likith

non neutron | shit posting crypto, AI and web3

[Vijay @imvijaymode]

What is a context window in AI?

Sometimes AI forgets what you told it earlier.

That happens because AI has a limit.

It cannot keep the whole conversation in mind forever.

It can only work with a certain amount of text at one time.

That limit is called the context window.

Think of it like a study table.

If the table is small,
you can keep only a few papers in front of you.

If more papers keep coming,
some old ones have to move away.

AI works in a similar way.

It can only keep a limited amount of text in front of it while replying.

That is why long chats can go wrong.

If the conversation gets too long,
or the PDF is too big,
or the prompt has too many instructions,

AI may start to:

• forget older details
• miss instructions
• repeat itself
• give weaker answers

A context window is not permanent memory.

It is just the amount of text AI can actively look at while answering you.

A bigger context window means AI can handle more at once:

• longer chats
• larger documents
• more instructions in one go

That is useful.

But even big context windows still have limits.

So the practical lesson is simple:

If you want better answers,
keep your prompts clear,
keep long chats organized,
and break very big tasks into smaller parts.

Takeaway:
A context window is the amount of text AI can keep in front of it at one time. If too much fills it, AI can lose track of older details.

#AIMadeSimple

[Vijay @imvijaymode]

AI is not a brain. It is a pattern machine.

When most people hear “AI,” they imagine a robot brain.

That is the first mistake.

AI does not think like a human.

It learns patterns from huge amounts of text, images, code, voice, and data, then predicts the most likely useful output.

Think about your phone keyboard.

You type “Happy”
and it suggests “birthday.”

Your phone does not understand birthdays.

It just learned the pattern.

AI works in a similar way, just at a much bigger scale.

That is why it can write a strong email in seconds, and still give a wrong answer confidently.

Because pattern prediction is not the same as real understanding.

Starting today, I’m sharing one simple AI idea at a time under #AIMadeSimple.

Takeaway: AI does not think like a human. It predicts using patterns.

#AIMadeSimple

[Vijay @imvijaymode]

What is an LLM in AI?

One word keeps showing up in AI.

LLM.

So what is it?

LLM stands for Large Language Model.

That sounds complex.

But the idea is simple.

Let’s break it down.

Large means it has learned from a huge amount of text.

Language means it works with words.

Model means it is the trained system doing the job.

Put that together:

An LLM is an AI system trained on a huge amount of text so it can work with human language.

That is why AI tools can:

• answer questions
• write emails
• summarize documents
• translate text
• explain ideas
• chat in a natural way

In simple words:

An LLM is what helps AI read your words and reply with words.

That is why tools like ChatGPT, Claude, Gemini, and others feel so natural to talk to.

They are built to work with language.

Not because they are human.

Not because they truly understand everything.

But because they have learned patterns from a huge amount of human writing.

That is the big idea.

If you understand what an LLM is, AI starts feeling less like magic.

And more like a language machine.

Takeaway:
An LLM is an AI system trained on lots of text so it can work with human language.

#AIMadeSimple

[Vijay @imvijaymode]

Why AI gives bad answers sometimes ?

A lot of people think AI is bad.

Sometimes, it is not.

Sometimes, the instruction is bad.

That is the real problem.

If you tell AI:

“Write something for me”

the answer may be weak.

Too vague.
Too random.
Not useful.

But if you say:

“Write a short and polite email to my manager asking for leave tomorrow”

the answer gets much better.

Why?

Because AI works better when your instruction is clear.

That instruction has a name.

It is called a prompt.

Now forget the fancy word.

A prompt is just what you ask AI to do.

That’s it.

If you ask clearly, AI usually helps better.

If you ask badly, AI usually replies badly.

Just like asking a human for help.

If you say:

“Do this”

they may get confused.

But if you say:

“Help me write a birthday message for my friend in a funny tone”

now they understand what you want.

AI works in a similar way.

A good prompt does not need big words.

It does not need smart tricks.

It just needs to be clear.

What do you want?
How do you want it?
What details matter?

That is why prompting is important.

Not because it is technical.

But because clear instructions give better results.

Takeaway:
A prompt is just your instruction to AI. The clearer the instruction, the better the answer.

#AIMadeSimple

[Vijay @imvijaymode]

How AI learns from examples?

AI does not wake up smart.

It has to learn first.

But it does not learn the way humans do.

You learn by living.

By making mistakes.
By asking questions.
By trying again.

AI learns in a different way.

It learns from a huge number of examples.

That is the big idea.

Before you use AI, it is trained on many examples.

That training is how it learns patterns.

Later, when you ask it something, it uses those patterns to respond.

That is why it can help from the very first message.

If an AI is trained on lots of text, it starts seeing patterns in language.

If it is trained on lots of images, it starts seeing patterns in pictures.

If it is trained on code, it starts seeing patterns in programming.

That is how AI becomes useful.

Not because it understands life like a human.

But because it has seen so many examples that it becomes good at spotting patterns.

Think about a student who solves 10 math problems.

They learn a little.

Now imagine they solve 10 lakh similar problems.

They start seeing patterns much faster.

AI works in a similar way, but on a much bigger scale.

That is why training matters.

The more useful examples an AI learns from, the better it usually becomes at giving useful outputs later.

This also explains why AI can:

• write
• summarize
• translate
• generate images
• suggest code

It has seen enough examples to recognize what usually comes next.

That is the foundation.

AI learns from examples.

Then it uses those patterns to respond when you ask it something.

Takeaway:
AI learns from a huge number of examples first, then uses what it learned to respond.

#AIMadeSimple

[Vijay @imvijaymode]

What is an AI agent, really?

Most people think all AI does the same thing.

You ask a question.
It gives an answer.

That is only one kind of AI.

An AI agent is different.

It is called an agent because it does not just talk to you.

It tries to act for you.

A normal chatbot mostly replies.

An AI agent tries to do.

That is the simplest way to understand it.

If you ask a chatbot:

“Help me find the cheapest flight”

it may give you suggestions.

If you ask an AI agent:

“Find the cheapest flight for me”

it may search, compare options, and help complete the task step by step.

That is the difference.

A chatbot is mostly about conversation.

An agent is about action.

In simple words:

An AI agent is an AI system that is given a goal and then takes steps to try to complete it.

Those steps may include:

• searching for information
• comparing options
• using tools
• making a plan
• doing tasks in order

That is why people are excited about AI agents.

They are not just answering.

They are starting to handle small pieces of real work.

But this is also where people should stay careful.

Not every AI tool is an agent.

And not every agent is reliable.

It can still make mistakes.
It can still misunderstand the goal.
It can still need human checking.

So the big idea is this:

A chatbot gives you answers.
An agent tries to get things done for you.

Takeaway:
An AI agent is an AI system that takes a goal and tries to complete it step by step.

#AIMadeSimple

[99th Floor 🇮🇳 @99thFloorIN]

99th floor X @NexaMoney
We’re not just touring cities
we’re building ecosystems.

Nexa Otoplo India Tour lands in Odisha at Centurion University, alongside @odisha_dao
A new wave of builders is rising.
More ideas. More collaboration. More execution.

Same mission. Same energy.

[Vijay @imvijaymode]

Finally got my DGEN1 @freedomfactory today . The box looks like it’s carrying some top secret crypto tech . Inside you get a classy leather case , charger , earphones and an NFC recovery card you need for the initial setup .

The phone feels super light and the UI is smooth like butter . Did my first transaction with just a fingerprint and damn it actually works . The matrix lights on the back look straight out of a sci-fi movie .

It does heat up quite fast though , maybe the phone’s also mining in spirit . Still trying to figure out what that laser thing does .

Overall feels like holding a piece of the future . Yet to try the eSIM and physical SIM but first impression is solid .

Paisa Vasool for crypto nerds ✌️

[QuaiSquad @QuaiSquad]

SOAP is serious tech but who said it can’t look cute? 🫧
Introducing SOAPY, a PFP collection for every nationality, season and mood.

Mint to your heart’s content: https://t.co/170wREoqS7

Quote-tweet this post with your SOAPY set as PFP + mint proof and we’ll be kipper tipping each one of you 🧼

SOAPY’s looking for a home. Are you game? ✨

[QuaiSquad @QuaiSquad]

🧼 @QuaiSquad doesn’t make memes for laughs.

--> We make them to show the truth.

Other PoW projects choke on outdated economics.
SOAP reloads Proof-of-Work with fresh yield + survival.

That’s why we put the SOAP right where it matters.

@QuaiNetwork @QuaiNetworkDesi

[QuaiSquad @QuaiSquad]

Happy Weekend Everyone 😀

In the DIP all our Supported Buy Orders Executed 😅

[QuaiSquad @QuaiSquad]

Try Guessing the Correct Answer Below Questions and Get some Quai Through Kipper....

1. Guess the Domain.
2. What's Exactly we are cooking try Guessing.

@QuaiNetwork @QuaiNetworkDesi

[QuaiSquad @QuaiSquad]

--> New Collection is Out at https://t.co/eWs5J3SQWa.

Go and check it.

[Vijay @imvijaymode]

🚨 @baseapp has 8M+ wallets but 99% have ZERO identity

Bots are draining $100M+ from protocols yearly

I just built the solution: PoEP - privacy-first identity that proves you're human without revealing who you are

Here's how it works 🧵👇

[Bhupalam Venkata Datta Sai🛹 @BhupalamS]

99th Floor squad cooking up something CRAZY 🔥 Web3 has NEVER seen this before. It’s gonna take time… but when it drops, the space won’t be ready. 🤫💥 Stay locked in.
@xlikith,@AlthafP65,@bhumi_125
@Kshitij_web3,@thisisSadananda
@r_mayank_sen

[99th Floor 🇮🇳 @99thFloorIN]

99th floor X Orbit X
Community Call Incoming

Join us on 7th June 2025 at 6 PM IST with
@OrbitX_Pay × @ThePrayasu × @harshamv
What’s the agenda?

Shhh... it’s a surprise 👀
You’ve got to be there to find out!
For more details Dm us

[likith @xlikith]

Just watched Khaleja 🎬 — booked via @bookmyshow and paid in USDC using https://t.co/AXsl2RTOuz on Base!

Used the Scan & Pay feature to settle the QR directly with crypto. 📷

Flawless checkout experience. Massive thanks to @P2Pdotme for making it so effortless!

Order ID: 7132

[likith @xlikith]

@99_thFloor LFG 💥💥

[Move Club India @MoveClubIN]

Shoutout to our community partners (2/3):

@Cryptoholic_Soc
@HackTourIND
@ActualOnexyz
@DecentralGuild
@99_thFloor

Quizzes, merch, banter, shilling - and one surprise guest you’ll want a selfie with!

Come meet the Aptos team, connect with founders, and chill at the Community Hub.

Sign up now: https://t.co/GjQIn4apZ3

PS - Limited slots only
approvals on a first come, first serve basis!

[Prayasu @ThePrayasu]

OnchAin Island is a Scam.

It is heartbreaking for us and the whole team to see how the community Narratives have taken a turn.

Just because a couple of them find a way to figure out an engagement out of it.

But what exactly went down, here is a thread

[likith @xlikith]

"🍕Pizza sooo good, it’s worth 10,000 BTC!" 💥 Celebrating #BitcoinPizzaDay with max flavor and crypto swagger
With @avax , @reddtactt, @AlthafP65 and @BhupalamS 🎉🎉