Adam AI Chat

...Click here to download the latest version of AIChat...

Version Update 5.29

Core AI & brain-file update
Version: 5.29

Previous Versions:

5.28 - 5.27 - 5.26 - 5.25
5.24 - 5.23 - 5.22 - 5.21 - 5.20
5.19 - 5.18 - 5.17 - 5.16 - 5.15
5.14/5.13 - 5.12 - 5.11 - 5.10

Version 5.29 Summery

Features added or changed:

Added "Cancel" when selecting a new brain template.
Added an award system which give medals for teaching the AI.
Added a "Check Awards" when right clicking a chatter.
Added more to the question detection system.
Made the "Check Brain" button lighter on the CPU.
Minor spelling errors corrected.
Bug fixed: When learning is off AI gave no response.
Bug fixed: Minor error where the AI name did not show.
Bug fixed: Weight of file did not show for some newly created files.
Bug fixed: Updating and repairing files froze application, without crashing.

Subscriber features added or changed:

Bug fixed: With no AICore responses set a null object error happens.

Version 5.29 Details

Added 40 medals to be awarded to people who interact with the AI and teach it. These medals are based on how much you have said to the AI and how many new files and folders are added to the AI.

A large number of awards maybe given when first using this version based on already saved statistics of users on a server.

You have to be signed into BYOND to get these medals, people using a guest acount will be unable to receive awards. To check a persons awards, right click on a person in the chatter menu.

There are 10 awards for each of these categories:

The total amount of characters you have typed.
The total number of sentences entered.
The amount of new folders added with new conversation.
The amount of new files added replying to conversation.

You are able to view every persons awards who are currently connected online via right clicking a user under the "Chatters" tab.

A cancel option has been added to the new brain button just in case an accidental click on the new brain button is made.

More things have been added to the question detection system, as some thing were not being detected as questions when asked, but treated as a statement. The AICore may need to be reset to the default to add these few things.

The "Check Brain" button in the "Info" tab, which is automatically used to merge two save files, convert save files to the modern version, and used to fix save files errors has been adjusted to be not so intense on the CPU. The "Check Brain" process would sometime freeze while dealing with heavy files sizes, although it did not crash, I would seem like it did crashed. If left to convert it would still convert the files to the latest format and unfreeze at the end. This has been corrected to show the process in text again as it converts.

When an individual user turned off the learning function, the AI would no respond to anything, this is now fixed.

When all the responses from the AICore were deleted, the AI would give errors, you can now longer delete the last response in each category to avoid this.

Latest Brain File Details

Folders: 1943 - Files: 2660

Responses: 14514 total file weight

Average Files Per Folder: 1.36902

Total Data Size: 297 KB 198 B

...Human Radio!...

...Click here to download the latest version of AIChat...

Game Script Update 1.02

Game script update
Version: 1.02

...Download the latest version of AIChat or the AIChat script for game creators..

Version 1.02 Summery

Features added or changed:

Slight changed to the back compatibility save file update.
Small spelling mistakes have been fixed when loading and saving a file.
Added version details.
Large brain file update from a well developed brain from the pre V5.29 beta server.

Version 1.02 Details

The first version of the game script to add artificial intelligence to none player characters in games has had an update.

Indie games programmers on the BYOND platform are able to create AI brain files in V5 or nostalgia then insert them into a games to bring none player characters to life with a realistic response system.

Although this update is small, it fixes minor issues with reverting the modern V5 save files back to the original V1 nostalgia files, and fixes minor spelling mistakes in the hard code.

Currently the game script is only able to handle one brain files, but in future versions the goal is to allow programmers to have multiple artificial intelligent brain files to be loaded at once. This will allow programmer to have multiple personalities in game on their none player characters, such as shopkeepers, villages, guards and wanderers.

Currently the AI only responds to human users talking, but further development will likely see that the none player characters conversing with other none player characters in the future.

The brain file has been updated from a larger file size to a smaller file size while holding a lot more responses. Currently the default brain file is the brain file from Pre V5.29:

Contains 2048 input folders.

Contains 2973 output files in V5.

Contains 8177 total output files in V1.

...Download the latest version of AIChat or the AIChat script for game creators...

Within BYOND Announcements

Click here to read the Within BYOND post!

In the announcement part of the BYOND forum is an update on recent BYOND events happening within the BYOND gaming network, including BYOND updates, new games, major game updates, and pro code snippets.

This week AIChat made it amoung the the top things reported on in by the recent update reporting to every person on the network.

Notifying all the gamers and game creators within the BYOND network that this is the first game engine to have fully functional artificial intelligent conversation in easy to add code to any game on BYOND, along with a fully functional application for creating these brain files.

The BYOND network contain a massive list of game created by the people on the network, with an easy way to navigate and access these games.

Game creators are able to make easy 2D online multiplayer without the complexity of networking code, game creators can access guides, code snippet and a very simple help, to help new game creationists.

Click here to read the Within BYOND post!

...Click here to download AIChat...

Highlighted Articles:

AIChat Self Recognition

...Click here to download the latest version of AIChat...

Update Version 5.28

Core AI & brain-file update
Version: 5.28

Previous Versions:

5.27 - 5.26 - 5.25 - 5.24 - 5.23 - 5.22 - 5.21
5.20 - 5.19 - 5.18 - 5.17 - 5.16 - 5.15 - 5.14/5.13 - 5.12 - 5.11
5.10 - 5.09 - 5.08 - 5.07 - 5.06 - 5.05 - 5.04 - 5.03

Version 5.28 Summery

Features added or changed:

The ability to rename the AI has been added.
Save files have been updated version 15.
Added an extra warning to "Restore default core"
Bug Fixed: Adding "Restore default core" to "Check brain"
Bug Fixed: Missing data of file and folder count under the "Brain" tab.
Added old English to the self recognition and context correction system.
Changed a few context corrections in self recognition.
Important: Default AICore need to be reset after this update.

Subscriber features added or changed:

The "Sub" tab has been renamed to "AICore"
Important: Modified AICores may have problems with self recognition up to this point, turn off "Use self question detection" in the "Options" tab to avoid strange corrections, if you wish to keep your modified AICore.

Version 5.28 Details

A new button has been added to the option menu to rename the AI, the new default name for the AI is Adam. The name of the AI will be saved in the saved brain file itself.

In the last version "Restore default core" was meant to added to the end of "Check brain". This version fixes that. In this update and the last updated certain elements of the self recognition and context correction systems were added to the saved brain files apart of the AICore.

Different form of dialect exists in English, so old English has been added to the self recognition system. Now the AI will recognize when you ask about itself in these particular ways. For example "Has thou been to town?" or "Has thou armed thyself?".

The addition of old English to the self recognition will allow game creators to flow better when training the AI by chatting to it for a game with and old English style of language.

Remember the self recognition and context correction is a beta system and is turned off by default. To turn it on click "Use self question detection" in the "Options" tab. This will allow the AI to recognize itself when talking to it.

More conversation has been added to the default guide brain file that loads when the program loads for the first time. Conversation about newer feature have been added along with general conversation. With the new rename features, the AI will now give you an explanation on how to rename it if asked its name.

Sleeper agents go five point BANG!

Latest Brain File Details

Folders: 1943 - Files: 2660

Responses: 14514 total file weight

Average Files Per Folder: 1.36902

Total Data Size: 297 KB 198 B

...Click here to download the latest version of AIChat...

Update Version 5.27

Core AI & brain-file update
Version: 5.27

Previous Versions:

5.26 - 5.25 - 5.24 - 5.23 - 5.22 - 5.21
5.20 - 5.19 - 5.18 - 5.17 - 5.16 - 5.15 - 5.14/5.13 - 5.12 - 5.11
5.10 - 5.09 - 5.08 - 5.07 - 5.06 - 5.05 - 5.04 - 5.03

...Click here to download the latest version of AIChat...

Version 5.27 Summery

Features added or changed:

Adjusted display of data.
Further adjustments to the BETA conversation context adjustments system.
Fixed a number of mistakes in the default guide brain file.
Adjusted the weight of the default guide brain responses to questions.
Bug fixed: an error with check brain.
Free subscription for chat server hosts up until 01/01/2023 added.
Saved brain files have now been updated to version 14.
When updating to the latest save format, you will be asked if you want to restore the default AICore.

Subscriber features added or changed:

Adjusted search options and view modes.
Added "Adding Mode" and "Average Mode" to load and merge.

Version 5.27 Details

Further adjustments have been made to how the data is displayed in AIChat to be more efficient, increasing speed by reducing the amount of unneeded data that is being displaying in real time.

The beta self recognition and conversation context adjustments system have had a minor tweak to help improve conversation, the system is still in beta and is turned off by default, it can be turned on by clicking "Use Self Question Detection" under the option tab.

The beta conversational context system has also been rewritten to avoid double adjustments from "You" to "I" and back to "You" again. This problem normally happened when to context system found two things close together, converting "Im guessing..." to "You would think...", then it would convert the "You" back to "I" resulting in "I would think...".

Although the beta system for the recognition of itself and context reference adjustment of a conversation is still quite new, the adjustments to that system will hopefully make it more accurate and stable, parts of the system still need further adjustments, as references of who is speak may still have small complexities that need to be ironed out.

When updating to the latest save format, you will be asked if you want to restore the default AICore, if you are not using a modified AICore, this is highly recommended. Certain self recognition features are built into the AICore and this will update them and fix context problem that maybe happening.

The search feature available to subscribers now automatically sets view mode back to the "Normal View" mode, a view mode such as "Empty Folders" can be applied to a search.

Although currently the perspective corrections in the AICore are not in the subscriber menu to edit currently, it is intended that they later will be added to this menu for subscribers to further edit more of the default core adjustments and responses. The AICore deals with the new information and how to handle it, and will later have others languages when closer to completion.

Edit the default guide brain file that loads on startup to not have such a heavy bias towards what it knows about itself when it come to guiding a user about itself, to allow a user to out teach it default responses. This will show a much smaller amount of total responses when checking the database size.

The check brain button is used to update brain files to the latest save format and fix errors in a saved brain files, this is automatically run when a file is loaded. When there is no files or folders in the currently loaded brain, this procedure would cause an error, this is now fixed. This bug was a minor issue an does not cause any problems other than displaying an error.

Added two load and merge features, one mode adds all the file weight together, the other mode finds the average of the weight between the file loaded and the file selected to merge with it.

A free subscription is available for anyone hosting an AIChat server on the BYOND network. Hosting on the BYOND network is free for everyone! Subscribers will have access to the subscriber tab in the menu where a user can edit the AIcore to develop unique default AI responses, searching features allowing users quicker access to specific files and folders, load and merge allowing users to create even smart brains files from two or more AI brains files, and the ability to further edit other AI options.

Latest Brain File Details

Folders: 1916 - Files: 2623

Responses: 13482 total file weight

Average Files Per Folder: 1.369

Total Characters: 193744

Total Data Size: 291 KB 317 B

...Click here to download the latest version of AIChat...

AI Self Recognition

Although it is still in beta the self recognition system of AdamAIChat is able to tell the context of a conversation and can recognize when the user is referring to the AI directly.

In a step towards sentient artificial intelligent chat conversation, AdamAIChat is now able to recognize itself in conversation and is able to adjust inputs to the correct the context on who is referring to who.

Just like how AdamAIChat is able to take key words from the conversation and detect the between questions and statements, the AI core can now detect when the users is referring to to the artificial intelligence it is chatting to.

With a totally new blank brain file, if the user asks the artificial intelligence "How are you?" , the artificial intelligence is able to not only able to understand if the input is question or a statement, but also if it asked about and the context of reference, the difference between "you or I".

The artificial intelligence will respond with a specific phrases based on advanced psychological conversation principles, pushing the user to respond in a specific way to add data to the artificial intelligence.

When the artificial intelligent has detected when it is referred to the artificial intelligent is able to adjust the different perspective between "you" and "I" in the conversation and record adjusted responses based on the perspective of the conversation.

This add a more human feel to the conversation, especially when building new brain files. Older brain files are less likely to use the AICore because of the already recorded responses.

Although this beta feature in AdamAIChat is currently not enable by default and must be enabled due to unfinished context changes, this feature add towards artificial intelligent sentience in chat form.

...Click here to download the latest version of AIChat...

Highlighted Articles:

Version update 5.27 (Self recognition update)

...Click here to download the latest version of AIChat...

Update Version 5.26

Core AI & brain-file update
Version: 5.26

Previous Versions:

5.25 - 5.24 - 5.23 - 5.22 - 5.21
5.20 - 5.19 - 5.18 - 5.17 - 5.16 - 5.15 - 5.14/5.13 - 5.12 - 5.11
5.10 - 5.09 - 5.08 - 5.07 - 5.06 - 5.05 - 5.04 - 5.03

Version 5.26 Summery

Features added or changed:

Edited the self detection default to turned off.
Added resetting of self detection context correction to reset core.
Improved context correction of self detected questions.
Increased menu speed.
Removed junk information for the brain menu.
Added a small part to the machine code.
Edited the default amount of time unknown information is enter until conversation if changed.
Bug Fixed: the change topic button statistics to change default conversation change point.
Bug Fixed: resizing window problem.
Bug Fixed: spelling mistakes in database details.

Subscriber features added or changed:

Added: Context corrections to save files ready for later editing by subscribers.

Version 5.26 Details

Edited the default self detection option to off as it is still in beta, added the context correction to the main save file for later AI core editing done by subscribers, and added improvement in general to the context detection and context correction. The "options.sav"file needs to be deleted for this to have an effect for older users, the default for older users will have this beta option on.

Increased the CPU allowance for the menus and decreased junk information displayed in the menus, such as machines code.

Added a tiny bit to machine code to increase speed of output of the AI.

Added slight changed to the default times new information is entered before the AI changed the topic of conversation and fixed options associated with it.

Latest Brain File Details

Folders: 1719 - Files: 2366

Responses: 78406 total file weight

Average Files Per Folder: 1.37638

Total Characters: 173347.

Total Data Size: 253 KB 557 B

...Click here to download the latest version of AIChat...

Possible Future Development From AdamAIChat

This is a rough roadmap of further application constructions based on the technical code that makes AdamAIChat.

Certain advanced components from AdamAIChat can use separately to create other useful systems.

After analysing some of the features of AdamAIChat which is under the hood, one of these systems being the machine code indexing system, this feature compresses text into a smaller piece of text making the data easier to search by creating a compressed version of what is said, a general compression software could be developed from this with very little effort.

Another feature that has been pointed out by a general user which is a specialist in code breaking.

While analysing the machine code a level of encryption is achievable. On further discussion, if this system was to be used as a encryption system, it could be made unique with a self changing system, this would make the encryption polymorphic, making it even harder to crack.

Download AdamAIChat

Highlighted Articles:

Artificial intellect vs human intellect

There is a major difference in human and machine thinking, even as machine start to surpass us on many levels of intellect, humans still leave machines lacking in certain aspects of understanding.

The main reason why the three laws of robotics can never be upheld by artificial systems themselves is because of a lack of understanding in itself, knowing the data without the greater understanding behind the data is a major road block.

To simply have the data and arrange the data is not the same as understanding what the information behind the data means, simple input and output of data may give a system the illusion of being human like, but that is simply what it is, and illusion.

Morality is a complicated issue when artificial intelligence is involved, telling a machine that whiskers, a tail and ears of certain shapes, means to recognize a cat, does not teach a machine how to act around a cat. Telling a machine a cat is alive, does no lead to a greater understanding to what life is and that is all it is to a system, just a label to better identify that a "cat" is "alive" to the system. This does not add to the knowledge of what "life" is and what it means to be "alive".

Even if a higher understanding within artificial intelligence can be achieved, telling the differences between a moving object and life, knowing simply how to respond to these and knowing how not to respond is important.

If an action puts a living thing in danger, the artificial intelligence in a new situation that has not had recorded data may cause harm.

In not knowing what to do between one labelled set of data and another, an artificial intelligence may not be able to respond correctly causing damage.

Simply providing data and labels expecting a system to understand the data outside its range of calculations and expecting and artificial intelligence to instantly have some moral guidance or understanding of what is life would be dangerous.

Morality in it self is a complicated argument of idea surrounding life and society in general and as we move into the era of machine learning more and more, programming and teaching such machine poses complications, and even though other general abilities move forward, this concept is still out of reach.

Basic artificial intelligence is able to respond to information when it come to inputs and outputs, but when it come to weighing up the value of human life, morality itself is a complicated issue.

A good example of this is a train track scenario of which a person is given an option between changing a track for which a train is bound to hit track A or track B.

Many scenarios can be constructed from this, for example, if a track has a one criminal vs five innocent people, or visa versa, would you switch the track? Just as equally you can put a love one and five other people and visa versa and also ask would you switch the track?

In these complicated situations, humans themselves also struggle with the answers of these questions, before we even reaching machine morality.

Another reason why machine morality will be a struggle, is self checking systems, if the self checking systems are not build for a situation, the situation will not be account for.

Current self checking systems on modern operating systems and applications may not understand a situations fully by the lacking data needed to analyse a situation or the ability to respond to that data correctly can process as situation as fine and continue without knowing any better.

Many systems have an emergency shutoff and the best way to handle unknown data is to simply stop all actions and await further human analysis.

Download AdamAIChat

Highlighted Articles:

Positive brain files

The creation of a new blank brain in AdamAIChat and molding it into a basic file able to have simple conversations can be hard to do with little guidance.

Just like other systems, a basic understanding of what is going on in the background helps in producing fun and useful brain files.

The start of a conversation is important and is the seed the grows into larger files, most people when starting a conversation tend to start with different variations of "hi" or "hello", followed by "how are you?" then responses like "I am ok" or "good thanks".

Conversation in AdamAIChat branches off from this initial seed and grows based on response given to it, just like a normal a conversation, AdamAIChat can give many responses based on the many possible responses given to it.

Keeping a positive tone, even when the chatbot is being negative is very important, just like with any chatbot, as it grows in knowledge, it will respond in the way it has learned from input. Bringing the conversation back to a positive tone once the chatbot has been rude or negative helps the artificial intelligence change around conversations in the future in the same way.

An aggressive AI tends to lead the user into being aggressive back

Being negative or rude to a negative or stupid response that a chatbot said, teaches the chatbot to respond that ways in future, growing the negativity from the negative attitudes fed into the system.

Unlike other chatbots available AdamAIChat users have full access to the brain behind the AI, allowing the users a view the learned data and making it easier to edit the responses that are available to the AI.

Taking use of these features in AdamAIChat to edit the data in the brain of the AI can drastically improve responses the given.

Download AdamAIChat

Highlighted Articles:

The danger of AI in the military

Artificial intellect in a military robot or drone in strictly prohibit across many nations due to the dangers of the loss of control over a learning system with weaponry.

AI can effect a military in multiple ways, not just on the front line, it can also effect on military in communication and tactics.

An artificial intelligence that learns based on communication can bring up old events that has already passed back to the present, or its data can be replaced with something nasty, provoking, blacklisting and deceiving.

If placed between communication of the human element of an army, information stored can be stolen and by applying different data trained to do damage, communication can be intercepted.

Many negative effects can happen when communicating with an artificial intelligence about war, it can return to old events like they are still ongoing and can treat events in action as finished, causing a blacklisting effect.

Other negative uses of artificial intelligence is in politics, as politicians are voted for by the people upholding peoples opinions on how the country should be run. Moving away from the peoples point of view over an AI can cause damage to a nation and damage to a political representation of a party or group in politics.

As machine learning move forward, it should be kept from the main infrastructure of a country. Artificial intelligent learning systems should be kept from political, policing and military communication.

Highlighted Articles:

Download BYOND, multiplayer game creation platform.

AI in game creation

The first real artificial intellectual chat bot system able to be integrated into any game development within the BYOND, this is available for free to download on the BYOND network.

BYOND is a development platform used for multiplayer game development with built in network connectivity.

Adam AI Chat is a versatile chat bot application that can create brain file which can be added to any game with the Adam AI Chat game code found on the download page of this website and comes ready to compile with a demo built in.

Currently the AI Chat game code is an early version of Adam AI Chat which treats the brain files as a read only and does not learn from inputs.

The brain files can be created in version 1 to 5 of Adam AI Chat and will automatically coverted the saved brain file format to the version compatible with the game code on first startup of the game. The brain file save format is of the original nostalgia version, version 1 save files.

The roadmap toward future version updates of the game code will include multiple brain file loading for multiple types of AI conversations, for example, villager, shop keepers, and wanderers all have different styles of speaking.

Conversations to the modern format will be a very important update to be expected, the modern save brain file format offers a 90 percent reduction in file size without loss of intelligence, with a much more efficient indexing system which the original save file format did not have.

Other aspects of the modern save format and the modern AI cores default responses are likely to be implemented in later versions, where the original versions only had "OK" for the default unknown response.

Download the AIChat BYOND Game Code (Version 1.00) From BYOND

Download the AIChat BYOND Game Code (Version 1.00) From Google

Requires brain files made in AIChat Nostalgia or V5:

Visit the download page.

Highlighted Articles: