to someone different in each module with no memory of chat history.
• Modules get repeated and it feels repetitive and predictable.

What I find about these 2 is, or you get 1 or other. More ramdomize you get, less will looks like the same person. You could try to work with moods to get this.

• Normal Stroking parts can be boring

Yeah. I think I could cut 50% of the stroking part in HoT. Since you can call modules at any point, you can work on that to break the circle. I am happy with the solution I got when the sub are denied for a long time. In this case, almost all "breaks" lead to a specific module series about that, skiping the stroking part.

Another way is create alternative stroking. I am working on a Mazu (from GuideMe) inspirated "stroking part" now. I will ask the author if I can include in a future public version.

[ski23]

Good luck

I took a look at your doc, and this could ending super complex (and a lot of work for 1 person only)

• Flow feels disjointed like you’re talking to someone different in each module with no memory of chat history.
• Modules get repeated and it feels repetitive and predictable.

What I find about these 2 is, or you get 1 or other. More ramdomize you get, less will looks like the same person. You could try to work with moods to get this.

• Normal Stroking parts can be boring

Yeah. I think I could cut 50% of the stroking part in HoT. Since you can call modules at any point, you can work on that to break the circle. I am happy with the solution I got when the sub are denied for a long time. In this case, almost all "breaks" lead to a specific module series about that, skiping the stroking part.

Another way is create alternative stroking. I am working on a Mazu (from GuideMe) inspirated "stroking part" now. I will ask the author if I can include in a future public version.

1. First off, yes it is a lot of work but, I've been thinking about some of the details for months. I've done bigger projects so it's definitely doable but it is a big undertaking. To begin with though, I'm setting up the structure of the personality and all the code to make the personality itself run in a way that eliminates these issues. Once that's done, if some other devs want to come on, I'd welcome them as long as they follow the paradigm.

2. You are right at least to a certain extent that these 2 issues are fighting each other. I mention this a bit in the doc. So the issue about modules feeling repeatable is easily solvable by having a lot of content and using a history. However, obviously it would take a lot of manual labor to get that much content. And for the other one involving flow, having long modules will help the flow. So, my thought is the best compromise is to use long modules with lots of random phrases (while keeping the same general meaning) to try to improve both of these issues without affecting the other too much. Also using a tree structure as I mentioned in the doc could be an elegant solution (will need to test that one though because its just a theory).

3. I agree that the stroking parts can get boring and it is easier in TAJ to eliminate them. I'm not entirely sure what you mean by alternative stroking but that sounds potentially interesting.

Thanks for the reply!

[FATALES]

The personality should be fully adjustable.
For example. I want edging , buttplug, ice cube, exerciese but I don't want cbt, pegging etc.
I also wondered how to make each session unique - yes you need a lot of modules.
My idea:
1.Start sission and now depending on what the user has set. For example: CBT and edge.
2.Randomly choose whether cbt or Edge
3.If if stroke will follow and vice versa.
4. Pussy lick.
5.Randomly choose whether cbt or edge.
6.Randomly choose whether cbt or edge.
7. Edge
8.**********
9.**********
10.**********
11. Denial or Ruin - End session
It is not easy to program such a thing. Random phrases may also be selected.

[holaba_be]

I think you have identified all of the major issues. But I guess it mostly comes down to "more content". It seems to me a lot of the content from the most popular & complete TAI personalities (MB & HoT) is interconnected to the entire personality. Which makes it probably more interesting from a realism stance, but I assume it also makes it harder to transfer content because modules now assume for example that certain personality dependent flags exist. So anything you can make to ease transfer of content is a bonus. It would help you to import some "standalone" content from other personalities, and would help other creators to import some of your content.

I think the current games are not really an interesting part of TAI(J). They take too long and are in no way sexual. Getting a punishment or reward at the end is not interesting enough to warrant the time spent on the game. If you do a short session, spending 10 minutes on Risky Pick, is not the most exciting thought. So if games are a part of your personality, I would try to make sure they are at least somewhat sexual during the actual game as well.

It would help if was more clear what some answers do when being asked what the player wants to do. For example, if the Domme asks about anal or CEI for the first time, and you say "no", does that mean she won't ever ask about it again, or that she might try to convince you later. I know some personalities have some content about the Domme trying to convince the player to do something he initially does not want. Which is a really cool idea, but if as a player I don't know how long I can "refuse" before she gives up, I might miss out on the content. I know this might be somewhat immersion breaking of course, so I don't know how to implement this in a good way.

A better history system would also help the previous issue. Knowing what's been going on during the previous sessions would allow for longer "story lines". For example, the Domme might have her heart set on forcing CEI, so she keeps denying the player at the end of every session until he agrees to it. Currently personalities have a check at the beginning of each session to ask about something that might have been asked at the end of the previous one. Expanding on this might make it far more immersing. It would probably also increase variety when there is a better history of which module has already been done. Since content is not endless, some of it is bound to repeat. But it should try to repeat only when each other module has already been done (unless it's some highly specific module that should be rare by design).

Repetitiveness is the biggest bane of personalities. Both because of the stroke-module-stroke-module repetition, but also because some content just is repetitive. For example, each time I install a new personality (update), I do a search on the MultipleEdges command, to replace it with more sensible numbers. Having someone do dozens of edges, with little or no talking or anything else in between is just boring. But I imagine this can be different for everyone, so perhaps some kind of "Maximum multiple edges" setting, like the "Maximum edge holding time", might help here.

What I also think is less important:
* Irrelevant pictures: I think this can only be solved by a lot of tagged pictures. Which is going to scare away anyone who does not want to go through that. The alternative is providing pre-tagged pictures, but I think the fact that you can now choose the Domme to be anyone you like, is too important to give up.
* Extra user interaction should not be overdone. Like I've said in the other topic, if a lot of content depends on user interaction, chances are high it will be missed.

[ski23]

The personality should be fully adjustable.
For example. I want edging , buttplug, ice cube, exerciese but I don't want cbt, pegging etc.
I also wondered how to make each session unique - yes you need a lot of modules.
My idea:
1.Start sission and now depending on what the user has set. For example: CBT and edge.
2.Randomly choose whether cbt or Edge
3.If if stroke will follow and vice versa.
4. Pussy lick.
5.Randomly choose whether cbt or edge.
6.Randomly choose whether cbt or edge.
7. Edge
8.**********
9.**********
10.**********
11. Denial or Ruin - End session
It is not easy to program such a thing. Random phrases may also be selected.

I also agree that this is a major problem in some personalities where you cant customize your fetishes or even if you do, the domme still has you do stuff you don’t like. I’ll add this to the doc. This is a good one.

[ski23]

I think you have identified all of the major issues. But I guess it mostly comes down to "more content". It seems to me a lot of the content from the most popular & complete TAI personalities (MB & HoT) is interconnected to the entire personality. Which makes it probably more interesting from a realism stance, but I assume it also makes it harder to transfer content because modules now assume for example that certain personality dependent flags exist. So anything you can make to ease transfer of content is a bonus. It would help you to import some "standalone" content from other personalities, and would help other creators to import some of your content.

I think the current games are not really an interesting part of TAI(J). They take too long and are in no way sexual. Getting a punishment or reward at the end is not interesting enough to warrant the time spent on the game. If you do a short session, spending 10 minutes on Risky Pick, is not the most exciting thought. So if games are a part of your personality, I would try to make sure they are at least somewhat sexual during the actual game as well.

It would help if was more clear what some answers do when being asked what the player wants to do. For example, if the Domme asks about anal or CEI for the first time, and you say "no", does that mean she won't ever ask about it again, or that she might try to convince you later. I know some personalities have some content about the Domme trying to convince the player to do something he initially does not want. Which is a really cool idea, but if as a player I don't know how long I can "refuse" before she gives up, I might miss out on the content. I know this might be somewhat immersion breaking of course, so I don't know how to implement this in a good way.

A better history system would also help the previous issue. Knowing what's been going on during the previous sessions would allow for longer "story lines". For example, the Domme might have her heart set on forcing CEI, so she keeps denying the player at the end of every session until he agrees to it. Currently personalities have a check at the beginning of each session to ask about something that might have been asked at the end of the previous one. Expanding on this might make it far more immersing. It would probably also increase variety when there is a better history of which module has already been done. Since content is not endless, some of it is bound to repeat. But it should try to repeat only when each other module has already been done (unless it's some highly specific module that should be rare by design).

Repetitiveness is the biggest bane of personalities. Both because of the stroke-module-stroke-module repetition, but also because some content just is repetitive. For example, each time I install a new personality (update), I do a search on the MultipleEdges command, to replace it with more sensible numbers. Having someone do dozens of edges, with little or no talking or anything else in between is just boring. But I imagine this can be different for everyone, so perhaps some kind of "Maximum multiple edges" setting, like the "Maximum edge holding time", might help here.

What I also think is less important:
* Irrelevant pictures: I think this can only be solved by a lot of tagged pictures. Which is going to scare away anyone who does not want to go through that. The alternative is providing pre-tagged pictures, but I think the fact that you can now choose the Domme to be anyone you like, is too important to give up.
* Extra user interaction should not be overdone. Like I've said in the other topic, if a lot of content depends on user interaction, chances are high it will be missed.

1. I can definitely use other personalities content to some degree but likely a lot will have to be modified to fit in with this newer design. I definitely will utilize previously written content though.
2. I also agree repetitiveness is one of the biggest problems. Having a large amount of content and a history system will help with that. Also what you’re talking about with too many edges and whatnot is also part of one of the other issues I brought up about there isnt much customization about how much time is spent stroking/edging/ talking. In existing personalities, the multipleEdges thing uses preset number of edges and is repetitive with not much chat while edging. In this personality, I will use a function similar but the number of edges will be determined based on your settings and there will be a lot more interesting dialogue during and between edges.
3. This one will get solved mostly by the program im making that uses deep learning to automatically tag images.
4. I’d like to have a lot of user interaction but I will make sure there are multiple ways to get to a module regardless of your input so the content isnt “missed”. Also I can document the types of interaction and what they can do.

[fusiontermite]

The personality should be fully adjustable.
For example. I want edging , buttplug, ice cube, exerciese but I don't want cbt, pegging etc.

I agree to this, mainly also because you said your goal was to create a personality that can be played by almost all users (i.e. regardless of their sexual preferences).

For me, what being able to adjust the domme's behaviour to your preferences means, is, one the one hand, that the sub can agree or disagree to each fetish, and, on the other hand, that the domme checks the fetishes in much detail. For example, I like both edging and kneeling (so the domme should create a flag for both of it), but I can't stand edging or holding an edge while on my knees. Similary, I really love playing with clothespins but I hate putting them on my nipples (when I'm given the command to do so, I just put them somewhere else, but preferably, the domme wouldn't even tell me to).

With regard to clothespins this means that the domme does not only check if the sub likes playing with clothespins but also what body parts he agrees to put them on. When introducing clothespins, the domme would ask if the sub is ok with playing with clothespins in general, but also if he agrees to put them on his nipples, his tongue, his balls, his cock or his body. She would also check wearing clothespins for an entire session separately. Similarly, the domme also checks the other fetishes in detail. For example, edging is not the same as holding an edge and holding an edge is different to having to hold 10 edges in a row or holding an edge while kneeling.

For the technical implementation this means that the personality needs to create separate flags for edging, kneeling, edging while kneeling, clothespins in general, clothespin on nipples, clothespins on tongue, etc.

In addition to having flags that state whether or not the sub has agreed to a certain fetish ("agreed_edgekneel"), there could also be flags stating that the sub particularly likes or dislikes a certain fetish (e.g. "hates_edgekneel", "loves_edgekneel"). Having these flags allows the modules to adapt to the sub's preferences even more. For example, in a module that involves putting clothespins to your body the domme would check the "loves_pinstongue" flag and tell the sub "Since you like it so much I want you to put two clothespins on your tongue as well". Similarly, the same module could also check the "loves_kneel" flag and tell the sub to do it all while kneeling.

The "hates_" flags can be useful to make statements like "I know you don't like kneeling that much / But you will just have to get through that, #name". Besides, it can also be combined with the domme's mood, so that the domme chooses to make the sub things he dislikes when she is in a bad mood. (Please note that "hates_" flag refer to fetishes the sub has agreed to, but simply doesn't like especially much. The domme would never make the sub do something he has not agreed to.)

Lastly, there should be a limit for days without orgasm. During normal sessions, this would guarantee an orgasm after a certain amount of days. During a long denial period, it would end the long denial period (or at least the domme would ask the sub if he wants it to end, i.e. at least giving him the chance to end it) and make sure the sub is allowed to cum (because the Tease AI settings could still make the domme deny the sub although long denial has ended).

I would like to build a new personality unlike any other from the ground up. With this new personality, I will attempt to solve as many problems as I can that cause existing personalities to get old. The end goal here is to be able to create 1 large single personality that most users could enjoy using frequently for a long time without it getting boring, repetitive, or predictable. I have wanted to do this for a long time and have lots of ideas on how to address many existing issues.

BTW: This sounds amazing!! Good luck and thanks for all your efforts!

[FATALES]

The personality should be fully adjustable.
For example. I want edging , buttplug, ice cube, exerciese but I don't want cbt, pegging etc.

I agree to this, mainly also because you said your goal was to create a personality that can be played by almost all users (i.e. regardless of their sexual preferences).

For me, what being able to adjust the domme's behaviour to your preferences means, is, one the one hand, that the sub can agree or disagree to each fetish, and, on the other hand, that the domme checks the fetishes in much detail. For example, I like both edging and kneeling (so the domme should create a flag for both of it), but I can't stand edging or holding an edge while on my knees. Similary, I really love playing with clothespins but I hate putting them on my nipples (when I'm given the command to do so, I just put them somewhere else, but preferably, the domme wouldn't even tell me to).

With regard to clothespins this means that the domme does not only check if the sub likes playing with clothespins but also what body parts he agrees to put them on. When introducing clothespins, the domme would ask if the sub is ok with playing with clothespins in general, but also if he agrees to put them on his nipples, his tongue, his balls, his cock or his body. She would also check wearing clothespins for an entire session separately. Similarly, the domme also checks the other fetishes in detail. For example, edging is not the same as holding an edge and holding an edge is different to having to hold 10 edges in a row or holding an edge while kneeling.

For the technical implementation this means that the personality needs to create separate flags for edging, kneeling, edging while kneeling, clothespins in general, clothespin on nipples, clothespins on tongue, etc.

In addition to having flags that state whether or not the sub has agreed to a certain fetish ("agreed_edgekneel"), there could also be flags stating that the sub particularly likes or dislikes a certain fetish (e.g. "hates_edgekneel", "loves_edgekneel"). Having these flags allows the modules to adapt to the sub's preferences even more. For example, in a module that involves putting clothespins to your body the domme would check the "loves_pinstongue" flag and tell the sub "Since you like it so much I want you to put two clothespins on your tongue as well". Similarly, the same module could also check the "loves_kneel" flag and tell the sub to do it all while kneeling.

The "hates_" flags can be useful to make statements like "I know you don't like kneeling that much / But you will just have to get through that, #name". Besides, it can also be combined with the domme's mood, so that the domme chooses to make the sub things he dislikes when she is in a bad mood. (Please note that "hates_" flag refer to fetishes the sub has agreed to, but simply doesn't like especially much. The domme would never make the sub do something he has not agreed to.)

Exactly except that she would have you kneeling , sometimes not.
It is similar to tying balls. Will be tied to session or next ? Definitely domme. When you have to tie them at the beginning , in the middle , before the end, you should untie them during the session ? I don't like CBT :D

[ski23]

The personality should be fully adjustable.
For example. I want edging , buttplug, ice cube, exerciese but I don't want cbt, pegging etc.

I agree to this, mainly also because you said your goal was to create a personality that can be played by almost all users (i.e. regardless of their sexual preferences).

For me, what being able to adjust the domme's behaviour to your preferences means, is, one the one hand, that the sub can agree or disagree to each fetish, and, on the other hand, that the domme checks the fetishes in much detail. For example, I like both edging and kneeling (so the domme should create a flag for both of it), but I can't stand edging or holding an edge while on my knees. Similary, I really love playing with clothespins but I hate putting them on my nipples (when I'm given the command to do so, I just put them somewhere else, but preferably, the domme wouldn't even tell me to).

With regard to clothespins this means that the domme does not only check if the sub likes playing with clothespins but also what body parts he agrees to put them on. When introducing clothespins, the domme would ask if the sub is ok with playing with clothespins in general, but also if he agrees to put them on his nipples, his tongue, his balls, his cock or his body. She would also check wearing clothespins for an entire session separately. Similarly, the domme also checks the other fetishes in detail. For example, edging is not the same as holding an edge and holding an edge is different to having to hold 10 edges in a row or holding an edge while kneeling.

For the technical implementation this means that the personality needs to create separate flags for edging, kneeling, edging while kneeling, clothespins in general, clothespin on nipples, clothespins on tongue, etc.

In addition to having flags that state whether or not the sub has agreed to a certain fetish ("agreed_edgekneel"), there could also be flags stating that the sub particularly likes or dislikes a certain fetish (e.g. "hates_edgekneel", "loves_edgekneel"). Having these flags allows the modules to adapt to the sub's preferences even more. For example, in a module that involves putting clothespins to your body the domme would check the "loves_pinstongue" flag and tell the sub "Since you like it so much I want you to put two clothespins on your tongue as well". Similarly, the same module could also check the "loves_kneel" flag and tell the sub to do it all while kneeling.

The "hates_" flags can be useful to make statements like "I know you don't like kneeling that much / But you will just have to get through that, #name". Besides, it can also be combined with the domme's mood, so that the domme chooses to make the sub things he dislikes when she is in a bad mood. (Please note that "hates_" flag refer to fetishes the sub has agreed to, but simply doesn't like especially much. The domme would never make the sub do something he has not agreed to.)

Lastly, there should be a limit for days without orgasm. During normal sessions, this would guarantee an orgasm after a certain amount of days. During a long denial period, it would end the long denial period (or at least the domme would ask the sub if he wants it to end, i.e. at least giving him the chance to end it) and make sure the sub is allowed to cum (because the Tease AI settings could still make the domme deny the sub although long denial has ended).

I would like to build a new personality unlike any other from the ground up. With this new personality, I will attempt to solve as many problems as I can that cause existing personalities to get old. The end goal here is to be able to create 1 large single personality that most users could enjoy using frequently for a long time without it getting boring, repetitive, or predictable. I have wanted to do this for a long time and have lots of ideas on how to address many existing issues.

BTW: This sounds amazing!! Good luck and thanks for all your efforts!

This is a great point and something I've thought about regarding further customization of when to call fetish modules beyond just whether the user doesnt like that fetish or not. There are also times where a user is into a specific fetish but only as a punishment. I'll add this to the doc and make sure it's something that is planned.

[avatarbr]

4. I’d like to have a lot of user interaction but I will make sure there are multiple ways to get to a module regardless of your input so the content isnt “missed”. Also I can document the types of interaction and what they can do.

Something I did for that was the Domme "leading" the use to say something. Like this:

Domme: I wonder if this pic will make your balls ache even more (put a @Timeout here, so the script don't stop if the user don't say anything)
Use: Yes/ it makes/etc. (trigger the response file for "balls aching)

I used this in cases where I created something big enough from interactions, and I want to give more chances someone will see it.


Other case, MissBlue have a lot of responses triggered only if some flags are active. This helps to keep the conversation more real.

[clayk0]

just an idea i have:
making the domme use the webcam as she likes, giving you things to do while taking pics and teasing you with them sometimes. "you want to cum? cum to yourself"

if you ever got the auto-tagger integrated, she can even tell if you did some things correct or censor yourself from yourself.. ouch..

another thing the auto-tagger can do is start censoring your whole porn section. "You liked that pic? well, here you have it, nice and shiny- and censored"

[FATALES]

It's a difficult task to program AI

[ski23]

4. I’d like to have a lot of user interaction but I will make sure there are multiple ways to get to a module regardless of your input so the content isnt “missed”. Also I can document the types of interaction and what they can do.

Something I did for that was the Domme "leading" the use to say something. Like this:

Domme: I wonder if this pic will make your balls ache even more (put a @Timeout here, so the script don't stop if the user don't say anything)
Use: Yes/ it makes/etc. (trigger the response file for "balls aching)

I used this in cases where I created something big enough from interactions, and I want to give more chances someone will see it.


Other case, MissBlue have a lot of responses triggered only if some flags are active. This helps to keep the conversation more real.

Right now I’ve been leaning toward the missblue sort of style with responses that get triggered with active flags.

[ski23]

just an idea i have:
making the domme use the webcam as she likes, giving you things to do while taking pics and teasing you with them sometimes. "you want to cum? cum to yourself"

if you ever got the auto-tagger integrated, she can even tell if you did some things correct or censor yourself from yourself.. ouch..

another thing the auto-tagger can do is start censoring your whole porn section. "You liked that pic? well, here you have it, nice and shiny- and censored"

Yes being able to censor is something that might be possible. I havent decided yet on that. Its possible right now but I might need to remove that capability.