Home Blog Page 4

Enhancing Research Skills with Workshop on Research Methodology & Article Writing – Universal College

Namaste and welcome to DataSagar. As an educator and tech guy, I believe that sharing knowledge and experiences can go a long way in promoting growth and development in our field. With this in mind, I am excited to share my insights and learning from the recent workshop on Research Methodology and Article Writing organized by Universal College. This post is an effort to not only share my experiences but also to encourage others to participate in such training programs and to continue to learn and grow in their respective fields. Moving forward, I plan to continue this knowledge-sharing initiative by writing more posts on various topics related to research and academia in the future as well.

To begin with, attending the one-day seminar cum workshop on Research Methodology & Article Writing organized by Universal College Today(4/29/2023) was an enlightening experience for me. The workshop was aimed at equipping participants with the knowledge and skills necessary to write compelling research reports and journal articles.

The workshop was delivered by Dr. David Prasad Bashyal, an experienced researcher and scholar who has published numerous research papers and articles in reputable journals. Dr. Bashyal’s insights and practical tips on the research process, data collection, analysis, and reporting were invaluable.

For Your Information,

Dr. David Prasad Basyal is a highly qualified and experienced professional with a Ph.D. in Public Management from Incheon National University (INU), South Korea, an MPhil and MA in Economics from TU, Nepal, and an MBA in Marketing from the University of Gloucestershire (UoG), UK. He has diverse work experiences ranging from education to marketing, sales, HR, medicine, rural development, and more. Dr. Basyal has an extroverted nature and a keen interest in the daily happenings of people, the economy, government, public policy, and business activities. 
His expertise in monitoring and evaluation has been instrumental in his work as a Monitoring and Evaluation Expert at Social Welfare Council Nepal. His knowledge and experience in various fields made his session insightful and highly relevant to the attendees of the workshop. Dr. Basyal's contribution to the workshop was highly appreciated, and we look forward to his future contributions to promoting research and knowledge creation.

During the workshop, we learned about the different types of research and the importance of choosing the right research methodology. We also learned about the various steps involved in the research process, including data collection, analysis, and interpretation.

Dr. Basyal also discussed the prerequisites of research writing, including citation and referencing. He emphasized the importance of selecting a research topic that has significance, novelty, curiosity, scope, and actionability. He also discussed the process of writing the introduction of a research paper, focusing on creating a skeleton and writing it after the completion of the paper. Furthermore, Dr. Basyal discussed the critical elements of the literature review, including the aim, what to include, methods, and more. He also highlighted the importance of selecting target journals and considering the Journal Impact Factor (JIF) when submitting research articles. Overall, the workshop was a valuable experience that equipped attendees with the essential skills and knowledge required for research writing.

The workshop also covered the art of journal article writing, which is a critical component of the research process. Dr. Bashyal provided us with practical tips on how to structure and format research articles, as well as how to write clearly and concisely.

In addition to learning about research methodology and article writing, the workshop also introduced us to several research and citation tools that can help make the research process more efficient and effective. One of the tools we discussed in the workshop was Mendeley, a popular reference management software. Mendeley is a user-friendly and powerful tool that allows researchers to organize their references, generate citations and bibliographies, and collaborate with others on research projects.

Mendeley can be used to store and manage references from different sources, such as academic journals, books, and websites. The software also has a powerful search engine that allows users to quickly and easily find relevant literature for their research projects. Another great feature of Mendeley is its citation generation capability. Mendeley supports various citation styles, including APA, MLA, Chicago, and Harvard, making it easy for researchers to create accurate and consistent citations and bibliographies. During the workshop, we learned how to use Mendeley effectively with MS Word Plugin and how it can help us streamline the research process. By using Mendeley, we can save time and avoid errors when organizing our references and generating citations.

Moreover, The program was focused to promote a research culture among faculties from which our students can get benefited later on. The Universal College publishes a journal every year which is called “The Harvest”. The main objective of the workshop was to enhance and refine research skills among educators and ensure the quality of the articles in “The Harvest” as well.

For your knowledge, The Harvest is a renowned publication of Universal College, affiliated with Tribhuvan University in Nepal, that aims to promote research in multidisciplinary fields among management as well as technology diaspora. This annual publication includes research-based articles from management fields to humanities, social sciences, and technology, etc. which are reviewed by a double-blind peer-review process. The journal follows APA and MLA format systems and includes articles that shed light on research reviews, analyses, and discussions that contribute to the stock of knowledge.

The concluding session of the workshop was marked by the remarks of BBA Coordinator Mr. Sibahari Mainali and the Principal of Universal College, Dr. Shiva Datta Gnawali . They appreciated the efforts of Dr. Basyal and the attendees for making the workshop a success. Participants also shared their experiences and expressed their gratitude for the valuable insights they gained from the workshop. At the end of the program, a Certificate of Participation was awarded to all the attendees, marking the successful completion of the workshop. The workshop was undoubtedly a significant step towards enhancing research skills and promoting research-based activities in the academic community.

These kinds of skill-enhancing programs are sure to enhance faculty members and students develop independent critical thinking skills, expand their understanding, discover and create new knowledge, and disseminate it to others.

Overall, The workshop on Research Methodology & Article Writing, which I attended, was a fantastic opportunity to learn about research methodology and article writing as well as to learn more about research skills and tools such as Mendeley. I am confident that the knowledge and skills I gained will be invaluable as I continue to pursue my research interests in the field of Data Science, Business Intelligence, and Electronic Tourism. I am looking forward to applying my learning and sharing my own research findings with the tech community very soon with an article of my own.

I would like to express my gratitude to Universal College for organizing such a useful and informative workshop, and to Dr. David Prasad Bashyal for sharing his insights and expertise.

Thanks and happy learning.

Microsoft developing its own AI chip Athena to reduce Machine Learning costs — DataSagar

Welcome back to yet another blog post on DataSagar! This post is a continuation of my blog series on Artificial Intelligence. As a technology educator and e-business consultant, I aim to provide insightful articles and helpful tips to keep you up-to-date with the latest developments in the world of tech. In this article, we’ll be discussing Microsoft’s latest project: developing its own artificial intelligence (AI) chip code named “Athena.”

Microsoft is reportedly working on their new chip project called Athena which is supposed to be an AI processor intended to power the technology powering AI chatbots like ChatGPT, which will be used to train the substantial language models required by generative AI. Massive volumes of data are provided by these chatbots, which can also spot trends and imitate human speech. Microsoft expects Athena to outperform the chips it presently purchases from other vendors, saving time and money on its pricey AI research and development.

The market for these chips is presently dominated by Nvidia, but Microsoft aims to change that with Athena. Developed in 2019, the chip is now being tested by a small team of Microsoft and OpenAI staff members. Athena is reportedly being developed by more than 300 employees, and Microsoft anticipates that it will be offered to Azure users by the end of the year.

According to Dylan Patel, principal analyst at SemiAnalysis, “Microsoft wants to use large language models in all of their applications, including Bing, Microsoft 365, and GitHub.” The corporation might save tens of billions of dollars as a result.

As a result of ChatGPT’s success, Microsoft is quickening the rollout of Athena. In an effort to compete with Google and capitalize on its cooperation with OpenAI, Microsoft earlier this year released Bing AI, its own AI-powered search engine.

Microsoft’s upcoming Surface PCs will also come with an NPU (neural processing unit) to speed up on-device artificial intelligence and machine learning (ML), in addition to Athena. Microsoft now employs chipsets with NPUs from AMD, Intel, and Qualcomm, all of which are currently adding them to their microprocessors.

It is obvious that Microsoft is making significant investments in AI, and with the creation of Athena, it aspires to dominate the AI chip industry. Although Nvidia presently rules the market, it is still unclear how Microsoft’s chip will function and whether it will be able to challenge the top chip maker. Whatever the case, AI and technology as a whole are in an exciting phase right now, and we can’t wait to see what the future brings.

The idea of developing AI chips internally is not new; companies like Amazon, Google, and Facebook already produce AI chips. It’s an increasing trend as tech behemoths grasp AI’s potential and how it can completely transform the market. These businesses can save costs on their time- and money-consuming AI endeavors with the development of AI chips.

Additionally, the epidemic has increased technology use and raised demand for AI-powered products and services. In light of this, Microsoft’s investment in AI chips may prove to be a big step forward for the business, allowing it to increase its product offerings and competence in the AI sector.

You should know that ChatGPT and other language models need high-capacity AI devices to do their computations effectively. The size and complexity of the model determine how much processing power is needed.

Modern language generation models like ChatGPT, which are built on the GPT-3 architecture and trained using enormous amounts of data, demand a lot of computing power. Depending on the situation, ChatGPT may consume more or less processing power, although it normally needs strong GPU clusters or specialist AI chips.

For instance, a supercomputer with approximately 3,000 graphics processing units (GPUs) and 192 server-class CPUs was used to train the greatest version of GPT-3, which has 175 billion parameters. Although ChatGPT is a more compact model than GPT-3, it still needs a lot of computing power to function properly.

And yes, whatever it be, Microsoft’s creation of its own AI chip is a positive step forward for the field of artificial intelligence. Microsoft expects that Athena will outperform the chips it presently purchases from other vendors, saving time and resources on its pricey AI research and development. AI is undoubtedly the technology of the future, and by making this investment, Microsoft is establishing itself as a significant player in the AI chip industry. It will be interesting to see how Athena does on the market and how it counters Nvidia’s hegemony.

Can AI Build Another AI System More Advanced Than Itself? Exploring the Possibility and Implications

In recent years, the topic of artificial intelligence (AI), which is quickly developing, has attracted a lot of attention. Being a technopreneur and computer science educator, I’ve been paying great attention to this subject as the AI revolution started to hit the world more strongly than ever. I’ll try to cover whether AI can create a system that is more potent than itself in this blog post.

Firstly, Let’s define AI before putting forward an opinion. The term Artificial Intelligence (AI) can be understood simply as computer programs that can carry out tasks that ordinarily require human intelligence. Content Generation and Translation of languages, speech recognition, decision-making, and visual perception are a few of them. A subset of Artificial Intelligence(AI) called machine learning enables systems to learn from data and enhance their performance over time.

Image by Freepik

So, the main question is, “Can AI thereby develop more potent AI systems?”. I think Yes, that is feasible. AI systems can be trained to create increasingly complex AI systems as they continue to learn and collect more data. Also, Machine learning that is automated is known as AutoML. You might already have heard about this term. A number of IT firms are already employing AutoML to build intricate AI models that can be leveraged to build even more sophisticated systems. This sets up a cycle of exponential growth that might someday result in the development of incredibly potent AI.

Secondly, It’s important to remember, though, that the emergence of more sophisticated AI systems does not portend the extinction of humanity. Despite the fact that AI is capable of performing many activities that people do, it cannot replace human experience and emotions at least in near future. In many businesses, empathy, inventiveness, and interpersonal contact are essential skills that machines cannot take the place of.

In addition, massive volumes of data are needed for AI systems to learn and advance. Humans must contribute this data, and restrictions must be put in place to guarantee that AI systems are operating properly. In order to guarantee that AI is used ethically and responsibly, human oversight is essential.

There is a chance that AI could create more sophisticated AI systems in the future. However, it’s imperative to go slowly and make sure that AI is used in an ethical and responsible manner. Humans will still be necessary for giving data and supervising AI systems as they develop. It’s more likely that AI will be utilized to boost productivity and efficiency across a range of industries than it will be to completely replace people.

At last, What do you think about AI and the possibilities for developing more sophisticated AI systems? Comment below with your thoughts, and don’t forget to share this article with anyone you think would be interested in the subject. Have a good day!

Twitter “no longer exists” as a result of the merger with X Corp

Well, folks, it looks like Twitter is no more! That’s right, the beloved social media platform has gone through a bit of an identity crisis and decided to merge with a new shell firm called X Corp. It’s kind of like when your friend gets a new boyfriend/girlfriend and suddenly start going by a different name – except in this case, it’s a billion-dollar company and we’re all left wondering what the heck is going on. It’s almost too funny to say that Twitter has become X, but here we are! haha!

Elon Musk’s intentions for the social media platform have been the subject of rumors following the recent merging of Twitter and the newly created X Corp. Twitter “no longer exists” as a result of the merger with X Corp, claims a document filed in a California court for a case brought by conservative activist Laura Loomer against the company and its former CEO Jack Dorsey. The document is getting attention on social media platforms since yesterday.

Although the effects of this move are not yet evident, it is important to remember that Twitter has experienced considerable adjustments since Musk purchased the business for $44 billion last year. The billionaire has previously made suggestions that Twitter might become a “everything app” under the X name, perhaps competing with Chinese social media giant WeChat.

Despite the absence of official statements, there are rumors that this merger may mark the beginning of the development of the X app. It is also conceivable that this action may lead to the creation of a larger parent organization called X that would govern all of Musk’s ventures, including Tesla and SpaceX.

It’s interesting to note that Twitter is now headquartered in Nevada rather than Delaware under X Corp. However, the new X Corp will inherit all of the debts and lawsuits that Twitter and Musk faced before to the merger.

The fact that Elon Musk now owns the “X.com” domain, which he previously used for a payment service that was incorporated into PayPal, fuels even more rumors about what he has in store for the X name.

On Tuesday, Musk tweeted about the change with just the letter “X.” as well.

It is crucial to remember that Twitter is still a fully functional social media network, but debate still surrounds the types of content and individuals it allows and disallows. Many members of the IT community are turning to other platforms for safety as a result as well.

Concerns regarding the future of the social media platform have been sparked by the merging of Twitter and X Corp. Even while it’s yet unclear whether this merger will result in the development of the X app or whether it’s simply a part of a bigger reorganization of Musk’s properties, it’s obvious that it’s a big deal for Twitter and its future.

Threats and attacks in Network Environment

A critical component of the contemporary digital environment is network security. The prevalence of internet-connected gadgets and cloud computing has increased the frequency of network security risks and attacks. The infrastructure, data, and reputation of a business could all suffer severe harm as a result of these risks and attacks. We will talk about many dangers and attacks in a networked environment one-by-one in this article.

Malware, often known as malicious software, is a category of a program intended to harm, interfere with, or allow unauthorized access to a computer system. Malware is one sort of attack. Malicious software, including viruses, trojan horses, worms, and spyware, can spread throughout a network and infect numerous computers, seriously jeopardizing the security of the entire system. This means Malware can take many different shapes, including viruses, trojan horses, worms, and spyware. These dangerous applications have the potential to proliferate throughout the network and infect several systems, seriously jeopardizing network security.

Another danger is phishing, which employs deceptive strategies to coerce people into disclosing personal data like passwords and usernames. These attacks may take the shape of emails, pop-up windows, or phony websites made to look like the real thing. In another word, Phishing is a type of social engineering attack that uses dubious methods to get people to divulge private information like usernames and passwords. These assaults frequently take the shape of emails, pop-up windows, or phony websites that imitate trustworthy ones. Once the attacker obtains the data, they can access the network and jeopardize the security of the system.

Network security can also be compromised by denial-of-service (DoS) and distributed denial-of-service (DDoS) attacks. These attacks are intended to flood a network or a particular system with traffic, rendering it inaccessible to authorized users. DoS and DDoS assaults have the potential to seriously disrupt enterprises and result in large financial losses.

Attacks known as “man-in-the-middle” (MitM) occur when an attacker eavesdrops on a conversation between two parties in order to access and alter the data being transferred. Attackers can insert harmful code into databases through SQL injection attacks, which target poorly protected databases.

Attacks using passwords are another danger. Automated password attacks are possible and can use a variety of methods, including brute force, dictionary, and rainbow table attacks. An attacker can access the network and compromise its security once they have the password.

Last but not least, insider threats can be deliberate or accidental. When a worker or contractor with access to the network harmed the network purposefully or accidentally, it is referred to insider threat. Data theft, sabotage, and illegal network access are some examples of insider risks.

Safeguarding the security of electronic commerce networks requires an understanding of the various threats and assaults. Networks can be protected from potential threats and attacks by implementing the right security measures, such as firewalls, intrusion detection systems, and anti-malware software.

How to enable Microsoft Loop Application from Microsoft 365 Admin Panel

Namaste and Greetings to everyone. I’m here with yet another blog post on a very common problem that your educational institution might be facing while trying to use a productivity app called Microsoft Loop.

If you’re receiving this kind of message while trying to log into Microsoft Loop then, this blog is absolutely for you. If you’re not of the kind, also don’t worry. You’ll be learning an awesome productivity tool from Microsoft for Microsoft 365 users.

Image: datasagar

To begin with,

FYI, Microsoft Loop is a brand-new productivity tool that enables flexible and visual work organization and collaboration. You can make and manage notes, tasks, and documents all in one location using Loop. We’ll go over how to activate the Microsoft Loop application in Microsoft 365 in this article.

Step 1: Verify that Microsoft Loop is accessible through your Microsoft 365 account.
Make sure your Microsoft 365 account has access to Loop before you enable the Microsoft Loop application. Only a select group of subscribers, including Microsoft 365 Enterprise and Education customers, have access to Loop at this time. Consult your IT administrator if you’re not sure if your account has access to Loop.

Step 2: Log in to Microsoft 365
You must sign in to your Microsoft 365 account in order to enable Microsoft Loop. Enter your email address and password on the Microsoft 365 login page at https://login.microsoftonline.com.

Step 3: Access the Microsoft 365 Admin center
To access the Microsoft 365 Admin center, click the Admin app after logging into Microsoft 365. This application is normally found in the screen’s upper left corner, in the app launcher.

Step 4: Turn on Microsoft Loop for your business.
Click Settings > Org settings in the Microsoft 365 Admin center. Click Microsoft Loop under the Services section after scrolling down. Toggle the option to enable Microsoft Loop for your company after that.

Step 5: Access Microsoft Loop
Microsoft Loop may be accessed after it has been activated for your company by heading to the Microsoft 365 app launcher and clicking on the Loop icon. A new tab will open with the Loop app.

Step 6: You’re done! Start utilizing Microsoft Loop
You can begin using the Microsoft Loop app to arrange your work as soon as it has been launched. You can import existing notes, tasks, and documents from other Microsoft 365 programs like Word and Excel, or you can create new ones. By sharing your Loop workspaces with others and giving team members assignments, you can collaborate with them as well.

ChatGPT’s VSCode Extension – Write Code with Natural Language Instructions

The capacity of ChatGPT, a sizable language model developed by OpenAI, to produce text that resembles human speech has made it more and more well-liked. With recent improvisation in ChatGPT 4, it can support many different languages, including Nepali as well. For your information, The addition of Nepali language support in ChatGPT 4 enables Nepali users to use it for a range of tasks, including text generation, question-and-answer translation, and text generation in Nepali. This can be very helpful for people or companies operating in Nepal or dealing with clients or customers that speak Nepali. With wonderful features and capabilities being advanced day by day, the OpenAI team is doing a wonderful job.

One remarkable application of ChatGPT is a VS Code addon that enables programmers to produce full blocks of code and code snippets using natural language instructions.

I’ll try to highlight the capabilities and advantages of the ChatGPT VS Code addon in this article and discuss how it might enhance your development experience along with steps to install it.

The ChatGPT VS Code Extension’s features

A number of features offered by the ChatGPT VS Code addon enable programmers to create code in response to orders given in natural language. Some of the extension’s main characteristics are listed below:

  1. Code generation utilizing natural language commands is possible with the ChatGPT VS Code addon. For instance, the extension will generate the code for the function when a developer types “make a function that adds two numbers”.
  2. The add-on offers a variety of code snippets that are simple to put into your code. These code snippets, which are intended to save developers time and effort, are available in a wide variety of programming languages.
  3. Based on the context of your code, the addon also offers code completion ideas. Developers may write code more quickly and with fewer errors as a result.
  4. The plugin can identify flaws in your code and offer recommendations for how to solve them. Developers may spend less time and effort troubleshooting their code as a result.
  5. You can modify the extension to meet your needs as well. You have a variety of options, including the programming languages you want to use and the code samples you want to utilize.

The ChatGPT VS Code Extension’s advantages

Developers can gain from the ChatGPT VS Code addon in a number of ways. Here are a few of the main advantages:

  1. Time-saving: By creating code snippets and completing code blocks, the addon can help developers save time. Developers may write code more quickly and with fewer errors as a result.
  2. Error Prevention: By offering advice on how to resolve them, the plugin can assist in preventing errors in your code. Developers may spend less time and effort troubleshooting their code as a result.
  3. Productivity Enhancement: By offering code completion recommendations and mistake detection, the addon can assist in increasing developer productivity.
  4. Accessibility: Those who are unfamiliar with programming languages may find coding to be more approachable thanks to the extension. This may contribute to the democratization and inclusivity of coding.
  5. Customization: The extension can be tailored to your preferences, which can enhance and make coding more pleasurable for you.

The ChatGPT VS Code addon is, all things considered, a useful tool for developers who wish to increase the quality of their software. We’ll now go towards How To Install ChatGPT extension in VS Code.

Steps to install ChatGPT Extension in VS Code

Image: www.DataSagar.com
  1. On your computer, launch Visual Studio Code (VS Code).
  2. On the VS Code window’s left side, select the “Extensions” icon (it looks like four squares).
  3. Type “ChatGPT” into the search box at the top of the window.
  4. Search for and choose the “ChatGPT” extension from the search results.
  5. Go to the extension page and click the “Install” button.
  6. Wait the extension’s installation. To enable the addon, VS Code might need to be reloaded.

After installing the extension, you can use it by entering natural language commands in the VS Code editor to begin utilizing it.

Test your installation with Ctrl+Shift+A shortcut and You may be asked to enter Open AI API key to store your sessions detail only. You can get Open AI API for ChatGPT VS Code Extension by going to this link Account API Keys – OpenAI API

By selecting “Preferences: Open User Settings” and looking up “ChatGPT,” you can further personalize the extension’s settings. The accessible programming languages and code snippets in the extension can then be modified from there.

Congratulations, You’re now set to experience ChatGPT with your VS Code, a strong tool that can assist you in writing code using natural language instructions. A variety of capabilities offered by the extension can help developers like you save time and effort and work more efficiently. Additionally, by enabling those who are not conversant in programming languages to write code, the extension can increase accessibility and inclusivity in the field of coding.

Microsoft vs Google – Who’ll win the Generative AI Race?

Many tech behemoths, like Microsoft and Google, have made artificial intelligence a significant area of research. These businesses have made significant investments in the creation of generative AI models that can generate creative media such as music, films, and photos. The competition between Microsoft and Google to build generative AI will be discussed in this article, as well as how each company’s methods have significantly advanced the area.

On the basis of patterns and data inputs, generative AI models are intended to produce creative content. These models create new content based on patterns they find and repeat using deep learning techniques. Since they have been used to produce everything from virtual art to music, generative AI models have grown in popularity.

Research on Generative AI at Microsoft

Microsoft has been at the forefront of generative AI research, with their research team publishing multiple papers on the topic. Microsoft’s DeepDream algorithm, which creates surreal, hallucinogenic pictures using neural networks, is one of the company’s most well-known generative AI initiatives. Microsoft also created the GPT-3 language model, which can produce writing that is grammatically accurate and cohesive.

Microsoft’s Image Captioning Algorithm, which use deep learning to evaluate images and produce insightful captions, is another noteworthy effort. Also, Microsoft researchers created an AI system that can create accurate 3D models of goods from a single image, which has the potential to enhance product display in e-commerce.

Research on Generative AI at Google

The Google Brain research team has been working hard to understand generative AI models, and they have published multiple papers on the subject. The Magenta project at Google, which focuses on employing AI algorithms to produce music, is one of the company’s most noteworthy generative AI initiatives. Magenta has created some original works that have received attention for their excellence and originality.

Google has developed the DeepDream algorithm, which is similar to Microsoft’s algorithm in that it uses neural networks to generate strange, hallucinogenic pictures. Also, the research group at Google created the WaveNet generative model, which can produce music and speech that seem realistic.

Competition between Google and Microsoft

Much progress has been made in the field of generative AI as a result of the rivalry between Microsoft and Google. These businesses make significant investments in R&D and are continually pushing the limits of what is achievable with generative AI models.

Microsoft and Google are approaching generative AI research from various angles, with Microsoft concentrating on language models and picture processing and Google concentrating on music and speech. Yet, both businesses are creating novel algorithms and models that are expanding the realm of generative AI’s capabilities.

Who will win the race then?

Both Google and Microsoft have made important strides in generative AI, and both firms have developed models that are well-respected, such as Google’s GPT model and Microsoft’s Turing-NLG model. Together with Google and Microsoft, these companies have created models like BigGAN and DALL-E that can produce high-resolution images from textual descriptions.

Both businesses have made progress in recent years in creating AI systems that can communicate with people in normal language, with Microsoft’s Xiaoice and Google’s Meena serving as famous examples.

Therefore, it’s tough to proclaim a clear winner in the race for generative AI, as both companies have made remarkable advancements in this discipline. It’s important to keep in mind that there are a lot of other businesses and groups working on generative AI, and the subject is always changing, so the current situation may change in the future.