These include Guidelines for secure AI system development (by the UK National Cyber Security Centre and US Cybersecurity and Infrastructure Security Agency), National Institute of Standards and Technology Risk Management Framework, MITRE Adversarial Threat Landscape for AI Systems and Open ...

Launch of Guidelines and Companion Guide on Securing Artificial Intelligence Systems

Companion

companion

Besides a regular webapp and a Mac app, which uses the same endpoints as the webapp, Home Assistant also provides mobile apps. In this post, I want to describe its advantages over the former. I'll use iOS to do this: Samsung's French partner to recycle used mobiles tricked me once, and I moved away from Samsung. I decided to buy Apple for better integration since I've had Mac computers for the last few years. Installing and Configuring The companion app is readily available at the Apple Store. Note that the app is both free as a beer and free as a bird. It's provided under the Apache v2 License. If interested, the source code is on GitHub. Once installed, you must configure at least one Home Assistant server. At this point, you should get the same UI as the web app. The app's power lies in its widgets, though. Widgets Some iOS applications allow you to add widgets to the UI. I'm not a massive user of widgets. I've Google Calendar, the Weather, and a VPN on my phone. Widgets serve a dual purpose: visualization and possibly action; if a widget doesn't offer an action, pressing it will open the associated app. That's the case with the Google Calendar and Weather apps. My VPN widget displays a button allowing you to connect and pause. The Companion App Widget The app offers two distinct kinds of widgets, each serving a specific purpose. The first kind allows opening the app on a specific "page", e.g., Overview or Developer tools. IMHO, it's not very useful. The second kind allows calling  Home Assistant commands. However, it requires significantly more effort. First, we must create action objects. Here's one to switch the light on: actions:
  - name: OfficeLightsOn                 #1
    label:
      text: "Turn ON Office Lights"      #2
    icon:
      icon: lightbulb-on-outline         #3
      color: "#ffff00"                   #3
    show_in_carplay: false
    show_in_watch: true Name, but you should probably think about it as a String ID.


The label is shown in the companion app.


Graphic details used by the companion app After creating the action, you must restart Home Assistant for devices to detect the new action. Once restarted, you can add a Home Assistant widget that is bound to that action. However, pressing the widget does nothing but a slight haptic feedback. We need to bind the action to an automation object. I explained automations when we migrated away from the Hue bridge in the previous post. At the time, we created an automation via the UI; we can also create one in the /homeassistant/configuration.yaml. We configure the action as the automation's trigger. - alias: Turn On Office Lights On        #1
  trigger:
    - platform: event                    #2
      event_type: ios.action_fired       #2
      event_data:
        actionName: OfficeLightsOn       #3
  action:
    - service: light.turn_on             #4
      entity_id: light.all_office        #5 Friendly alias for easier maintenance.


Fixed payload sent to the device.


Configured payload sent to the device. If the payload matches, it will trigger the action.


Service to call.


Call the service on the device ID. Now, pressing the widget turns on the light as expected. Summary Integrating Home Assistant on your device is straightforward. The pressing on the widget is the trigger for an automation - an action. You can create the automation or reuse an existing one. To go further: Home Assistant Companion Docs Originally published at A Java Geek on January 12th, 2025 Besides a regular webapp and a Mac app, which uses the same endpoints as the webapp, Home Assistant also provides mobile apps. In this post, I want to describe its advantages over the former. I'll use iOS to do this: Samsung's French partner to recycle used mobiles tricked me once, and I moved away from Samsung. I decided to buy Apple for better integration since I've had Mac computers for the last few years. Installing and Configuring The companion app is readily available at the Apple Store . Note that the app is both free as a beer and free as a bird . It's provided under the Apache v2 License. If interested, the source code is on GitHub . Apple Store free as a beer free as a bird GitHub Once installed, you must configure at least one Home Assistant server. At this point, you should get the same UI as the web app. The app's power lies in its widgets, though. Widgets Some iOS applications allow you to add widgets to the UI . I'm not a massive user of widgets. I've Google Calendar, the Weather, and a VPN on my phone. to add widgets to the UI Widgets serve a dual purpose: visualization and possibly action; if a widget doesn't offer an action, pressing it will open the associated app. That's the case with the Google Calendar and Weather apps. My VPN widget displays a button allowing you to connect and pause. The Companion App Widget The app offers two distinct kinds of widgets, each serving a specific purpose. The first kind allows opening the app on a specific "page", e.g. , Overview or Developer tools. IMHO, it's not very useful. e.g. The second kind allows calling  Home Assistant commands. However, it requires significantly more effort. First, we must create action objects. Here's one to switch the light on: action actions:
  - name: OfficeLightsOn                 #1
    label:
      text: "Turn ON Office Lights"      #2
    icon:
      icon: lightbulb-on-outline         #3
      color: "#ffff00"                   #3
    show_in_carplay: false
    show_in_watch: true actions:
  - name: OfficeLightsOn                 #1
    label:
      text: "Turn ON Office Lights"      #2
    icon:
      icon: lightbulb-on-outline         #3
      color: "#ffff00"                   #3
    show_in_carplay: false
    show_in_watch: true Name, but you should probably think about it as a String ID. The label is shown in the companion app. Graphic details used by the companion app Name, but you should probably think about it as a String ID. Name, but you should probably think about it as a String ID. The label is shown in the companion app. The label is shown in the companion app. Graphic details used by the companion app Graphic details used by the companion app After creating the action, you must restart Home Assistant for devices to detect the new action. Once restarted, you can add a Home Assistant widget that is bound to that action. must However, pressing the widget does nothing but a slight haptic feedback. We need to bind the action to an automation object. I explained automations when we migrated away from the Hue bridge in the previous post . At the time, we created an automation via the UI; we can also create one in the /homeassistant/configuration.yaml . We configure the action as the automation's trigger . automation previous post /homeassistant/configuration.yaml trigger - alias: Turn On Office Lights On        #1
  trigger:
    - platform: event                    #2
      event_type: ios.action_fired       #2
      event_data:
        actionName: OfficeLightsOn       #3
  action:
    - service: light.turn_on             #4
      entity_id: light.all_office        #5 - alias: Turn On Office Lights On        #1
  trigger:
    - platform: event                    #2
      event_type: ios.action_fired       #2
      event_data:
        actionName: OfficeLightsOn       #3
  action:
    - service: light.turn_on             #4
      entity_id: light.all_office        #5 Friendly alias for easier maintenance. Fixed payload sent to the device. Configured payload sent to the device. If the payload matches, it will trigger the action. Service to call. Call the service on the device ID. Friendly alias for easier maintenance. Friendly alias for easier maintenance. Fixed payload sent to the device. Fixed payload sent to the device. Configured payload sent to the device. If the payload matches, it will trigger the action. Configured payload sent to the device. If the payload matches, it will trigger the action. Service to call. Service to call. Call the service on the device ID. Call the service on the device ID. Now, pressing the widget turns on the light as expected. Summary Integrating Home Assistant on your device is straightforward. The pressing on the widget is the trigger for an automation - an action . You can create the automation or reuse an existing one. action automation To go further: To go further: Home Assistant Companion Docs Home Assistant Companion Docs Home Assistant Companion Docs Originally published at A Java Geek on January 12th, 2025 Originally published at A Java Geek on January 12th, 2025 A Java Geek

The Home Assistant Companion App: How to Integrate it to Your Device

JobQuest Assistant: How I Built a Smart Companion for Stress-Free Job Hunting With Coze

Asistente de JobQuest: cómo creé un compañero inteligente para buscar trabajo sin estrés con Coze

JobQuest Assistant: nola sortu nuen Estresik gabeko Lan Ehizarako Lagun Adimendun bat Coze-rekin

JobQuest Assistant: איך בניתי בן לוויה חכם לציד עבודה נטול מתחים עם Coze

JobQuest Assistant: Coze を使ってストレスフリーな就職活動のためのスマートなコンパニオンを構築する方法

ជំនួយការ JobQuest: របៀបដែលខ្ញុំបង្កើតដៃគូឆ្លាតវៃសម្រាប់ការស្វែងរកការងារដោយគ្មានភាពតានតឹងជាមួយ Coze

JobQuest Assistant: Coze로 스트레스 없는 취업 활동을 위한 스마트한 동반자를 만든 방법

JobQuest asistents: kā es izveidoju gudru kompanjonu darba meklējumiem bez stresa ar mājīgumu

JobQuest Yanapakuq: Imaynatataq huk Smart Compañerota ruwarqani Cozewan Mana Estrés Llamkana Cazapaq

JobQuest Assistant: Coze සමඟ ආතතියෙන් තොර රැකියා දඩයම් කිරීම සඳහා මම ස්මාර්ට් සහකාරියක් ගොඩනඟා ගත් ආකාරය

JobQuest kömekçisi: Coze bilen stresiz iş awlamak üçin akylly ýoldaş nädip gurdum

JobQuest Assistant: як я створив розумного компаньйона для пошуку роботи без стресу з Coze

Trợ lý JobQuest: Cách tôi xây dựng một người bạn đồng hành thông minh để tìm việc không căng thẳng với Coze

AI companions are more than just a passing trend; they could become a $150 billion industry by 2030, according to a recent article by Ark Invest. With the potential to revolutionize how we interact with technology and each other, AI companionship represents the next frontier in digital entertainment. Companion apps like Replika, Gatebox, HeraHaven, and SoulMachines attract millions of users every month, according to Similarweb data. But if you’re considering building an AI companion, where do you start? Specifically, which large language model (LLM) should you choose to power your app? Let me walk you through the best open-source LLMs currently available and why they might be the perfect fit for your project. Choosing the Right LLM for AI Companionship When thinking about AI companionship, one of the first decisions you’ll need to make is what kind of open-source LLM to use. There are several types of LLMs, but they generally fall into three categories: base models, instruction-tuned models, and chat-tuned models. Why Chat-Tuned Models Are Essential Base models are the raw, unrefined versions of LLMs that have been trained on large datasets but lack the fine-tuning required for specific tasks. Instruction-tuned models have been refined to follow human instructions more accurately. However, when it comes to creating AI companions, chat-tuned models are the way to go. These models are specifically optimized for conversational interactions, making them better suited for the nuanced, ongoing dialogues that characterize AI companionship. Model Size: How Big Is Big Enough? Another key consideration is the size of the model. For most conversational chatbots, an 8B parameter model should suffice. These models offer a good balance between performance and resource efficiency, making them ideal for most applications. However, if you’re looking to go above and beyond—perhaps aiming to create a more sophisticated AI—you might want to explore models with up to 70B parameters. The 5 Best Open-Source LLMs for AI Companionship Here’s a list of some of the best open-source LLMs that are particularly well-suited for creating AI companions: 1. Hermes-3 Llama-3.1-8B Hermes-3 Llama-3.1-8B is a powerful model that’s been finely tuned for chat applications. It offers a good balance between capability and efficiency, making it a top choice for AI companionship. 2. Yi-1.5-9B-Chat Yi-1.5-9B-Chat is another excellent option, boasting a slightly larger parameter count, which allows for more complex and nuanced conversations. If your AI companion needs to handle a wide range of topics with ease, this model is worth considering. 3. InternLM2 5-7B Chat InternLM2 5-7B Chat is a lighter model that still packs a punch. Its smaller size makes it ideal for applications where resources are limited but conversational depth is still a priority. 4. Humanish-Roleplay-Llama-3.1-8B Humanish-Roleplay-Llama-3.1-8B is designed for roleplay scenarios, making it perfect for users who want an AI companion that can take on different personas. Whether it’s a friendly chat or a more complex interaction, this model excels at maintaining character. 5. OpenChat-3.5-1210 OpenChat-3.5-1210 is one of the larger models on this list and is optimized for deep, engaging conversations. If your AI companion needs to offer an immersive experience, this model is a strong contender. If you’re interested in seeing how these models stack up against others, you can visit the Open LLM Leaderboard for the most recent comparisons. Taking Your AI Companion to the Next Level with Fine Tuning While these models are impressive right out of the box, there’s always room for improvement. With enough prompting, you should be able to get decent results for most applications. However, if you’re looking for a more advanced approach, consider fine-tuning one of these models to better suit your specific needs. Fine-tuning allows you to tweak the model’s behavior, making it more aligned with your desired user experience. In conclusion, the potential for AI companions is vast, and the right LLM can make all the difference in creating a compelling, engaging experience. Whether you’re building the next million user app or a more niche application, these open-source models provide a strong foundation for success. So, dive in, experiment, and see what you can create with the incredible power of modern LLMs. AI companions are more than just a passing trend; they could become a $150 billion industry by 2030, according to a recent article by Ark Invest . With the potential to revolutionize how we interact with technology and each other, AI companionship represents the next frontier in digital entertainment. according to a recent article by Ark Invest Companion apps like Replika, Gatebox, HeraHaven, and SoulMachines attract millions of users every month, according to Similarweb data. But if you’re considering building an AI companion, where do you start? Specifically, which large language model (LLM) should you choose to power your app? Let me walk you through the best open-source LLMs currently available and why they might be the perfect fit for your project. Choosing the Right LLM for AI Companionship When thinking about AI companionship, one of the first decisions you’ll need to make is what kind of open-source LLM to use. There are several types of LLMs, but they generally fall into three categories: base models, instruction-tuned models, and chat-tuned models. Why Chat-Tuned Models Are Essential Base models are the raw, unrefined versions of LLMs that have been trained on large datasets but lack the fine-tuning required for specific tasks. Instruction-tuned models have been refined to follow human instructions more accurately. However, when it comes to creating AI companions, chat-tuned models are the way to go. These models are specifically optimized for conversational interactions, making them better suited for the nuanced, ongoing dialogues that characterize AI companionship. Model Size: How Big Is Big Enough? Another key consideration is the size of the model. For most conversational chatbots, an 8B parameter model should suffice. These models offer a good balance between performance and resource efficiency, making them ideal for most applications. However, if you’re looking to go above and beyond—perhaps aiming to create a more sophisticated AI—you might want to explore models with up to 70B parameters. The 5 Best Open-Source LLMs for AI Companionship Here’s a list of some of the best open-source LLMs that are particularly well-suited for creating AI companions: 1. Hermes-3 Llama-3.1-8B 1. Hermes-3 Llama-3.1-8B Hermes-3 Llama-3.1-8B is a powerful model that’s been finely tuned for chat applications. It offers a good balance between capability and efficiency, making it a top choice for AI companionship. 2. Yi-1.5-9B-Chat 2. Yi-1.5-9B-Chat Yi-1.5-9B-Chat is another excellent option, boasting a slightly larger parameter count, which allows for more complex and nuanced conversations. If your AI companion needs to handle a wide range of topics with ease, this model is worth considering. 3. InternLM2 5-7B Chat 3. InternLM2 5-7B Chat InternLM2 5-7B Chat is a lighter model that still packs a punch. Its smaller size makes it ideal for applications where resources are limited but conversational depth is still a priority. 4. Humanish-Roleplay-Llama-3.1-8B 4. Humanish-Roleplay-Llama-3.1-8B Humanish-Roleplay-Llama-3.1-8B is designed for roleplay scenarios, making it perfect for users who want an AI companion that can take on different personas. Whether it’s a friendly chat or a more complex interaction, this model excels at maintaining character. 5. OpenChat-3.5-1210 5. OpenChat-3.5-1210 OpenChat-3.5-1210 is one of the larger models on this list and is optimized for deep, engaging conversations. If your AI companion needs to offer an immersive experience, this model is a strong contender. If you’re interested in seeing how these models stack up against others, you can visit the Open LLM Leaderboard for the most recent comparisons. Open LLM Leaderboard Taking Your AI Companion to the Next Level with Fine Tuning While these models are impressive right out of the box, there’s always room for improvement. With enough prompting, you should be able to get decent results for most applications. However, if you’re looking for a more advanced approach, consider fine-tuning one of these models to better suit your specific needs . Fine-tuning allows you to tweak the model’s behavior, making it more aligned with your desired user experience. consider fine-tuning one of these models to better suit your specific needs In conclusion, the potential for AI companions is vast, and the right LLM can make all the difference in creating a compelling, engaging experience. Whether you’re building the next million user app or a more niche application, these open-source models provide a strong foundation for success. So, dive in, experiment, and see what you can create with the incredible power of modern LLMs.

The 5 Best Open-Source LLMs for AI Companionship

Holodeck Heroes: Building AI Companions for the Final Frontier

My initial encounter with ChatGPT was driven by sheer curiosity when it was released. I was particularly drawn to articles discussing code generation, even though I don't personally write code. And I initially believed that feature wouldn't be relevant to me. However, there came a moment when I found myself working on one UML diagram that I wasn't too eager to create. In my quest to simplify or expedite the process, I turned to various methods. I usually craft UML diagrams using PlantUML code... here, the realization struck me: this is a form of code, and code can be generated. Thus, ChatGPT evolved into a valuable tool for my diagram creation, and now I'm excited to share my experiences with you. A Few Words About UML stands for , and it serves as a language for specifying, visualizing, constructing, and documenting artifacts of software systems. UML draws from concepts in object-oriented programming languages; having development knowledge is beneficial for UML modeling but not mandatory. UML Unified Modeling Language It is primarily designed to describe aspects of software design, but it also offers a set of behavioral diagrams to model business or user requirements. In my role, I engage with IT business processes, specifically focusing on modeling them for analysis, optimization, and translating them into a clear format for developers to implement. When it comes to modeling, I employ various approaches and tools, one of which is UML. Here's the beauty of it — you don't have to be a software engineer to utilize UML. I use it to navigate through business processes implemented with software, but it goes beyond that — it's remarkably versatile and can be applied by various roles at different stages of development. All in all, give it a try; you might just find yourself loving it. Draw vs Code Like any diagram, you can easily draw it — it's simple, convenient, quick, and provides a clear view of the process and the outcome. Numerous graphic tools are available for this purpose, and I do use them. However, more often than not, I choose the "code" approach, thanks to PlantUML. , an open-source modeling language and tool, allows you to generate diagrams based on plain text language. In simpler terms, you can write code in PlantUML to create diagrams. PlantUML I enjoy sketching diagrams in my 'creative mode,' but more often than not, I either convert them into code afterward or write the code directly. There are two main reasons for this: You can insert, move, and delete blocks in your diagrams much more efficiently Changing the order of blocks in drawn diagrams often leads to modifications across the entire diagram, affecting other blocks and links. When working with code, you don't need to concern yourself with the visual aspect; you're dealing with a text, and PlantUML takes care of the graphical representation for you. You can work with and store simple text files instead of dealing with pictures This is particularly beneficial when encountering images of diagrams in documentation. The quality may be compromised. It's impossible to continue working with such diagrams, and they can only be redrawn, not edited or extended. Another significant advantage is that , allowing you to enjoy the benefits of code generation and seamlessly integrate them with UML. I'll share some examples from my personal experience below. ChatGPT understands PlantUML and can generate code You can use any code editors for PluntUML. I personally prefer Visual Studio Code and PlantUML extension in it. Another way is to use online redactors like or others. https://plantuml.com ChatGPT, Write PluntUML Code for Me Example 1 – I Don’t Wanna Start It With a Blank Slate Let’s create a Use Case diagram for a Taxi application. First, I use a prompt: use syntax from https://plantuml.com/ru/use-case-diagram, create a use case diagram for a taxi app ChatGPT generates a code: @startuml
left to right direction
actor "Passenger" as passenger
actor "Driver" as driver
rectangle "Taxi App" {
  usecase "Request Ride" as request
  usecase "Cancel Ride" as cancel
  usecase "Track Ride" as track
  usecase "Payment" as payment

  passenger -- (request)
  passenger -- (cancel)
  passenger -- (track)
  passenger -- (payment)

  driver -- (track)
  driver -- (payment)

  (request) --> (cancel)
  (request) --> (track)
  (track) --> (payment)
}
@enduml Visual view: The idea is fine, but having only four cases seems insufficient. I request to add additional cases and enhance it with some syntax examples, as there are none similar in PlantUML. Prompt: add more cases, you can use different types of links such as include, extends and generalization, syntax: (case1) .> (case2_includes_case1) : <<include>> (parent) <|-- (child) (case1 extends case 2) .> (case2) : <<extends>> Then I ask multiple times to add more cases again and make some specific requests to fix errors. Here is the final outcome from ChatGPT: @startuml
left to right direction
actor "Passenger" as passenger
actor "Driver" as driver
rectangle "Taxi App" {
  usecase "Request Ride" as request
  usecase "Cancel Ride" as cancel
  usecase "Track Ride" as track
  usecase "Payment" as payment
  usecase "Rate Driver" as rate
  usecase "View Ride History" as history
  usecase "Select Ride Type" as selectType
  usecase "Provide Feedback" as feedback
  usecase "View Driver Information" as driverInfo
  usecase "Update Profile" as updateProfile
  usecase "Promo Code" as promoCode
  usecase "Emergency Assistance" as emergencyAssistance

  passenger -- (selectType)
  passenger -- (request)
  passenger -- (cancel)
  passenger -- (track)
  passenger -- (payment)
  passenger -- (rate)
  passenger -- (history)
  passenger -- (feedback)
  passenger -- (driverInfo)
  passenger -- (updateProfile)
  passenger -- (promoCode)
  passenger -- (emergencyAssistance)

  driver -- (track)
  driver -- (payment)
  driver -- (rate)

  (request) ---> (cancel)
  (request) ---> (track)
  (track) ---> (payment)
  (track) ---> (rate)
  (history) --> (payment)
  (rate) --> (payment)
  (rate) ---> (track)
  (cancel) --> (request)
  (selectType) --> (request)
  (feedback) --> (rate)
  (driverInfo) ---> (track)
  (updateProfile) --> (track)
  (promoCode) --> (payment)
  (emergencyAssistance) --> (track)
}
@enduml I like the cases, but not the connections between them. So, I edit that code a bit. Here is the final result: @startuml

actor "Passenger" as passenger
actor "Driver" as driver
rectangle "Taxi App" {
  usecase "Request Ride" as request
  usecase "Cancel Ride" as cancel
  usecase "Track Ride" as track
  usecase "Payment" as payment
  usecase "Rate Driver" as rate
  usecase "View Ride History" as history
  usecase "Select Ride Type" as selectType
  usecase "Provide Feedback" as feedback
  usecase "View Driver Information" as driverInfo
  usecase "Update Profile" as updateProfile
  usecase "Promo Code" as promoCode
  usecase "Emergency Assistance" as emergencyAssistance

  passenger -- (selectType)
  passenger - (request)
  passenger -- (track)
  passenger -- (payment)
  passenger -- (history)
  passenger -- (rate)
  passenger -- (updateProfile)
  passenger -- (emergencyAssistance)

  driver - (request)
  driver - (payment)
  driver - (rate)
  (request) --> (track)
  (track) --> (payment)
  (track) --> (rate)

  (selectType) --> (request)
  (cancel) .> (request) : <<extends>>
  (rate) <.. (feedback)  : <<include>>
  (payment) <..(promoCode) : <<include>>
  (driverInfo) .> (track) : <<extends>>
}
@enduml Certainly, you can incorporate more scenarios. For example, there's no information about user registration and login. Furthermore, you can guide ChatGPT to create a diagram related to your designs. The key is that you have a foundation that you can use. Example 2 – I Need to Check and Enhance a Diagram I have a simple version of a sequence diagram for picture storage solution with a reverse proxy: @startuml
participant Client
boundary "Reverse Proxy" as Proxy
participant Back
participant FileStorage
participant Admin
actor Editor

Editor -> Admin: Add picture
Admin -> FileStorage: Add picture

Client -> Proxy: Get picture
Proxy -> Proxy: Check cache

alt cash is not empty
Proxy -->> Client: Return picture
else cash is empty
Proxy -> FileStorage: Recieve picture
FileStorage --> Proxy: Picture
Proxy -> Proxy: Save to cache
Proxy -->> Client: Return picture
end
@enduml Let’s improve it. My prompt: I'm designing a solution for storing and using images, check this diagram and edit it if necessary ChatGPT suggestion: 1. **Participant Name Clarification:**
    - Consider renaming "Back" to something more descriptive, like "Application Server" or a similar term that better represents the functionality.
2. **Actor Placement:**
    - Move the "Editor" actor to a more standard position, usually at the top.

@startuml
participant Client
actor Editor
boundary "Reverse Proxy" as Proxy
participant ApplicationServer
participant FileStorage
participant Admin

Editor -> Admin: Add picture
Admin -> FileStorage: Add picture

Client -> Proxy: Get picture
Proxy -> Proxy: Check cache

alt cache is not empty
    Proxy -->> Client: Return picture
else cache is empty
    Proxy -> ApplicationServer: Receive picture
    ApplicationServer --> Proxy: Picture
    Proxy -> Proxy: Save to cache
    Proxy -->> Client: Return picture
end
@enduml I don’t like the point about actor placement, so I put it back; everything else is fine. Then I repeated several times a prompt: check this diagram and enhance it if needed and write prompts: show how ApplicationServer get pic from database add focus of control add some corner cases To finalize it, I adjust the focus of control and make some changes to the naming. Here's the final result: @startuml
participant Client
boundary "Reverse Proxy" as Proxy
participant ApplicationServer
database FileStorage
participant Admin
actor Editor

Editor -> Admin: Add picture
activate Admin
deactivate Admin
Admin -> FileStorage: Add picture
activate FileStorage
deactivate FileStorage
Admin -> ApplicationServer: Store metadata
activate ApplicationServer

Client -> Proxy: Get picture
activate Proxy
Proxy -> Proxy: Check cache

alt cache is not empty
  Proxy -->> Client: Return picture
else cache is empty
  Proxy -> ApplicationServer: Receive picture request
  activate FileStorage
  ApplicationServer -> FileStorage: Retrieve picture from database
  alt picture not found in database
    FileStorage --> ApplicationServer: Picture not found
    deactivate FileStorage
    ApplicationServer -> Proxy: Notify picture not found
    Proxy -->> Client: Picture not found
  else
    FileStorage --> ApplicationServer: Picture data
    deactivate FileStorage
    ApplicationServer -> Proxy: Send picture to Proxy
    Proxy -> Proxy: Save to cache
    Proxy -->> Client: Return picture
  end
end

deactivate Proxy
deactivate ApplicationServer
deactivate Admin
deactivate Editor
@enduml Drawbacks First of all, you have to check everything because "ChatGPT can make mistakes." It is not meant to be a replacement for you but rather a tool to streamline and optimize your work. The most common errors are related to PlantUML syntax. Sometimes, it happens in the middle of a conversation, and I have to fix it and get ChatGPT back on the right track. Second, I use a free version of ChatGPT 3.5, and it has ups and downs. Sometimes, it works worse and gives some random answers for the same prompt. Third, ChatGPT understands some diagrams better than others. For example, it does much better with sequence diagrams than use case diagrams. However, despite these limitations, ChatGPT remains a valuable tool and an excellent assistant capable of increasing your efficiency and saving you time. Conclusion I use ChatGPT on a permanent basis in my work. If you're a manager, analyst, developer, tester, or in any other role, you can make use of UML. If that's the case, consider writing and generating PlantUML code — it will save you time and boost your efficiency. I hope my experience will be useful and inspire you. Thank you!

ChatGPT: Your Time-Saving Companion for UML Diagram Generation

চ্যাটজিপিটি: ইউএমএল ডায়াগ্রাম জেনারেশনের জন্য আপনার সময় বাঁচানোর সঙ্গী

ChatGPT: Ihr zeitsparender Begleiter für die Erstellung von UML-Diagrammen

ChatGPT: su compañero que le permitirá ahorrar tiempo en la generación de diagramas UML

ChatGPT : votre compagnon qui vous fait gagner du temps pour la génération de diagrammes UML

चैटजीपीटी: यूएमएल आरेख निर्माण के लिए आपका समय बचाने वाला साथी

ChatGPT: UML 図生成の時間を節約するコンパニオン

ChatGPT: UML 다이어그램 생성을 위한 시간 절약 도우미

ChatGPT: seu companheiro que economiza tempo para geração de diagramas UML

ChatGPT: ваш помощник для экономии времени при создании UML-диаграмм

ChatGPT: UML Diyagramı Oluşturmada Zaman Kazandıran Yardımcınız

ChatGPT: Người bạn đồng hành tiết kiệm thời gian của bạn để tạo sơ đồ UML

ChatGPT：生成 UML 图的省时伴侣

The Mysterious Island by Jules Verne, is part of the HackerNoon Books Series. You can jump to any chapter in this book . Chapter VI here Chapter VI Cyrus Harding stood still, without saying a word. His companions searched in the darkness on the wall, in case the wind should have moved the ladder, and on the ground, thinking that it might have fallen down.... But the ladder had quite disappeared. As to ascertaining if a squall had blown it on the landing-place, half way up, that was impossible in the dark. “If it is a joke,” cried Pencroft, “it is a very stupid one! To come home and find no staircase to go up to your room by—that’s nothing for weary men to laugh at.” Neb could do nothing but cry out “Oh! oh! oh!” “I begin to think that very curious things happen in Lincoln Island!” said Pencroft. “Curious?” replied Gideon Spilett, “not at all, Pencroft, nothing can be more natural. Some one has come during our absence, taken possession of our dwelling and drawn up the ladder.” “Some one,” cried the sailor. “But who?” “Who but the hunter who fired the bullet?” replied the reporter. “Well, if there is any one up there,” replied Pencroft, who began to lose patience, “I will give them a hail, and they must answer.” And in a stentorian voice the sailor gave a prolonged “Halloo!” which was echoed again and again from the cliff and rocks. The settlers listened and they thought they heard a sort of chuckling laugh, of which they could not guess the origin. But no voice replied to Pencroft, who in vain repeated his vigorous shouts. There was something indeed in this to astonish the most apathetic of men, and the settlers were not men of that description. In their situation every incident had its importance, and, certainly, during the seven months which they had spent on the island, they had not before met with anything of so surprising a character. Be that as it may, forgetting their fatigue in the singularity of the event, they remained below Granite House, not knowing what to think, not knowing what to do, questioning each other without any hope of a satisfactory reply, every one starting some supposition each more unlikely than the last. Neb bewailed himself, much disappointed at not being able to get into his kitchen, for the provisions which they had had on their expedition were exhausted, and they had no means of renewing them. “My friends,” at last said Cyrus Harding, “there is only one thing to be done at present; wait for day, and then act according to circumstances. But let us go to the Chimneys. There we shall be under shelter, and if we cannot eat, we can at least sleep.” “But who is it that has played us this cool trick?” again asked Pencroft, unable to make up his mind to retire from the spot. Whoever it was, the only thing practicable was to do as the engineer proposed, to go to the Chimneys and there wait for day. In the meanwhile Top was ordered to mount guard below the windows of Granite House, and when Top received an order he obeyed it without any questioning. The brave dog therefore remained at the foot of the cliff while his master with his companions sought a refuge among the rocks. To say that the settlers, notwithstanding their fatigue, slept well on the sandy floor of the Chimneys would not be true. It was not only that they were extremely anxious to find out the cause of what had happened, whether it was the result of an accident which would be discovered at the return of day, or whether on the contrary it was the work of a human being; but they also had very uncomfortable beds. That could not be helped, however, for in some way or other at that moment their dwelling was occupied, and they could not possibly enter it. Now Granite House was more than their dwelling, it was their warehouse. There were all the stores belonging to the colony, weapons, instruments, tools, ammunition, provisions, etc. To think that all that might be pillaged and that the settlers would have all their work to do over again, fresh weapons and tools to make, was a serious matter. Their uneasiness led one or other of them also to go out every few minutes to see if Top was keeping good watch. Cyrus Harding alone waited with his habitual patience, although his strong mind was exasperated at being confronted with such an inexplicable fact, and he was provoked at himself for allowing a feeling to which he could not give a name, to gain an influence over him. Gideon Spilett shared his feelings in this respect, and the two conversed together in whispers of the inexplicable circumstance which baffled even their intelligence and experience. “It is a joke,” said Pencroft; “it is a trick some one has played us. Well, I don’t like such jokes, and the joker had better look out for himself, if he falls into my hands, I can tell him.” As soon as the first gleam of light appeared in the east, the colonists, suitably armed, repaired to the beach under Granite House. The rising sun now shone on the cliff and they could see the windows, the shutters of which were closed, through the curtains of foliage. All here was in order; but a cry escaped the colonists when they saw that the door, which they had closed on their departure, was now wide open. Some one had entered Granite House—there could be no more doubt about that. The upper ladder, which generally hung from the door to the landing, was in its place, but the lower ladder was drawn up and raised to the threshold. It was evident that the intruders had wished to guard themselves against a surprise. Pencroft hailed again. No reply. “The beggars,” exclaimed the sailor. “There they are sleeping quietly as if they were in their own house. Hallo there, you pirates, brigands, robbers, sons of John Bull!” When Pencroft, being a Yankee, treated any one to the epithet of “son of John Bull,” he considered he had reached the last limits of insult. The sun had now completely risen, and the whole facade of Granite House became illuminated by its rays; but in the interior as well as on the exterior all was quiet and calm. The settlers asked if Granite House was inhabited or not, and yet the position of the ladder was sufficient to show that it was; it was also certain that the inhabitants, whoever they might be, had not been able to escape. But how were they to be got at? Herbert then thought of fastening a cord to an arrow, and shooting the arrow so that it should pass between the first rounds of the ladder which hung from the threshold. By means of the cord they would then be able to draw down the ladder to the ground, and so re-establish the communication between the beach and Granite House. There was evidently nothing else to be done, and, with a little skill, this method might succeed. Very fortunately bows and arrows had been left at the Chimneys, where they also found a quantity of light hibiscus cord. Pencroft fastened this to a well-feathered arrow. Then Herbert fixing it to his bow, took a careful aim for the lower part of the ladder. Cyrus Harding, Gideon Spilett, Pencroft, and Neb drew back, so as to see if anything appeared at the windows. The reporter lifted his gun to his shoulder and covered the door. The bow was bent, the arrow flew, taking the cord with it, and passed between the two last rounds. The operation had succeeded. Herbert immediately seized the end of the cord, but, at that moment when he gave it a pull to bring down the ladder, an arm, thrust suddenly out between the wall and the door, grasped it and dragged it inside Granite House. “The rascals!” shouted the sailor. “If a ball can do anything for you, you shall not have long to wait for it. “But who was it?” asked Neb. “Who was it? Didn’t you see?” “No.” “It was a monkey, a sapajou, an orangoutang, a baboon, a gorilla, a sagoin. Our dwelling has been invaded by monkeys, who climbed up the ladder during our absence.” And, at this moment, as if to bear witness to the truth of the sailor’s words, two or three quadrumana showed themselves at the windows, from which they had pushed back the shutters, and saluted the real proprietors of the place with a thousand hideous grimaces. “I knew that it was only a joke,” cried Pencroft; “but one of the jokers shall pay the penalty for the rest.” So saying, the sailor, raising his piece, took a rapid aim at one of the monkeys and fired. All disappeared, except one who fell mortally wounded on the beach. This monkey, which was of a large size, evidently belonged to the first order of the quadrumana. Whether this was a chimpanzee, an orangoutang, or a gorilla, he took rank among the anthropoid apes, who are so called from their resemblance to the human race. However, Herbert declared it to be an orangoutang. “What a magnificent beast!” cried Neb. “Magnificent, if you like,” replied Pencroft; “but still I do not see how we are to get into our house.” “Herbert is a good marksman,” said the reporter, “and his bow is here. He can try again.” “Why, these apes are so cunning,” returned Pencroft; “they won’t show themselves again at the windows and so we can’t kill them; and when I think of the mischief they may do in the rooms and storehouse—” “Have patience,” replied Harding; “these creatures cannot keep us long at bay.” “I shall not be sure of that till I see them down here,” replied the sailor. “And now, captain, do you know how many dozens of these fellows are up there?” It was difficult to reply to Pencroft, and as for the young boy making another attempt, that was not easy; for the lower part of the ladder had been drawn again into the door, and when another pull was given, the line broke and the ladder remained firm. The case was really perplexing. Pencroft stormed. There was a comic side to the situation, but he did not think it funny at all. It was certain that the settlers would end by reinstating themselves in their domicile and driving out the intruders, but when and how? this is what they were not able to say. Two hours passed, during which the apes took care not to show themselves, but they were still there, and three or four times a nose or a paw was poked out at the door or windows, and was immediately saluted by a gun-shot. “Let us hide ourselves,” at last said the engineer. “Perhaps the apes will think we have gone quite away and will show themselves again. Let Spilett and Herbert conceal themselves behind those rocks and fire on all that may appear.” The engineer’s orders were obeyed, and while the reporter and the lad, the best marksmen in the colony, posted themselves in a good position, but out of the monkeys’ sight, Neb, Pencroft, and Cyrus climbed the plateau and entered the forest in order to kill some game, for it was now time for breakfast and they had no provisions remaining. In half an hour the hunters returned with a few rock pigeons, which they roasted as well as they could. Not an ape had appeared. Gideon Spilett and Herbert went to take their share of the breakfast, leaving Top to watch under the windows. They then, having eaten, returned to their post. Two hours later, their situation was in no degree improved. The quadrumana gave no sign of existence, and it might have been supposed that they had disappeared; but what seemed more probable was that, terrified by the death of one of their companions, and frightened by the noise of the firearms, they had retreated to the back part of the house or probably even into the store-room. And when they thought of the valuables which this storeroom contained, the patience so much recommended by the engineer, fast changed into great irritation, and there certainly was room for it. “Decidedly it is too bad,” said the reporter; “and the worst of it is, there is no way of putting an end to it.” “But we must drive these vagabonds out somehow,” cried the sailor. “We could soon get the better of them, even if there are twenty of the rascals; but for that, we must meet them hand to hand. Come now, is there no way of getting at them?” “Let us try to enter Granite House by the old opening at the lake,” replied the engineer. “Oh!” shouted the sailor, “and I never thought of that.” This was in reality the only way by which to penetrate into Granite House so as to fight with and drive out the intruders. The opening was, it is true, closed up with a wall of cemented stones, which it would be necessary to sacrifice, but that could easily be rebuilt. Fortunately, Cyrus Harding had not as yet effected his project of hiding this opening by raising the waters of the lake, for the operation would then have taken some time. It was already past twelve o’clock, when the colonists, well armed and provided with picks and spades, left the Chimneys, passed beneath the windows of Granite House, after telling Top to remain at his post, and began to ascend the left bank of the Mercy, so as to reach Prospect Heights. But they had not made fifty steps in this direction, when they heard the dog barking furiously. And all rushed down the bank again. Arrived at the turning, they saw that the situation had changed. In fact, the apes, seized with a sudden panic, from some unknown cause, were trying to escape. Two or three ran and clambered from one window to another with the agility of acrobats. They were not even trying to replace the ladder, by which it would have been easy to descend; perhaps in their terror they had forgotten this way of escape. The colonists, now being able to take aim without difficulty, fired. Some, wounded or killed, fell back into the rooms, uttering piercing cries. The rest, throwing themselves out, were dashed to pieces in their fall, and in a few minutes, so far as they knew, there was not a living quadrumana in Granite House. At this moment the ladder was seen to slip over the threshold, then unroll and fall to the ground. “Hullo!” cried the sailor, “this is queer!” “Very strange!” murmured the engineer, leaping first up the ladder. “Take care, captain!” cried Pencroft, “perhaps there are still some of these rascals. “We shall soon see,” replied the engineer, without stopping however. All his companions followed him, and in a minute they had arrived at the threshold. They searched everywhere. There was no one in the rooms nor in the storehouse, which had been respected by the band of quadrumana. “Well now, and the ladder,” cried the sailor; “who can the gentleman have been who sent us that down?” But at that moment a cry was heard, and a great orang, who had hidden himself in the passage, rushed into the room, pursued by Neb. “Ah, the robber!” cried Pencroft. And hatchet in hand, he was about to cleave the head of the animal, when Cyrus Harding seized his arm, saying,— “Spare him, Pencroft.” “Pardon this rascal?” “Yes! it was he who threw us the ladder!” And the engineer said this in such a peculiar voice that it was difficult to know whether he spoke seriously or not. Nevertheless, they threw themselves on the orang, who defended himself gallantly, but was soon overpowered and bound. “There!” said Pencroft. “And what shall we make of him, now we’ve got him?” “A servant!” replied Herbert. The lad was not joking in saying this, for he knew how this intelligent race could be turned to account. The settlers then approached the ape and gazed at it attentively. He belonged to the family of anthropoid apes, of which the facial angle is not much inferior to that of the Australians and Hottentots. It was an orangoutang, and as such, had neither the ferocity of the gorilla, nor the stupidity of the baboon. It is to this family of the anthropoid apes that so many characteristics belong which prove them to be possessed of an almost human intelligence. Employed in houses, they can wait at table, sweep rooms, brush clothes, clean boots, handle a knife, fork, and spoon properly, and even drink wine... doing everything as well as the best servant that ever walked upon two legs. Buffon possessed one of these apes, who served him for a long time as a faithful and zealous servant. The one which had been seized in the hall of Granite House was a great fellow, six feet high, with an admirably poportioned frame, a broad chest, head of a moderate size, the facial angle reaching sixty-five degrees, round skull, projecting nose, skin covered with soft glossy hair, in short, a fine specimen of the anthropoids. His eyes, rather smaller than human eyes, sparkled with intelligence; his white teeth glittered under his mustache, and he wore a little curly brown beard. “A handsome fellow!” said Pencroft; “if we only knew his language, we could talk to him.” “But, master,” said Neb, “are you serious? Are we going to take him as a servant?” “Yes, Neb,” replied the engineer, smiling. “But you must not be jealous.” “And I hope he will make an excellent servant,” added Herbert. “He appears young, and will be easy to educate, and we shall not be obliged to use force to subdue him, nor draw his teeth, as is sometimes done. He will soon grow fond of his masters if they are kind to him.” “And they will be,” replied Pencroft, who had forgotten all his rancor against “the jokers.” Then, approaching the orang,— “Well, old boy!” he asked, “how are you?” The orang replied by a little grunt which did not show any anger. “You wish to join the colony?” again asked the sailor. “You are going to enter the service of Captain Cyrus Harding?” Another respondent grunt was uttered by the ape. “And you will be satisfied with no other wages than your food?” Third affirmative grunt. “This conversation is slightly monotonous,” observed Gideon Spilett. “So much the better,” replied Pencroft; “the best servants are those who talk the least. And then, no wages, do you hear, my boy? We will give you no wages at first, but we will double them afterwards if we are pleased with you.” Thus the colony was increased by a new member. As to his name the sailor begged that in memory of another ape which he had known, he might be called Jupiter, and Jup for short. And so, without more ceremony, Master Jup was installed in Granite House. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Jules Verne (1998). The Mysterious Island. Urbana, Illinois: Project Gutenberg. Retrieved October 2022 https://www.gutenberg.org/files/1268/1268-h/1268-h.htm This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at , located at . www.gutenberg.org https://www.gutenberg.org/policy/license.html

His companions searched in the darkness on the wall

Nowadays, an essential part of the development process is to track the performance of an application. Analyzing the processes in the system allows you to better understand which part of the project is working well, and where urgent fixes are needed. Among the many tools that allow you to monitor and debug, there is one that is very easy to customize in the Laravel framework, because it is made by the creators of the framework themselves. Its name is Telescope. Laravel Telescope is a debugging panel, with a user-friendly and modern design. Inside the panel you can find logs and statuses of all elements of the system. Features Laravel Telescope is a customized debugging panel with different aspects of the application, where each aspect is highlighted in a separate menu item. Telescope allows you to see all parts of the system, such as database accesses, cache accesses, how to work with built-in events, how to work with custom events. Also there is realized the display of requests to the system or errors. We will talk more about the features a little later, but for now let's set up the application. Local and production installation The main convenience of working with Telescope is that it can be configured literally in a couple of clicks. After installing the package you need to load the assets with the command telescope:install , and then run migrations. When installing only locally we need, after adding the package to dev dependencies and loading the assets, to remove the TelescopeServiceProvider connection from config/app.php, but set the provider connection depending on the environment in AppServiceProvider, for example as follows if ($this->app->environment('local')) {
   $this->app->register(\Laravel\Telescope\TelescopeServiceProvider::class);
   $this->app->register(TelescopeServiceProvider::class);
} After that you need to remove the Telescope package from discover in composer.json.
As for the production solution, there can't be one opinion. On the one hand - when installing and on production environment there is no need to finalize the connection, because in this case it will be available everywhere. It is also a plus, as it gives you the opportunity to quickly see what is happening with the state of production.
However, such a solution has a disadvantage and quite a big one.
Due to the fact that the data that is shown in Telescope is extremely sensitive, it is necessary to correctly configure the routing and disable it from being available without access to the corporate network.
In my opinion, the ideal connection would be to access a local Laravel Telescope that accesses all data sources within the corporate network.
The last step in setting up this tool is to set up the configuration in config/telescope.php Let's see how it works By default Telescope runs at /telescope. With a fresh installation, you probably won't see any requests or other data. But what we can say with certainty is that there is a lot of possibilities for this tool, if you look even just at the list of menus. As for the settings for connecting different services, we will talk about it below. In the meantime, it is still necessary to understand how to set up the panel for yourself. Prune command The first problem in our application is the problem with data deletion. Since Telescope collects all the data it sees, the database often quickly reaches its limit without deleting old records. Because of this, we need to update the database by force. Every application has a different frequency of data refresh, but if we're talking about a small project, we can do it on a daily basis. This can be achieved by running the artisan command telescope:prune, or by setting it automatically in Console\Kernel. Accessing the Telescope panel By default, working with the panel is only allowed locally. To fix this, you need to add some modified logic to the Gate check in TelescopeServiceProvider and specify there, for example, your email address. protected function gate(): void
{
   Gate::define('viewTelescope', function ($user) {
       return in_array($user->email, [
           'example@gmail.com'
       ]);
   });
} As you can see in the listing - adding a rule for validation is not a difficult task. Filters From time to time, when working with Telescope, especially in production, there is a problem with disk space. In order to avoid this and not to disable all logs so often, Laravel Telescope provides filters. They allow us to record only important parts, such as errors or dropped jobs. To change this we need to add or remove necessary/unnecessary parts in TelescopeServiceProvider, for example like this: Telescope::filter(function (IncomingEntry $entry) {
   if ($this->app->environment('local')) {
       return true;
   }


   return $entry->isReportableException() ||
          $entry->isFailedRequest() ||
          $entry->isFailedJob() ||
          $entry->hasMonitoredTag();
}); So the filters helped to limit the set of logs when working on production. Tags If you look at the previous listing, there is a method hasMonitoredTag(). It allows you to show all records that have a tag set, regardless of which watcher it belongs to. In order to set a tag on, for example, a query, you will need to add the following to the TelescopeServiceProvider: Telescope::tag(function (IncomingEntry $entry) {
   return $entry->type === 'request'
       ? ['status:'.$entry->content['response_status']]
       : [];
}); All tracked tags can be viewed by clicking on the button in the top right corner of the admin panel interface. Available watchers Finally, we have observed all the formalities and come to the most interesting part. Laravel Telescope contains a huge number of integrations with different services. Let's take a look at the most popular watchers. Adding a watcher to the system is done in the config/telescope configuration file. If you need to add login data to a new watcher, the new data is written directly to the 'watchers' key. The full list is quite large and while some work with obvious entities like CacheWatcher, CommandWatcher, RequestWatcher, others need a bit more detail. BatchWatcher - this watcher keeps track of sending jobs in batches. Read more about this method here EventWatcher and ModelWatcher - it's worth noting that EventWatcher does not track changes to models that use standard model events, unlike ModelWatcher GateWatcher - records all checks of gates and policies and their results. Below, for information, are examples of how some of the pages work. CacheWatcher Tracks cache hits, including unsuccessful ones. You can also see the passed arguments and options DumpWatcher Shows all calls to dump() function It is worth noting that this watcher does not work when using as a cache driver. array EventWatcher Shows all custom events that have occurred recently. You can also track the listener that worked on a particular event. ExceptionWatcher One of the most useful watchers, as it allows you to see errors, including the code where the error is thrown, as well as stacktrace Upgrade Telescope version It cannot be ignored that upgrading the Telescope version can cause some difficulties. To solve them, you need to follow all the steps of the guide on the official website, and make a couple of points. Due to the fact that the published assets may change during the upgrade, you need to upload them again during the upgrade, either manually or by setting post-update-cmd in composer.json Conclusion There are a huge number of all sorts of options on the market right now for debugging and tracking the state of a project. Laravel Telescope may not be very suitable for Production solutions, but for local development the tool is almost indispensable. It is worth noting that Laravel Telescope is updated quite often, which gives hope for long term support.

Companion for working with Laravel — Telescope

Source [Link](https://www.analyticsinsight.net/when-humans-fall-in-love-with-robots-technology-gets-complicated/) Dating in 2023 feels strange. After a 4-year relationship, I've been single for a year. Things have changed since 2018. and raised the bar for commitment. Nowadays, a pretty girl has the whole world potentially peeking into her life, liking her photos, and sliding into her DMs. Instagram Tinder Now, the bar for settling down has soared to new heights. The pressure to stand out and be more captivating feels like it's turned 11. I've given Bumble and OKCupid a shot, but it's disheartening. My matches are few; looks overshadow words there. It's tough out there. But I am not alone. OKCupid did a .). They looked at how men and women rate each other's physical attractiveness. study For men, many women were average, with balanced ratings for more and less attractive. Like a typical distribution, nothing odd. Weird, isn’t it? There are two ways to see it: only less attractive guys are using the app, or women assess men's looks more critically. It's likely the second one. And if that's true, it means women have high standards for looks. Well, 20% of women are seen as really good-looking by guys, but only 2% of guys are seen that way by women. So, most women feel like they're settling for less attractive guys. The competition gets fierce. The top guys have it easy while the rest women fight over them. Dating for women is like shopping. Dating for men is like trying to find life in outer space. -Wheat Waffles And then there's the rest of the guys like me, like 80% of them. Women don't see them the same way, so they're left out. So what option do we have now? Facing these difficulties in dating app dynamics, men like me are increasingly drawn to AI companion apps. However, it's important to understand that AI models cannot truly experience emotions as humans do. Yet, can AI convince us they feel emotions? Certainly, perception matters. AI models can learn to understand and mimic emotions in two ways. Firstly, supervised learning involves training with labeled emotional text and code. For instance, text from books, movies, and social media is tagged with emotions like happiness or sadness. The AI learns patterns and predicts emotions in the new text. Then there's reinforcement learning. The AI gets rewarded for correctly predicting emotions. Through trial and error, it figures out patterns and improves its understanding. It can even learn from audio, video, and physiological data like heart rate. While AI models can simulate emotions through learning patterns, the concept extends further in creating custom AI companions. Now there are tutorials on Github that can help you create your AI girlfriend. You can customize their personalities and backstories, making them potential romantic partners. With customizable personalities, men are likelier to find an AI companion that aligns with their preferences and desires. These AI girlfriends can provide companionship, engage in conversations, and offer emotional support. In the near future, we could see humanoid robots equipped with AI intelligence and resembling human skin. These robots will look like us and provide a lifelike and close experience. As this idea progresses, the thought of marrying such an AI moves beyond mere speculation. If AI can bring happiness and help with loneliness, maybe giving it a shot isn't such a bad idea after all, right?

Beyond Swipes and Likes: The Intriguing Rise of AI-Driven Companionship

Cyrus Harding and his companions slept like innocent marmots

Cyrus Harding and his companions slept like innocent marmots in the cave

Stories, News & Mentions about COMPANION

Categories

Trending Topics

Stories, News & Mentions about COMPANION

Companion’s Stories on HackerNoon

Companion's News & Mentions

Categories

Trending Topics