DEV Community: Hasindu Nagolla

Artificial General Intelligence (AGI)

Hasindu Nagolla — Thu, 10 Sep 2026 01:12:13 +0000

At the moment, the significance of Artificial Intelligence grows steadily. We use AI in a wide variety of spheres, including mobile devices, web sites, education, healthcare, software development and so on. Still, modern AI systems are mostly used for specific purposes. The concept of AGI implies creating an AI system which will be able to learn and solve different tasks in the same manner as a human.

What is AGI?

AGI is the concept of AI system which is able to demonstrate general intelligence, i.e. understand information, learn new things, solve various problems and apply the acquired knowledge to different situations.

For example, one system can be very good at image recognition. Other AI can be great at writing texts or translating languages. All of these systems are focused on their own purposes. In contrast, AGI is focused not on one task, but on multiple types of tasks and usage of this knowledge in different contexts.

How is AGI different from normal AI?

The difference lies in ability to deal with different types of problems.

Modern AI systems are called narrow AI. Their purposes are predetermined – for example, a recommendation system is focused on suggesting movies to users, voice assistants understand speech, chess AI plays chess.

The efficiency of these systems is very high. But unlike a person, they do not possess the same degree of general ability. A chess AI can play chess perfectly. But it cannot learn suddenly to cook or repair a computer.

AGI is aimed to acquire a much broader set of abilities. It is able to learn different topics and apply this knowledge for solving different types of problems.

What can AGI do?

AGI can do a lot of types of work. For example, it can understand the human language, create program code, solve mathematical problems, analyze the information and support the research process.

For example, an AGI system can get information about a software project and help to develop it. After that, it can learn about some scientific problem and help to find possible solutions.

Moreover, AGI is able to learn something new. This is one of the key concepts of AGI. It is not bound by one strict set of instructions.

Benefits of AGI

There are many benefits which AGI can bring to people.

In education, AGI can help students in studying different subjects. It can explain difficult topics in a simple way and provide help depending on the needs of each student.

In healthcare, AGI can help doctors and researchers to analyze huge amount of information. It can also support medical researches.

In software engineering, AGI can help developers in writing, testing and improving the software. It can analyze a problem and help to find a solution.

AGI can also help in scientific researches. It can analyze the information and assist researchers in working with complex problems.

Risks of AGI

But AGI brings some serious risks too. A powerful AI system can cause serious problems if it will be used in harmful purposes.

Firstly, we should speak about jobs. As AGI will be able to perform many types of work, some jobs can change and even become unnecessary. And people will need to learn new skills so that they could cooperate with AI.

Also there are risks connected with safety of AGI. We should make sure that AGI systems are pursuing human goals and not causing any harm.

Also there are risks connected with privacy, security and misinformation. For this reason, AGI development requires very thorough planning and strict safety measures.

Is AGI available today?

AGI is mainly the topic of current research. There is no common definition of AGI yet, and there is no test which would clearly show that some AI system is AGI.

But currently existing AI systems are becoming increasingly capable. They can write texts, generate images, understand program code and perform many other tasks. But this does not necessarily mean that these systems are AGI.

Researchers continue studying the ways to create AGI – systems with increased general ability to learn and think.

Future of AGI

AGI can have a tremendous impact on the future. It can influence our work, learning and solving of problems.

AGI can assist humans in complicated tasks and help in spheres like science, education, medicine and technologies.

But creation of AGI is not only about increasing of AI's power. Safety is very important issue here. Researchers need to make sure that AGI systems are safe, reliable and useful for people.

Conclusion

AGI is one of the most promising directions in artificial intelligence. It implies creating of an AI system which can learn, understand, think and solve many different problems.

While normal AI systems are mainly oriented on specific tasks, AGI implies general intelligence. It provides many useful opportunities but causes many risks too.

AGI is currently being studied and there are many questions about its development. AGI will probably become an important technology in the future, so it is important to understand both its benefits and risks.

AWS vs DigitalOcean, Which Cloud Platform Should You Choose?

Hasindu Nagolla — Tue, 08 Sep 2026 03:07:05 +0000

Cloud computing has become an essential part of modern software development. Whether you are deploying a personal project, hosting a startup application, building a machine learning system or managing enterprise infrastructure, choosing the correct cloud provider has a major impact on cost, scalability, complexity and development speed.

Two platforms that developers frequently compare are Amazon Web Services (AWS) and DigitalOcean. Both can host applications, databases, APIs, containers and virtual machines. However, they prioritise different things. AWS focuses on breadth, scalability and enterprise grade cloud infrastructure. DigitalOcean focuses on simplicity, predictable pricing and developer experience.**** So, which one should you choose? The answer depends on what you are building.

AWS - The Giant Cloud Ecosystem

Amazon Web Services is the largest cloud platform in the world. Instead of being simply a server hosting provider, AWS provides a huge ecosystem of cloud services covering compute, storage, databases, networking, security, analytics, AI, serverless computing, containers, DevOps and so many things.

For example, a typical AWS architecture could combine,
EC2 for virtual machines
S3 for object storage
RDS for managed databases
Lambda for serverless functions
CloudFront for content delivery
Route 53 for DNS
Elastic Load Balancing for traffic distribution
EKS for Kubernetes
CloudWatch for monitoring
IAM for access management

This big ecosystem is one of AWS's biggest advantages. However, it can also be one of its biggest disadvantages for beginners. There are often multiple ways to solve the same problem and understanding how different AWS services interact can take considerable time.

Where is AWS great?

AWS is particularly attractive when you need something like,

1. Large-scale infrastructure
2. Advanced networking
3. Global deployment
4. Enterprise security and compliance
5. Complex architectures
6. Serverless applications
7. Large databases and data pipelines
8. Advanced AI/ML infrastructure
9. Integration with many managed services

In other words, AWS gives developers an enormous toolbox. The challenge is learning which tool to use.

DigitalOcean - Simplicity First

DigitalOcean takes a different way. Instead of presenting developers with a large ecosystem from the beginning, DigitalOcean focuses on making common cloud tasks simple.

Its virtual machines are called Droplets. A developer can select an operating system, choose a Droplet configuration, select a region and deploy a server within minutes.

DigitalOcean currently offers Droplets starting at $4/month, with bundled resources such as CPU, memory, SSD storage and bandwidth.

DigitalOcean also includes a considerable amount of outbound bandwidth with Droplet plans. It makes its pricing easier to understand for many small applications. Another important change is that DigitalOcean moved Droplets to per second billing in January 2026, with a 60 second minimum, giving developers more granular control over short lived workloads.

Pricing of AWS vs DigitalOcean

Pricing is one of the biggest differences between the two platforms. AWS generally uses a highly granular pricing model. You can pay separately for compute, storage, data transfer, IP addresses, requests, databases and other resources depending on your architecture. This flexibility is powerful, but it can make estimating the final bill more complicated.

DigitalOcean's approach is generally easier to understand. For example, a Droplet plan bundles CPU, RAM, storage and a specified amount of bandwidth into a predictable price. DigitalOcean's own comparison shows how bundled resources can make its basic compute offerings cheaper than comparable AWS configurations in some scenarios.

However, cheaper does not automatically mean better. AWS provides a larger set of managed services and enterprise capabilities. So comparing only price of a virtual machine does not tell the whole story.

Ease of Use

This is where DigitalOcean has a major advantage. Imagine you are a university student who wants to deploy

React/Next.js → Node.js API → PostgreSQL

On DigitalOcean, you can get a virtual machine, configure your environment, install Docker, deploy your application and manage everything through a relatively simple interface.

AWS can absolutely handle the same application. But AWS gives you many more architectural choices. You could use,

EC2 → RDS → S3 → CloudFront → Route 53 → Load Balancer → IAM → CloudWatch

That flexibility is extremely useful in production environments. But for someone who simply wants to deploy a project, it can introduce unnecessary complexity.

Scalability

AWS wins when the discussion moves to massive scale and complex architectures. AWS provides infrastructure designed for applications that may need to scale across regions, availability zones, services and millions of users.

DigitalOcean can also scale applications using multiple Droplets, load balancers, Kubernetes, databases and other services. However, AWS generally provides a deeper ecosystem for highly complex enterprise architectures. Therefore,

Small and medium applications - DigitalOcean can be extremely convenient
Large and complex enterprise systems - AWS is often the stronger choice

Security

Both platforms provide security features, but their approaches differ. AWS provides a very extensive security ecosystem. Including IAM, VPC, Security Groups, KMS, WAF, Shield, GuardDuty, CloudTrail and Security Hub. This gives organisations to detailed control over identity, networking, monitoring and compliance.

DigitalOcean provides simpler security features such as Cloud Firewalls, SSH-key authentication, private networking and other infrastructure controls. For a developer deploying a portfolio application or startup MVP, DigitalOcean's simpler approach may be enough.

For organizations with complex compliance, identity and security requirements, AWS provides considerably more depth.

What About AI and Machine Learning?

This is particularly interesting for developers working in AI/ML. mAWS has a huge AI/ML ecosystem, including services for model training, inference, data processing, GPU computing and managed machine learning workflows.

DigitalOcean has also expanded significantly into AI/ML. Its current platform includes GPU Droplets and AI/ML services. DigitalOcean's GPU offerings support workloads such as model training, inference, large-scale data processing and deep learning.

For someone experimenting with an AI application, the choice will depend more on the specific workload than the provider name. For example, for a simple AI API + backend, DigitalOcean can be attractive. For a Large ML pipeline + data lake + distributed training + enterprise infrastructure, AWS becomes better.

Global Infrastructure

AWS has a massive global infrastructure footprint and provides extensive options for deploying applications close to users around the world. DigitalOcean also operates data centers across multiple regions, including Singapore, Bangalore, London, Amsterdam, Frankfurt, New York, Toronto, Sydney and others.

For developers serving users in South Asia or Southeast Asia, choosing a nearby region can significantly improve latency. This means that instead of considering, "Which provider is better?" we need to consider "Which provider has right infrastructure for my users?"

Which One Should You Choose?

There is no universal winner. If you are building your first cloud project, deploying a portfolio website, hosting a small backend, experimenting with Docker or launching an MVP, DigitalOcean is often the easier starting point. You can focus on your application instead of spending a significant amount of time learning cloud infrastructure.

If you are building a large scale production platform, need advanced networking, require extensive managed services or expect complex global infrastructure, AWS is the more powerful choice.

My Suggestion

Learn both. For software engineers and DevOps engineers, understanding DigitalOcean will teach the fundamentals of Linux, networking, SSH, Docker, Reverse Proxies, Databases and CI/CD Deployment. After that, those concepts provide a clear foundation for moving to and understanding AWS much easier.

AWS then adds another layer to our foundations (IAM, VPC, Load Balancing, Auto Scaling, Serverless, Managed Services, Cloud Architecture)
I do not think the real question is AWS or DigitalOcean. The better question should be "What does my application actually need?"
If your application needs a simple server with predictable costs, DigitalOcean is an excellent choice. If your application needs a huge ecosystem of managed services and enterprise scale infrastructure, AWS is the best.

For students and early career developers, I would recommend learning the fundamentals on a simpler platform first and then expanding into AWS. Because ultimately, cloud engineering is not about memorizing cloud provider services. It is about understanding infrastructure.

Once you understand how servers, networking, containers, databases, storage, security, monitoring and deployment work, switching between cloud providers becomes much easier. And that is the real skill worth learning.

Understanding the Transformer Model in LLMs

Hasindu Nagolla — Fri, 04 Sep 2026 15:54:00 +0000

Introduction

When you are using ChatGPT, Google Gemini, Claude or any other modern AI Language Model, you are interacting with a transformer. The transformer architecture introduced in 2017 by Google researchers revolutionized the ability of computers to understand and generate human language. This article will walk you through what transformers are and how they work step-by-step. Before transformers, language models looked at one word at a time. It was like reading a book with a flashlight that only shows one word. This made it hard for the model to understand long sentences. As an Example, In the sentence "The bank was steep," the word "bank" could mean a financial bank or a riverbank. To choose the right meaning, the model needs to see the word "steep". Transformers fix this by looking at all words in a sentence at once. This helps the model see how each word connects to the others

Step 01 - Breaking Text Into Tokens

Before the transformer reads the text, it splits it into smaller units called tokens. Each token becomes a number (a vector) that the model understands. This is called tokenization and embedding.

Input - "I love artificial intelligence"
Tokens - ["I", "love", "art", "ificial", "intell", "igence"]

Step 02 - Adding Positional Information

Transformers look at all words at once, so they need to know the order. For example, "dog bites man" is very different from "man bites dog." Positional encoding adds order to the tokens. It's like giving page numbers to mixed-up book pages.

Step 03 - Attention Mechanism

This is the key part of transformers. Attention helps the model focus on important words when reading a sentence. Let's try to understand how Attention Works. Let's assume we need to translate "Le chat noir" (the black cat) into English. For this, the model needs to focus on "Le" and "noir" when working on "chat." The model uses three parts for this

Query - What word are we looking at now?
Key - What other words are there?
Value - What meaning do they give?

The model gives attention scores to all words to decide which ones are more important. Transformers use many attention heads at the same time (Multi-Head Attention). Each head looks at different things like grammar, meaning or logic. This helps the model understand better.

Step 04 - Feed-Forward Neural Networks

After attention, the model sends each word's information through a feed-forward neural network. This helps the model learn more and improve its guesses. It’s like asking, “Now that I understand the sentence better, now what else can I figure out about this word?”

Step 05 - Stacking Layers

Transformers use many layers, not just one. Each layer improves the output of the one before it.

Layer 1 - Finds simple word links
Layer 5 - Understands grammar better
Layer 10 - Handles deep meaning
Layer 20+ - Deals with complex ideas and reasoning

It’s like editing a paper many times to make it a better and clearer output.

Step 07 - Training the Model

Transformers learn by reading lots of text. During training:

The model reads billions of sentences
It guesses the next word
If wrong, it makes small changes to improve
It repeats this many times

This helps the model learn grammar, facts, patterns, and some logic.

Step 8 - How It Writes Text

When you ask a question:

Your text becomes tokens
Tokens go through all the layers
Attention helps the model find meaning
It guesses the next word
Adds it to the answer

This will repeat until the answer is complete

Conclusion

Transformers are more powerful because they can read all words at once, understand long texts well, work better with more data and can be used for many tasks. Transformers changed how AI understands language by focusing on what matters and using many layers of learning. It became the core of most modern language models. They are not magic, but they are clever tools that learn from huge amounts of text and help computers understand and create human language better than ever before.

Cyber war 😂

Hasindu Nagolla — Fri, 04 Sep 2026 04:56:53 +0000

Don't do that

What is the RAG (Retrieval-Augmented Generation) 🔥😳

Hasindu Nagolla — Fri, 04 Sep 2026 04:49:12 +0000

Think, you are writing an exam. We usually write exams based on what we have memorized. But RAG is like an open book exam. It means we will write the exam using the resources on time, not based on our memory. just like writing answers to the question using a book.

AI models like ChatGPT (actually LLMs) can answer based on their training data set. They cannot provide answers beyond their training data. After integrating the RAG technology, AI models can provide answers by accessing the new outside databases or documents. It will be more accurate than the common answers that return from the LLMs.

Why RAG? 🤔🤔

AI models (LLMs) like ChatGPT and Gemini are brilliant, but they have some limitations,

Outdated Data - they trained on past data, because they have no idea about on time information. So they do predict. (Knowledge Cutoff)
Hallucinations - sometimes they provide random answers instead of the correct answer.
Personal Data - general AI models have no idea about your personal information, like private details or company information.

We can completely mitigate these issues using RAG.

How does it work?

This entire process is done through 4 steps,

Asking a question - think you are asking "can i get a leave?"
Retrieve (search) - The AI model (RAG integrated) does not provide the answer directly; it will search your company's HR policies and regulations.
Augmented - then the system will collect your problem and the company information and send it to the AL model.
Generation - now, the AI model is providing an accurate answer after analysing your question with your company details.

Advantages of RAG, 😮😮

👉Low cost - no need to train an AI model from scratch
👉Real-time information - if you ask for real time news, you can get answers if they exist in the databases
👉Stopping hallucinations - because the AI model will generate answers based on provided real information.
👉Sources - The AI model can highlight references based on the provided documents or databases.

As a summary, RAG(Retrieval-Augmented Generation) is like a bridge between your data and a separate AI model like ChatGPT. We can get highly accurate answers from AI models by providing our information as an outside data source. 👍👍

Artificial Intelligence

Hasindu Nagolla — Fri, 14 Aug 2026 08:45:25 +0000

AI stands for Artificial Intelligence. It is the ability of a computer, or a robot controlled by a computer to mimic human Intelligence.AI is rapidly changing the world.

Subfields of artificial intelligence (AI)

Machine learning

Machine learning is a type of AI that allows machines to learn from data without being explicitly programmed.

Deep learning

Deep learning is a type of machine learning that uses artificial neural networks to learn from data.

Natural language processing

Natural language processing is a type of AI that allows machines to understand and process human language.

Computer vision

Computer vision is a type of AI that allows machines to see and understand the world around them.

Nowadays, AI is very familiar to most people. It is in everything we use today. for example, automated tasks, AI chatbots, image searching, generate content.

Automating tasks - AI can be used to automate tasks. such as post-scheduling, managing comments and responding to emails.
Generating content - AI can be used to generate content. such as blog posts, articles and social media posts.
AI chatbots - AI can be used to replace customer care centres. such as checking customer issues and auto-reply. (Dialog dia AI chat bot)
Image searching - Ai can be used to search images needed by users. example, google image search.

here are some examples of using Artificial Intelligence,

Self-driving cars
Virtual assistants (Apple Siri, Alexa, Google Assistant)
Spam filter
finance
healthcare

Conclusion

AI is still in its early stages of development, but it has the potential to revolutionize the industry. As AI technology continues to improve, we can expect to see more innovative ways to use AI in industry.

Building a Lag-Free Telegram Music Bot for Concurrent Streams in Python

Hasindu Nagolla — Fri, 10 Jul 2026 11:45:50 +0000

Hey DEV community! 👋

Building a simple Telegram bot is fun, but building a music streaming bot that streams audio to many voice chats at the same time is a completely different challenge.

For the past year, I have been developing HasiiMusicBot, an open-source Telegram music player that focuses on performance and scalability. I build it to solve many of the problems that existing music bots face.

🏗️ Tech Stack

Framework: Pyrogram (MTProto API)
Voice Streaming: PyTgCalls
Media Processing: FFmpeg & yt-dlp
Database: MongoDB with Motor
Deployment: Docker & Docker Compose

🧗‍♂️ Challenges I Solve

1. Handling Multiple Streams Without Lag

Many Python music bots have a problem with blocking tasks like downloading audio or generating thumbnails. These tasks can freeze the event loop and cause audio interruptions in all active voice chats.

I redesigned the bot to use non-blocking asynchronous tasks. This keeps the event loop free and allows the bot to handle concurrent streams smoothly with no audio drops.

2. Handling yt-dlp Anti-Bot Restrictions

Many developers now face the "Sign in to confirm you're not a bot" error when using yt-dlp with YouTube.

To reduce this issue, I improved the metadata extraction process and add secure cookie handling. This helps the bot fetch high-quality audio streams more reliably.

🤝 Open Source and Contributions

The project is now available under the GPL-3.0 License and is completely open source.

If you are interested in Python, audio streaming, or Telegram bots, please check out the project.

🔗 GitHub Repository: https://github.com/hasindu-nagolla/HasiiMusicBot

I welcome feedback, code reviews, and contributions. I also added some good first issue labels for beginners who want to make their first open-source contribution.

If you like the project, please give it a ⭐ and let me know your thoughts in the comments!