You have probably heard people talk about "AI" everywhere: on the news, at work, in your children's school and even in WhatsApp. Maybe a relative used ChatGPT to write a wedding speech, or your bank's website now has a chat assistant. It can sound complicated or even scary. It does not need to be. This is Part 39 of our Tech for Beginners series, and the first part of a new section on AI tools. Today we will explain what AI, chatbots and "large language models" are, without jargon. We will look at what they are good at, where they make mistakes, and the main AI assistants you can use in India today.
AI chatbots run over the internet, which we explained in Part 15. Using a chatbot is a bit like the searching we covered in Part 17, but with important differences you will learn below. If you want to explore one more assistant, our separate beginner's guide to Claude covers another popular option.
A note on changes: AI is changing faster than almost any other technology. Product names, model versions, features, free limits and prices change every few months. This guide explains ideas that stay true; always check the current details on each company's official website or app store page.
What is AI, in plain words?
Artificial intelligence (AI) means computer programs that do tasks we normally connect with human thinking: understanding language, recognising faces or voices, translating, or making suggestions.
You have already used AI for years without calling it that:
- Your phone's keyboard suggesting the next word.
- Google Maps predicting traffic and suggesting a faster route.
- YouTube recommending videos you might like.
- Your phone camera making a night photo brighter.
- Google Translate turning a Tamil sign into English.
- Your bank flagging an unusual card transaction.
What is an AI chatbot?
An AI chatbot (or AI assistant) is a program you can talk to by typing or speaking in normal language, as if chatting with a person. You ask a question or give an instruction, called a prompt, and it writes a reply. You can continue the conversation, ask follow-up questions or ask it to change its answer.
For example, you could type: "Explain what a fixed deposit is, in simple Hindi, as if to a 70-year-old." Within seconds, you get a clear explanation. Then you might say: "Now give me three questions I should ask at the bank," and it continues from there.
What is a "large language model"?
Modern chatbots are powered by something called a large language model (LLM). Here is the simple idea:
- The model is trained on an enormous amount of text: books, websites, articles and more.
- From this, it learns patterns in how words and ideas usually follow each other.
- When you write a prompt, it produces a reply by predicting, word by word, what a helpful answer would most likely look like.
Think of a very well-read person who has read millions of books but has no personal experience of the world, and who sometimes confidently "remembers" things that were never in any book. That is a useful picture of how these models behave.
Many chatbots can now also see images (you send a photo and ask about it), listen and speak (voice conversations), and search the web for recent information. These features differ between apps and between free and paid plans.
What AI chatbots are good at
- Explaining things simply: a medical term on a report, a tax word, a science topic for your child's homework.
- Writing and editing: a leave letter, a complaint email to a company, a birthday message, a resume.
- Summarising: turning a long article or document into a few key points.
- Translating: between English and many Indian languages, though quality varies by language.
- Brainstorming: ideas for a child's birthday party, a small business name or a weekly menu.
- Planning: a rough travel plan for a trip to Ooty or a study timetable.
- Practising: English conversation, interview questions or a new language.
Where AI chatbots make mistakes
This is the most important section. AI chatbots can be wrong, and they often sound confident even when they are.
- Made-up facts ("hallucinations"): a chatbot may invent a fact, a law, a statistic, a book or a web link that does not exist.
- Out-of-date information: unless it searches the web, a chatbot may not know about recent changes, such as new bank rules or prices.
- Maths and calculations: they can make errors, especially with long sums or EMI calculations. Check with a calculator.
- Indian specifics: local rules, government scheme details, state-level procedures and regional facts can be wrong or mixed up.
- Bias: because they learn from human writing, answers can reflect stereotypes.
- Not a professional: AI is not your doctor, lawyer or financial adviser. Use it to understand and prepare questions, then talk to a qualified person.
The golden rule: use AI as a helpful assistant, not as the final authority. Check important facts with official or trusted sources.
AI chatbot or search engine?
- A search engine shows you a list of websites. You decide which sources to trust, and you can see where information comes from.
- A chatbot writes one combined answer. It is faster and easier to read, but you may not see the sources unless it provides links.
For facts that matter, such as a government deadline or a medicine dose, use official websites. For explanations, drafting and ideas, a chatbot can save a lot of time. Many search engines now show AI-written summaries at the top of results, so it is worth knowing the difference.
The main AI assistants available in India
These are some widely used options at the time of writing. All have free versions with limits, and most offer paid plans.
- ChatGPT by OpenAI: one of the best-known chatbots, available on the web and as an app. We cover it in Part 40.
- Gemini by Google: built into many Android phones and Google apps. We cover it in Part 41.
- Microsoft Copilot: available on the web, as an app and inside Windows and Microsoft products.
- Meta AI: available inside WhatsApp, Instagram and Facebook in India.
- Claude by Anthropic: another popular assistant for writing and analysis.
Indian companies and government-supported projects are also building AI tools focused on Indian languages. For example, the government's Bhashini platform works on translation and speech tools for Indian languages.
A few words you will hear
- Prompt: the question or instruction you type or speak.
- Model: the AI "brain" behind a chatbot. Companies release new models often, with names and version numbers that change.
- Hallucination: when AI states something false as if it were true.
- Generative AI: AI that creates new text, images, audio or video.
- Deepfake: a fake video, photo or voice made with AI to look or sound like a real person.
Is AI safe to use?
Using a well-known AI assistant from its official app or website is generally fine, with some precautions:
- Download apps only from the official Play Store or App Store listing, and check the developer name. Fake "AI" apps exist.
- Do not share Aadhaar numbers, bank details, passwords, OTPs or private medical records.
- Remember that conversations may be stored and, depending on settings, used to improve the service. Check privacy settings.
- Be careful of AI-made fake voices and videos used by scammers.
We will look at these safety points in detail in Part 43.
Will AI take away jobs?
AI is changing how many jobs are done, especially work involving writing, customer support, data and design. Experts disagree on exactly how big the impact will be. What is clear is that people who learn to use AI tools well, while checking their output, are likely to be better placed. Learning the basics, as you are doing now, is a good start.
Frequently asked questions
Do I need to pay to use AI chatbots?
No. The main assistants have free versions that are enough for everyday use. Paid plans give more usage, newer features or more capable models. Start free and decide later.
Does AI understand Hindi and other Indian languages?
Many chatbots can read and write Hindi, Tamil, Telugu, Bengali, Marathi and other Indian languages, as well as mixed "Hinglish". Quality is usually best in English and varies between languages, so check important text with a native speaker.
Is the AI thinking or conscious?
No. It does not have feelings or understanding like a person. It produces text based on patterns it has learned, which is why it can sound human but still make strange mistakes.
Can children use AI chatbots?
Each service sets minimum ages in its terms, often 13 or above with parental permission for teens. Children should use AI under supervision, as a study helper rather than a way to copy homework.
What is next
Try this today: think of one thing you would like explained simply, such as a word on your electricity bill or a topic from the news, and write it down as a question. In the next part of Tech for Beginners, we will sign up for ChatGPT and try your first useful tasks.
