<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Priyam Jain on Medium]]></title>
        <description><![CDATA[Stories by Priyam Jain on Medium]]></description>
        <link>https://medium.com/@priyamjainofficial?source=rss-74a05d726fd9------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*8qOUDxGM5l61CXJeDkwoDw.jpeg</url>
            <title>Stories by Priyam Jain on Medium</title>
            <link>https://medium.com/@priyamjainofficial?source=rss-74a05d726fd9------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 07:04:09 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@priyamjainofficial/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Transforming the Indian Judicial System: How AI Can Ensure Fairness for All]]></title>
            <link>https://medium.com/@priyamjainofficial/transforming-the-indian-judicial-system-how-ai-can-ensure-fairness-for-all-ec90cab1127c?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/ec90cab1127c</guid>
            <category><![CDATA[judiciary]]></category>
            <category><![CDATA[law]]></category>
            <category><![CDATA[governance]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Thu, 19 Nov 2026 09:36:42 GMT</pubDate>
            <atom:updated>2026-11-19T09:36:42.852Z</atom:updated>
            <content:encoded><![CDATA[<h3>Introduction</h3><p>The Indian judicial system, often described as slow and biased, faces significant challenges. The most prominent criticism is that <strong>the system favors the rich while neglecting the poor</strong>. Wealthy individuals can afford top lawyers, influence legal proceedings, and delay cases strategically, while the underprivileged struggle for basic legal representation.</p><p>Artificial Intelligence (AI) presents a <strong>transformative opportunity</strong> to bridge this gap and create a fairer, more efficient, and transparent justice system. In this blog, we will explore the <strong>challenges in the Indian judiciary</strong>, how <strong>AI can mitigate these issues</strong>, and the <strong>changes we will start seeing with AI implementation</strong>.</p><h3>Challenges in the Indian Court System</h3><h3>1. Case Backlogs &amp; Delayed Justice</h3><ul><li><strong>India has over 4.4 crore pending cases</strong> in various courts. Many cases take <strong>decades</strong> to reach a verdict.</li><li>The rich exploit delays through <strong>adjournments, legal loopholes, and expensive lawyers</strong>.</li><li>The poor often <strong>spend their entire lives fighting for justice</strong> with no resolution.</li></ul><h3>2. Corruption &amp; Favoritism</h3><ul><li><strong>Wealthy individuals have access to influential lawyers and networks</strong> that allow them to manipulate proceedings.</li><li>Police investigations are often biased in favor of those who can <strong>pay bribes or exert political influence</strong>.</li><li>Poor individuals struggle to <strong>file FIRs, get bail, or receive fair trials</strong>.</li></ul><h3>3. Lack of Legal Awareness &amp; Accessibility</h3><ul><li>Most of India’s poor and marginalized communities <strong>don’t understand their legal rights</strong>.</li><li>They cannot afford lawyers, leading to <strong>unfair representation and wrongful convictions</strong>.</li><li><strong>Court documents are complex and difficult to understand</strong> for a common person.</li></ul><h3>4. Inconsistent Judgments &amp; Human Bias</h3><ul><li>Different courts often give <strong>contradictory judgments on similar cases</strong>.</li><li>Judges, being human, are prone to <strong>biases, errors, and external influences</strong>.</li><li>There is <strong>no centralized system to track all past judgments and ensure uniformity</strong>.</li></ul><h3>How AI Can Solve These Challenges</h3><h3>1. AI-Powered Legal Assistance for the Poor</h3><ul><li>AI-driven <strong>chatbots and legal assistants</strong> can provide <strong>free legal advice</strong> in multiple Indian languages.</li><li>AI can help the poor <strong>file FIRs, track case progress, and understand legal terms</strong> without relying on costly lawyers.</li><li><strong>Example:</strong> Online AI platforms can assist in filling court petitions automatically based on case details.</li></ul><h3>2. Automated Case Prioritization &amp; Faster Resolutions</h3><ul><li>AI can <strong>analyze case urgency and prioritize hearings</strong> accordingly, reducing delays.</li><li>By <strong>removing unnecessary adjournments and paperwork</strong>, AI-powered scheduling can ensure <strong>timely trials</strong>.</li><li><strong>Example:</strong> AI can suggest <strong>alternative dispute resolution (ADR)</strong> methods before a case goes to trial.</li></ul><h3>3. AI-Based Judge Assistants for Fair Judgments</h3><ul><li>AI can analyze <strong>millions of past judgments</strong> to suggest <strong>the most consistent and unbiased decision</strong>.</li><li>AI-powered systems can <strong>highlight legal precedents</strong> so that similar cases receive similar judgments.</li><li><strong>Example:</strong> The Supreme Court can use AI to create a <strong>centralized database of legal precedents</strong> to ensure uniformity in judgments.</li></ul><h3>4. AI for Corruption Detection &amp; Transparency</h3><ul><li>AI can track <strong>suspicious delays, influence patterns, and unusual legal favors</strong>.</li><li>AI-powered audits can ensure <strong>police investigations are unbiased</strong> and <strong>based purely on evidence</strong>.</li><li><strong>Example:</strong> AI-driven tools can analyze case progress and <strong>flag undue delays caused by influential parties</strong>.</li></ul><h3>5. AI-Driven Predictive Policing &amp; Crime Analysis</h3><ul><li>AI can analyze crime data to <strong>predict high-risk areas and prevent crimes before they happen</strong>.</li><li>AI systems can assist police by <strong>identifying inconsistencies in statements and detecting tampered evidence</strong>.</li><li><strong>Example:</strong> AI facial recognition and forensic analysis can ensure <strong>evidence isn’t manipulated to frame innocent people</strong>.</li></ul><h3>What Changes Will We See After AI Implementation?</h3><h3>1. Speedy Justice for All</h3><ul><li>AI-powered case prioritization will reduce the <strong>average court trial time from years to months</strong>.</li><li>The poor will <strong>no longer have to wait decades</strong> for a resolution.</li></ul><h3>2. Equal Legal Representation</h3><ul><li><strong>Wealthy individuals will no longer have an unfair advantage</strong>.</li><li>AI-driven legal assistants will provide <strong>free legal advice and representation</strong> to the poor.</li></ul><h3>3. Transparent &amp; Corruption-Free Courts</h3><ul><li>AI monitoring will <strong>prevent judges and lawyers from being influenced by powerful individuals</strong>.</li><li>Automated case tracking will expose <strong>unnecessary delays, biased verdicts, and corrupt practices</strong>.</li></ul><h3>4. Uniform and Fair Judgments</h3><ul><li>AI will ensure <strong>all similar cases get similar judgments</strong>, eliminating human bias.</li><li>Case histories will be accessible to all, preventing manipulation of legal precedents.</li></ul><h3>5. Strengthened Law Enforcement &amp; Crime Prevention</h3><ul><li>AI-powered crime analysis will make policing <strong>more evidence-driven and less influenced by politics</strong>.</li><li>Predictive AI can <strong>prevent wrongful arrests and police harassment of the poor</strong>.</li></ul><h3>Conclusion: AI — The Key to a Fair Judicial System</h3><p>The Indian legal system <strong>must evolve to serve justice equally for all</strong> — not just the privileged. AI is not just a technological upgrade but a <strong>moral necessity</strong> to ensure justice is <strong>fair, efficient, and corruption-free</strong>. With <strong>AI-powered legal assistance, unbiased judgments, and corruption detection</strong>, the <strong>gap between the rich and the poor in the justice system can finally be bridged</strong>.</p><p>The time has come for India to embrace AI and <strong>build a judicial system that delivers justice swiftly, transparently, and without bias</strong>. If implemented correctly, AI can be the <strong>true equalizer</strong> that ensures <strong>justice is no longer just for the rich but for every citizen of India.</strong></p><p><strong>What do you think?</strong> Do you believe AI can bring real change to India’s legal system? Let us know your thoughts in the comments!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ec90cab1127c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Reinforcement Learning: Changing the Way We Think About Learning Systems]]></title>
            <link>https://medium.com/@priyamjainofficial/reinforcement-learning-changing-the-way-we-think-about-learning-systems-6362c3aab605?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/6362c3aab605</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Thu, 19 Nov 2026 09:34:59 GMT</pubDate>
            <atom:updated>2026-11-19T09:34:59.963Z</atom:updated>
            <content:encoded><![CDATA[<p>When we first learn Machine Learning (ML) and Deep Learning (DL), we develop a very specific way of thinking about problems. We assume that learning is about finding a mapping from input to output. Given enough data, the model learns patterns and improves its predictions. Whether it is classification or regression, the goal remains the same: <strong>minimize error and predict correctly</strong>.</p><p>Reinforcement Learning (RL), however, quietly breaks this entire mental model.</p><p>In ML and DL, the model is trained on a fixed dataset. The data exists independently of the model, and learning is essentially about extracting patterns from it. In RL, there is no such fixed dataset. Instead, the model — called an <em>agent</em> — interacts with an environment and generates its own data. This single shift changes everything. The goal is no longer prediction. The goal becomes <strong>maximizing long-term reward through a sequence of decisions</strong>.</p><p>This creates a fundamental difference in mindset. While building ML/DL systems, we think in terms of “What is the correct output for this input?” But while designing RL systems, the question becomes “What action should I take now so that I get the best outcome over time?” The model is no longer just learning relationships; it is learning <strong>behavior</strong>.</p><h3>A Shift in Perspective: From Prediction to Decision-Making</h3><p>To make this difference clearer, consider the following simplified comparison:</p><pre>ML/DL Thinking:<br>Input → Model → Output<br>Goal: Predict correctly</pre><pre>RL Thinking:<br>State → Action → Reward → Next State → ...<br>Goal: Maximize cumulative reward over time</pre><p>In ML, each data point is usually independent. In RL, every action affects the next state, and therefore future rewards. This introduces a <strong>temporal dependency</strong> that does not exist in traditional ML setups.</p><p>Another interesting way to look at it is this:<br>the <em>dataset</em> in ML becomes the <em>environment</em> in RL. Instead of passively receiving data, the agent actively explores and generates it. Learning is no longer passive — it becomes <strong>interactive and iterative</strong>.</p><h3>Where RL Still Feels Familiar</h3><p>Despite these conceptual differences, RL is not completely alien. In fact, if you look at the training process, there are strong similarities with ML and DL.</p><p>In both cases, we:</p><ul><li>Initialize a model (or policy)</li><li>Collect data (either from dataset or interaction)</li><li>Update parameters</li><li>Repeat the process</li></ul><p>This loop — <strong>initialize → learn → update → repeat</strong> — is common across ML, DL, and RL. Optimization still plays a central role, and in modern RL (Deep RL), neural networks are often used just like in DL.</p><p>So structurally, the systems look similar. But the <strong>source of data and the objective function</strong> make them fundamentally different.</p><h3>The Foundation: Markov Decision Process (MDP)</h3><p>At the core of every RL problem lies a single framework: the <strong>Markov Decision Process (MDP)</strong>.</p><p>MDP defines the problem in terms of:</p><ul><li>States (where the agent is)</li><li>Actions (what it can do)</li><li>Rewards (feedback from the environment)</li><li>Transitions (how the state changes)</li></ul><p>This framework remains constant across all RL problems. Whether you are building a game-playing agent, a recommendation system, or a robot controller, the problem is always modeled using these components.</p><p>What changes is not the framework itself, but:</p><ul><li>How states are represented</li><li>What actions are available</li><li>How rewards are defined</li><li>How transitions behave</li></ul><p>So MDP acts as a <strong>general blueprint</strong>, while the actual implementation varies depending on the problem.</p><h3>Different Algorithms, Same Goal</h3><p>One of the most confusing aspects for learners is the variety of RL algorithms — Q-learning, SARSA, Deep Q Networks, Policy Gradients, and more. At first glance, they seem like completely different approaches.</p><p>But a useful way to think about them is this:</p><blockquote><em>All RL algorithms are trying to solve the same underlying problem defined by the MDP — they just do it in different ways.</em></blockquote><p>The difference lies in <strong>how they estimate and update values</strong>.</p><p>For example, consider Q-learning and Deep Q Learning:</p><ul><li><strong>Q-learning</strong> uses a table to store values for each state-action pair.</li><li><strong>Deep Q Learning (DQN)</strong> replaces this table with a neural network that approximates those values.</li></ul><p>The goal remains the same: estimate how good an action is in a given state. But the method changes depending on the complexity of the problem.</p><h3>A Key Contrast with Deep Learning</h3><p>In Deep Learning, we often use the same neural network architecture for very different tasks. A neural network can perform classification, regression, or even generation with only minor changes in the output layer or loss function.</p><p>In RL, however, the variation is deeper.</p><p>Different problems may require:</p><ul><li>Different ways of representing value (state-value vs action-value)</li><li>Different learning strategies (on-policy vs off-policy)</li><li>Different update mechanisms</li></ul><p>For example:</p><ul><li>In a simple grid-world problem, tabular Q-learning might be enough</li><li>In a complex game like Atari, we need Deep Q Networks</li><li>In continuous action spaces (like robotics), entirely different approaches such as policy gradients are used</li></ul><p>So unlike DL, where the <strong>model structure is often reused</strong>, in RL the <strong>learning approach itself changes significantly</strong> depending on the problem.</p><h3>The Bigger Picture</h3><p>If we step back, the entire field of RL can be seen as a layered structure:</p><pre>Problem → Defined by MDP<br>Core Idea → Bellman Equation (value depends on future)<br>Methods → Q-learning, DQN, SARSA, Policy Gradients<br>Implementation → Tables, Neural Networks, etc.</pre><p>Everything connects back to a simple recursive idea:<br><strong>the value of a decision depends not just on immediate reward, but on future consequences</strong>.</p><h3>Final Thought</h3><p>Reinforcement Learning is not just another tool in the ML toolbox. It represents a shift from <strong>learning patterns</strong> to <strong>learning behavior over time</strong>.</p><p>Once this shift in thinking happens, many confusing concepts start to make sense. Algorithms stop looking like isolated techniques and start appearing as different ways of solving the same fundamental problem.</p><p>And that is the real goal of learning RL — not memorizing algorithms, but understanding the <strong>way of thinking</strong> behind them.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6362c3aab605" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Where Do IP Addresses Come From?]]></title>
            <link>https://medium.com/@priyamjainofficial/where-do-ip-addresses-come-from-a-simple-paragraph-wise-story-0b339d20de1e?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/0b339d20de1e</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Sun, 28 Dec 2025 16:44:44 GMT</pubDate>
            <atom:updated>2025-12-28T16:45:44.965Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*YyIuj3RfBIrOiVYW.jpg" /><figcaption>Regional Authorities that Manage IPs for ISPs, cloud providers, governments</figcaption></figure><h3>Before the Internet</h3><p>Before the internet existed, computers were mostly <strong>standalone machines</strong> or connected in very small local networks. There was no need for a global addressing system because computers were not talking to computers across the world.</p><p>As networks slowly grew and started connecting with each other, a big problem appeared:</p><blockquote>How does one computer uniquely identify another computer in the world?</blockquote><p>This is where the idea of <strong>IP addresses</strong> was born.</p><h3>IP Addresses: Just Numbers With Meaning</h3><p>An IP address is simply a <strong>number used to identify a device on a network</strong>. Just like a house number helps locate a home, an IP address helps locate a computer.</p><p>The important part is not the number itself, but the <strong>agreement</strong>:</p><blockquote><em>Everyone must agree that a specific number refers to a specific device.</em></blockquote><h3>Who Decides These Numbers?</h3><p>To keep the internet organized, a global authority was created:</p><p><strong>IANA (Internet Assigned Numbers Authority)</strong></p><p>IANA manages the entire pool of IP addresses for the world. It does not give IPs to users directly. Instead, it distributes large blocks of IP addresses to <strong>regional authorities</strong>.</p><h3>🌐 Regional Internet Registries (RIRs)</h3><p>IANA distributes IP ranges to <strong>5 regional authorities</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7yL_vQPvPuw0KItPvRI1-w.png" /></figure><p>These bodies:</p><ul><li>Manage IPs for ISPs, cloud providers, governments</li><li>Enforce uniqueness worldwide</li></ul><p>These regional authorities then give IP ranges to:</p><ul><li>Internet Service Providers (ISPs)</li><li>Large organizations</li><li>Cloud providers like AWS</li></ul><p>So, <strong>AWS does not invent IP addresses</strong>. It receives them through this global system.</p><h3>The Problem of Limited IP Addresses</h3><p>IPv4 has a limited number of addresses (about 4.3 billion). As the internet grew, it became clear that this would not be enough for everyone.</p><p>To solve this, engineers introduced the idea of <strong>private IP addresses</strong>.</p><h3>Private IP Addresses</h3><p>Private IP addresses are defined by an internet standard (RFC 1918). These IP ranges:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/846/1*8X_XxpyRZVhnaRDnrv-0-A.png" /></figure><ul><li>Are <strong>not allowed on the public internet</strong></li><li>Can be reused by anyone</li><li>Are meant for internal networks</li></ul><p>Your home Wi-Fi, college network, and AWS VPCs all use these ranges.</p><h3>What Is a VPC in This Story?</h3><p>When you create a VPC in AWS and choose something like:</p><pre>10.0.0.0/16</pre><p>You are saying:</p><blockquote><em>“Inside my private cloud network, I will use this block of private IP addresses.”</em></blockquote><p>AWS simply isolates this range inside your account. No global authority is involved here.</p><h3>Subnets and EC2 (Simple Analogy)</h3><p>Think of it like this:</p><ul><li><strong>VPC</strong> → A large piece of land</li><li><strong>Subnet</strong> → Sections of that land</li><li><strong>EC2</strong> → A house built in one section</li><li><strong>Private IP</strong> → The house number</li></ul><p>Every EC2 instance <strong>always gets one private IP from its subnet</strong>.</p><h3>What About Public IPs?</h3><p>Public IPs are:</p><ul><li>Globally unique</li><li>Owned by organizations like AWS</li></ul><p>When an EC2 has a public IP:</p><ul><li>The EC2 still uses its private IP internally</li><li>The public IP just forwards traffic to it</li></ul><p>Important truth:</p><blockquote><strong><em>EC2 always lives on a private IP. A public IP is only a bridge to the internet.</em></strong></blockquote><h3>Final Understanding</h3><ul><li>IP addresses are not random</li><li>They exist because of global agreements</li><li>Public IPs come from internet authorities → AWS</li><li>Private IPs are freely used inside networks</li><li>A VPC is simply a private, isolated part of the IP world</li></ul><p>Once you see this flow, networking stops being confusing and starts making sense.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0b339d20de1e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ The Engine of Intelligence: Simple Guide to How Neural Networks Learn]]></title>
            <link>https://medium.com/@priyamjainofficial/the-engine-of-intelligence-simple-guide-to-how-neural-networks-learn-fef96581c13f?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/fef96581c13f</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Mon, 06 Oct 2025 17:57:29 GMT</pubDate>
            <atom:updated>2025-10-06T17:57:29.309Z</atom:updated>
            <content:encoded><![CDATA[<p>Artificial Neural Networks (ANNs) are like the brains of machines. But what really makes them <em>learn</em>?<br> It’s their <strong>Learning Rule</strong> — the method that decides how the connections (weights) between neurons should change when the network sees new data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/0*9YD74EjpE3Acu9J4.png" /></figure><p>In simple terms:</p><blockquote><em>Learning Rule = how the network adjusts itself to become smarter over time.</em></blockquote><p>These rules mainly fall into two groups:</p><ul><li><strong>Unsupervised (Correlation-based):</strong> The network learns by finding patterns and associations.</li><li><strong>Supervised (Error-correction-based):</strong> The network learns by comparing what it predicted with the correct answer and fixing its mistakes.</li></ul><p>Let’s explore both.</p><h3>🔹 1. Correlation-Based Learning (Unsupervised)</h3><p>These rules are inspired by how the human brain works.<br> The idea is simple — when two neurons activate together, their connection gets stronger.<br> There’s <strong>no teacher or target output</strong> — the system learns naturally from the patterns in data.</p><p>Rule How It Works Where It’s Used <strong>Hebbian Learning</strong> “Neurons that fire together, wire together.” The connection (weight) between two neurons increases if they are active at the same time. Associative memory systems like <strong>Hopfield Networks</strong> and <strong>Bidirectional Associative Memory (BAM)</strong> <strong>Competitive Learning (Kohonen Rule)</strong> “Winner takes all.” Only the neuron that best matches the input gets to adjust its weights toward that input. Used in <strong>clustering</strong> and <strong>Self-Organizing Maps (SOMs)</strong> for grouping similar data</p><h3>🔹 2. Error-Correction Learning (Supervised)</h3><p>Here, the network has a <strong>teacher</strong> — a target output that it tries to reach.<br> It measures how far its prediction is from the target (this is the <strong>error</strong>) and then changes the weights to reduce this error step by step.<br> This method is behind almost all modern AI systems.</p><p>Rule How It Works Where It’s Used <strong>Perceptron Learning</strong> If the output is wrong, adjust the weights. Works only for simple, linearly separable problems. Single-layer classifiers (the first neural model ever!) <strong>Delta Rule (Widrow–Hoff / LMS)</strong> Uses gradient descent to minimize the squared error. It’s a smoother version of the perceptron rule and the base for backpropagation. Simple networks like <strong>Adaline</strong> <strong>Backpropagation Algorithm</strong> Applies the Delta Rule layer by layer, moving backward through the network using the chain rule of calculus. It adjusts all hidden layers to reduce the total error. All <strong>modern deep learning models</strong> — CNNs, RNNs, MLPs, etc.</p><h3>⚙️ From “Correlation” to “Correction”: The Big Shift</h3><p>The move from Hebbian learning to Backpropagation was a turning point in AI history.<br> It changed the focus from just <strong>associating patterns</strong> to <strong>actively minimizing mistakes</strong>.</p><p>Aspect Correlation-Based (Hebbian) Error-Correction (Backpropagation) <strong>Type of Learning</strong> Unsupervised (learns by observation) Supervised (learns from feedback) <strong>Signal Type</strong> Local — depends only on connected neurons Global — depends on the overall network error <strong>Goal</strong> Builds associations &amp; finds features Reduces prediction error <strong>Complexity</strong> Simple and biologically realistic Mathematically complex (uses calculus) <strong>Network Type</strong> Works for small/single-layer networks Works for multi-layer/deep networks</p><h3>🚀 In Short</h3><ul><li><strong>Hebbian Learning</strong> helps the network <em>associate</em> things — “this goes with that.”</li><li><strong>Backpropagation</strong> helps the network <em>improve</em> — “I was wrong, let’s fix it.”</li></ul><p>While Hebbian rules explain how biological brains might learn naturally,<br> <strong>Backpropagation is the true engine</strong> behind the deep learning systems that power today’s AI — from chatbots to image recognition and recommendation systems.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fef96581c13f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Decoding Associative Memory: Goal, Architecture, and Algorithm ]]></title>
            <link>https://medium.com/@priyamjainofficial/decoding-associative-memory-goal-architecture-and-algorithm-3fb516660140?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/3fb516660140</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Wed, 01 Oct 2025 05:18:53 GMT</pubDate>
            <atom:updated>2025-10-05T05:50:47.657Z</atom:updated>
            <content:encoded><![CDATA[<h3>Decoding Associative Memory: The Big Picture in ANN 🧠</h3><p>Associative memory is the <strong>goal</strong><em> </em>and it is achieved through certain <strong>Artificial Neural Network (ANN)</strong> architectures like the <strong>Hopfield Network</strong> and <strong>Bidirectional Associative Memory (BAM)</strong>, which learn patterns through the <strong>Hebbian Learning Rule</strong>.</p><p>In this blog, we’ll make sense of this sentence — step by step — connecting the goal, the architecture, and the algorithm that bring associative memory to life.</p><h3>🧩 What Is ANN and How Does It Make Predictions?</h3><p>Artificial Neural Networks (ANNs) are inspired by how our brain processes information. They consist of layers of interconnected nodes (neurons) that learn from data.<br> When we train an ANN, it adjusts the <strong>weights</strong> between these neurons to minimize error — allowing it to make predictions, recognize patterns, and classify data.</p><p>Different tasks call for different network architectures:</p><ul><li><strong>Convolutional Neural Networks (CNNs)</strong> specialize in detecting spatial features — like recognizing objects in images.</li><li><strong>Recurrent Neural Networks (RNNs)</strong> specialize in handling sequences — like predicting the next word in a sentence or the next value in a time series.</li></ul><p>Each of these networks serves a <strong>goal </strong>— and one of the most fascinating goals is <strong>associative memory</strong>.</p><h3>🧠 What Is Associative Memory?</h3><p>In simple terms, <strong>associative memory</strong> is the ability to <strong>remember a complete pattern when only part of it is presented.</strong><br> It’s what allows your brain to recall an entire song from hearing just its first few notes.</p><p>Unlike the memory in a normal computer (which retrieves data based on an exact address), associative memory is <strong>content-addressable</strong> — it recalls data based on <strong>similarity or association.</strong><br> This makes it powerful for pattern completion and noise-tolerant recall.</p><h3>💡 Why Do We Need to Understand Associative Memory?</h3><p>Associative memory lies at the heart of how intelligent systems can <strong>recall</strong>, <strong>generalize</strong>, and <strong>think by association.</strong><br> It represents the <strong>goal</strong> of building machines that don’t just store data, but <strong>understand relationships</strong> between them — just as the human brain does.</p><p>In ANN terms, associative memory is not a single network, but a <strong>goal</strong> achieved using specific network architectures and learning rules.</p><h3>🏗️ The Architectures: Hopfield Network &amp; BAM</h3><p>Two main neural architectures help achieve associative memory:</p><p>Architecture Type Function Goal <strong>Hopfield Network</strong> Auto-associative Stores and retrieves a pattern itself from a noisy or incomplete input. Pattern completion <strong>Bidirectional Associative Memory (BAM)</strong> Hetero-associative Stores and retrieves one pattern (output) when presented with another associated pattern (input). Pattern association</p><p>So, if the <strong>goal</strong> is associative memory, then these are the <strong>structures</strong> built to achieve it.</p><h3>⚙️ The Algorithm: Hebbian Learning Rule</h3><p>Now comes the <strong>how</strong> — the learning algorithm that trains these networks to remember associations.</p><p>The <strong>Hebbian Learning Rule</strong> is beautifully simple yet powerful:</p><blockquote><em>“Neurons that fire together, wire together.”</em></blockquote><p>When two neurons activate at the same time, the connection (weight) between them strengthens. Over time, this encodes associations between input and output patterns — allowing networks like Hopfield and BAM to recall one when given the other.</p><h3>🧭 The Big Picture: Connecting the Dots</h3><p>Let’s tie it all together:</p><ul><li><strong>Goal:</strong> Associative Memory — the ability to recall patterns based on association.</li><li><strong>Architectures:</strong> Hopfield Network (auto-associative) and BAM (hetero-associative).</li><li><strong>Algorithm:</strong> Hebbian Learning — strengthens connections based on simultaneous activation.</li></ul><p>So, the next time you hear “associative memory in neural networks,” remember this map:<br> 👉 <strong>Goal (Associative Memory)</strong> → <strong>Architecture (Hopfield/BAM)</strong> → <strong>Algorithm (Hebbian Learning)</strong></p><p>It’s not just how machines remember — it’s how they <em>think by association</em>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3fb516660140" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ What the MERN x AI Hackathon Taught Me]]></title>
            <link>https://medium.com/@priyamjainofficial/what-the-mern-x-ai-hackathon-taught-me-f5c13cc7f397?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/f5c13cc7f397</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Mon, 05 May 2025 07:39:49 GMT</pubDate>
            <atom:updated>2026-01-28T06:18:59.276Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A reflection through real experiences</em></p><h3>About the Event</h3><p>The MERNxAI Hackathon at VIT Bhopal was a two-round event where students first submitted project ideas online on March 29. Selected teams then developed their projects using the MERN stack and AI tools in an offline round on March 30. Winners received trophies, gift coupons, and paid internship opportunities at <strong>Build AI Engine</strong>, along with free meals during the event.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FZite-pU5GUSuhBl8fGk8A.jpeg" /></figure><h3>🧠 1. Planning Ahead Can Save the Day</h3><p>Before the event even began, I stayed up till <strong>2:00 AM</strong> the night before helping with the <strong>PPT checking process</strong>. After <strong>Round One</strong>, we had shortlisted <strong>around 40–50 teams out of 97</strong>, and it was important to review their submissions thoroughly. Despite the late night, I woke up early and reached the venue by <strong>6:30 AM</strong>, earlier than any other organizing member.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2GmaF3Vd8AdiU2XnScB-hQ.jpeg" /><figcaption>Teams we Shortlisted</figcaption></figure><p>One of my main responsibilities was <strong>entry management</strong>, and I had already planned how I’d manage it. Thanks to my <strong>past event experience</strong> and using <strong>ChatGPT</strong>, I had mentally rehearsed everything — from verifying team details to seating plans.</p><p>To handle attendance and streamline entry, I created a <strong>Google Sheet</strong> with <strong>App Script integration</strong>. I even applied <strong>First Normal Form</strong> (from DBMS) to keep the data clean and efficient. This planning helped me start confidently, even though I was nervous at first.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6b44FoQJapP6VOArCqbgTw.png" /><figcaption>Created the Attendace Sheet from the list of selected teams</figcaption></figure><h3>🪑 2. Resourcefulness in the Face of Constraints</h3><p>There was a special constraint for seating: <strong>no two consecutive teams could sit back-to-back</strong>.<br> To manage this, I took a <strong>pen and paper</strong>, sketched out a <strong>table layout</strong>, and marked the positions of each team as they arrived. As more teams came in, I realized some seats were running out, so I communicated with our <strong>club president</strong> to check if there were still unassigned tables we hadn’t considered.</p><p>This moment taught me that sometimes <strong>on-paper plans</strong> don’t hold up in the real world — and that’s okay, as long as you stay resourceful and <strong>communicate effectively</strong>.</p><h3>💬 3. Communication is a Superpower</h3><p>There were two moments that highlighted this for me:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RQ7K4ZYCNfXUrQwyBJ_pyA.jpeg" /><figcaption>Engaging with Teams About Their Innovations</figcaption></figure><ul><li>When seats ran short, I immediately contacted our president and coordinated with the team to identify available space.</li><li>When I was talking to teams about their projects, I had to ask clear, meaningful questions while also making them feel comfortable. Some teams were nervous, especially first-years — but with the right words and tone, I was able to <strong>build trust and guide them</strong>.</li></ul><p>Also, there was a team whose seat was at the very end, and the space was tight. While interviewing them, I <strong>almost stepped on their charger</strong>. That awkward moment reminded me how <strong>important spatial awareness and sensitivity</strong> are during such interactions.</p><h3>❤️ 4. Leadership Is Quiet but Powerful</h3><p>Throughout the event, I helped several teams — especially freshers — who were unsure about <strong>GitHub</strong> and collaboration. It gave me a chance to be the one who steps in, listens, explains, and guides.</p><p>Observing the participating teams, I noticed a beautiful pattern:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hcMvd1UNRNUeqIjTa4uRjA.jpeg" /></figure><blockquote><em>One or two members often took the lead, dividing tasks, guiding others, and keeping everything on track.<br>That’s the kind of </em><strong><em>leader I aspire to be</em></strong><em> — skilled, helpful, and calm under pressure.</em></blockquote><h3>🍽️ 5. Small Moments Matter the Most</h3><p>Even though I didn’t return to my hostel during the day (it’s 2 km from campus), I had simple, memorable meals.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g7wrQKj5PLq-SGV_bTeQPQ.jpeg" /></figure><ul><li><strong>Breakfast</strong> was a sandwich in the canteen.</li><li><strong>Lunch</strong> was special — it was hosted by our <strong>Vice President Arushi Puri</strong>, as a gesture to appreciate our work.</li><li><strong>Dinner</strong> was just an <strong>aloo paratha</strong> with teammates, but it felt fulfilling after a long day.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*TCbEIpA_euHXQCcZ" /></figure><p>We also had <strong>lots of photography time</strong>, since while teams were busy coding, we had enough breathing room to capture memories.</p><h3>🪞 6. Self-Love and Emotional Growth</h3><p>This was unexpected.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T1RSsbN7GXNRtmbfo7fawg.jpeg" /></figure><p>During the event, we had some free time and ended up chatting about personal stories — many of which were about <strong>relationships</strong>. Most members had either been in one or were in one, and I found myself in a different space.</p><p>Due to my <strong>cultural values</strong>, I can’t get into relationships. At first, it felt like I was missing out. But then a realization hit me:</p><blockquote><em>What if I could be the one who gives myself that emotional support, that encouragement, that belief in myself?</em></blockquote><p>From those conversations, I learned about <strong>self-love</strong>. I understood that I could be the one who <strong>pushes myself</strong>, <strong>sees the good in me</strong>, and <strong>celebrates my wins</strong>, even the little ones.</p><h3>🎤 7. Confidence Doesn’t Need Perfection</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QdgPqT2Tct5sRbLjGajJSw.jpeg" /></figure><p>I love things to be perfect. But I realized during this event — from stage management, seating issues, to last-minute adjustments — that not everything needs to be flawless. What matters more is being <strong>present, prepared, and open to learning</strong>.</p><p>And from our <strong>club president, vice president, teammates, and even freshers</strong>, I saw how to carry yourself with <strong>grace and confidence</strong>, even when things are messy behind the scenes.</p><h3>✨ Final Note:</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*w0RCD8qj3NyEFa6GGtQNXg.jpeg" /></figure><p>I joined the club not just for coding — but for <strong>collaboration and communication</strong>. This event gave me <strong>everything I hoped for</strong>, and so much more. From managing the entry to helping teams, from meal breaks to deep conversations — it was a journey of technical, emotional, and personal growth.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f5c13cc7f397" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[☁️ Beyond the Cloud: Dew, Fog, and Edge Computing Explained Simply]]></title>
            <link>https://medium.com/@priyamjainofficial/%EF%B8%8F-beyond-the-cloud-dew-fog-and-edge-computing-explained-simply-3ace4dafbcbb?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/3ace4dafbcbb</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Fri, 02 May 2025 06:07:58 GMT</pubDate>
            <atom:updated>2025-05-02T06:07:58.905Z</atom:updated>
            <content:encoded><![CDATA[<p>In the world of cloud computing, everything feels far away — data, apps, and storage live on big servers you don’t see. But did you know there are “closer” ways to compute? Meet <strong>Dew</strong>, <strong>Fog</strong>, and <strong>Edge</strong> computing — the layers between your device and the cloud. Think of them as the atmosphere layers of modern tech!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*n0KxOfQq6D6hs6wv" /></figure><p>Let’s break them down with examples that make total sense👇</p><h3>☁️ Cloud Computing (The Sky)</h3><p><strong>What it is:</strong> Centralized computing over the internet.</p><p><strong>Example:</strong> Using Google Drive to edit documents stored online.</p><h3>🌫️ Fog Computing (The Mist Around You)</h3><p><strong>What it is:</strong> Extension of cloud services closer to the ground — usually via routers, gateways, or local servers.</p><p><strong>Purpose:</strong> Reduces delay (latency), improves speed, and supports IoT.</p><p><strong>Real-Life Example:</strong><br> Smart traffic lights in a city use fog computing to analyze traffic patterns in real-time via roadside units (not the central server), allowing them to adapt quickly without depending on distant cloud servers.</p><h3>🖥️ Edge Computing (Right on the Edge)</h3><p><strong>What it is:</strong> Computation directly on local devices (or near the data source) like smartphones, sensors, or IoT devices.</p><p><strong>Purpose:</strong> Ultra-low latency, works even with spotty internet.</p><p><strong>Real-Life Example:</strong><br> A drone capturing and processing video for object detection directly on-board (without uploading every frame to the cloud).</p><h3>💧 Dew Computing (On the Ground — Your Device)</h3><p><strong>What it is:</strong> Combines the power of local devices (like your laptop) with cloud backup. It works offline-first and syncs with the cloud later.</p><p><strong>Purpose:</strong> Offline support + cloud sync when online.</p><p><strong>Real-Life Example:</strong><br> Microsoft OneNote on your laptop. You can take notes offline. Once connected to the internet, it syncs with your cloud account.</p><h3>🧠 Quick Comparison Table:</h3><p>Feature Cloud Fog Edge Dew Location Centralized Intermediate At/near device On-device Latency High Low Very Low None (offline) Internet needed? Yes Yes (less) Not always No (until sync) Use case Storage, apps Smart cities Drones, AR Notes, caching</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1011/1*hPjgo7tjSZtHfgzFJ-crOg.png" /></figure><h3>🧵 Final Thought</h3><p>Cloud computing is powerful — but not always fast or efficient for real-time needs. That’s where fog, edge, and dew come in, making tech more responsive, reliable, and human-centered.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3ace4dafbcbb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[My Experience in ADVITYA’25]]></title>
            <link>https://medium.com/@priyamjainofficial/my-experience-in-advitya25-53717acc8443?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/53717acc8443</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Wed, 02 Apr 2025 11:41:43 GMT</pubDate>
            <atom:updated>2025-04-02T11:49:39.750Z</atom:updated>
            <content:encoded><![CDATA[<p>ADVITYA’25 is the annual techno-cultural fest of VIT Bhopal, bringing together students from diverse backgrounds to showcase their talents in technology, culture, and innovation. The fest features a variety of events, ranging from coding competitions and technical workshops to cultural performances and social gatherings. As one of the most anticipated events on campus, ADVITYA’25 provided an incredible platform for learning, networking, and pushing boundaries. My experience in this fest was both challenging and rewarding, as I took on multiple responsibilities across different college clubs, gaining hands-on experience in event management and technical execution.</p><p>During ADVITYA’25, I was involved in three college clubs: MERN Matrix, Ganga Bhumi, and Microsoft Technical Club. While I was a volunteer in the Microsoft Club, I was a core technical team member in the other two clubs. In these roles, my primary responsibilities included handling registration and managing event entries.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T5QjU4SG8GPaz-LhRUUcDg.jpeg" /></figure><h3>Event 1: Prom Day (20th Feb) — MERN Matrix</h3><p>For Prom Day, my responsibility was managing student entries, primarily ensuring smooth on-spot registration and verifying participant details. The event mostly attracted couples, and my role was to facilitate an efficient check-in process.</p><h3>Event 2: Ganga Bhumi Event (21st Feb)</h3><p>This event provided a great opportunity for me to implement a QR-based attendance system using Google Sheets and Forms. Within just two days, I developed and deployed this system, generating and distributing QR codes via bulk email to all participants. I was also responsible for checking QR codes at the entry points, ensuring seamless attendance management.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2UkkydZUjLsUGW-MbSdb3w.jpeg" /></figure><h3>Event 3: Code Garuda (22nd Feb)</h3><p>This was the most significant event for me, where I had the chance to apply a wide range of skills, from project coordination to technical execution. Although others could have taken charge of the technical management, I was given the responsibility, which was both challenging and rewarding.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wBmVg5MBJ8ciNKbNzreTEw.jpeg" /></figure><h4>Pre-Event Preparations</h4><ul><li><strong>Question Selection:</strong> I was involved in curating the problem statements for the competition.</li><li><strong>Google Sheet &amp; Drive Integration:</strong> I developed a structured Google Sheet linked to a shared Google Drive folder for efficient data management. Although it seemed simple, ensuring its usability and scalability required extensive planning.</li><li><strong>Initial Planning Meeting:</strong> A day before the event, Vanya and I had an impromptu discussion in the library to finalize the technical workflow. Initially, we considered using GitHub Gist for code submissions, but after analyzing the complexities of judgment and evaluation, we shifted our focus to optimizing the judging process instead. This led to the idea of organizing code submissions into slots and streamlining the swapping mechanism for teams.</li></ul><h4>During the Event</h4><ul><li><strong>Technical Execution:</strong> The biggest challenge was handling multiple technical issues, such as file upload errors, broken links, and ensuring proper code swapping between teams.</li><li><strong>Real-Time Communication:</strong> Despite the pressure, I found myself communicating effectively, making quick decisions to resolve participant concerns. I was in a high-pressure state, where everything was happening rapidly, and I was operating on instinct rather than overthinking.</li><li><strong>Team Coordination:</strong> My friends Puranjay and Virat played a crucial role in ensuring fair code swapping by grouping teams with compatible programming languages. They also handled the checking process along with Prem, ensuring smooth execution.</li><li><strong>Last-Minute Interview Round:</strong> Towards the end, the guest requested the technical team to conduct interviews with participants. In hindsight, I realized it could have been better structured. Instead of delegating the work, a more centralized evaluation process might have been more efficient.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*N5YG9bjRFBY3FBQD4jcuew.jpeg" /></figure><h4>Post-Event Reflections</h4><ul><li>The results were finalized automatically, and while I had some input in deciding second and third place, the overall process was fast-paced and somewhat overwhelming.</li><li>I found myself reflecting on how I handled the event. While I was happy with my contributions and the recognition I received, I also realized the importance of having a larger technical team for future events.</li><li>Throughout the event, I was highly engaged, balancing multiple responsibilities from technical management to sound system handling, often feeling like I was acting instinctively rather than strategically.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iCCJD4VUMfgptrIyl5z1Qw.jpeg" /></figure><h3>Final Thoughts</h3><p>ADVITYA’25 was a transformative experience for me. It pushed me beyond my comfort zone, allowing me to implement my technical skills in real-time while also enhancing my ability to handle pressure and communicate effectively. Although there were moments of self-doubt and rapid decision-making, I am grateful for the opportunities and lessons this event provided. Moving forward, I aim to refine my approach to event management and contribute even more effectively in future endeavors.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sVBY19oacUzPawQ4ZkptTw.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=53717acc8443" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Closest I got to universal principle that guarantees a successful product without stakeholder…]]></title>
            <link>https://medium.com/@priyamjainofficial/the-closest-i-got-to-universal-principle-that-guarantees-a-successful-product-without-stakeholder-f0f92811031a?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/f0f92811031a</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Sun, 02 Mar 2025 13:46:48 GMT</pubDate>
            <atom:updated>2025-03-02T13:47:12.043Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Closest I got to universal principle that guarantees a successful product without stakeholder input</h3><p>— — — — — — — — — — — — — — — — — — — — — — — — — — — — — —</p><p><strong>Note to Visitors:</strong></p><p>Welcome to My Research Blog!</p><p>This blog, titled <em>“The Closest I Got to Universal Principle That Guarantees a Successful Product Without Stakeholder Input,”</em> is a personal collection of insights and principles from various fields like product design, usability, and cognitive psychology. While the content here is still a work in progress, it serves as a reference point for me to gather ideas and research that I can come back to for further exploration.</p><p>The goal is to compile key design principles that help create successful products even without direct stakeholder feedback, focusing on universal patterns that appeal to users. This will serve as a foundation for future projects and learning.</p><p>Feel free to explore the concepts, and stay tuned for more updates as I continue researching and expanding the content!</p><p>— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —</p><h3>1. Human-Centered Design (HCD) &amp; Usability Heuristics</h3><p>A product should be designed with the user’s needs, behaviors, and psychology in mind. If a product follows <strong>common usability heuristics</strong> (like clarity, error prevention, and consistency), it becomes easy to use — even without testing with actual users.</p><p><a href="https://blog.logrocket.com/ux-design/human-centered-design-principles-and-applications/">Human-centered design: Principles and applications — LogRocket Blog</a></p><p>Norman model of interaction: <a href="https://www.educative.io/courses/intro-human-computer-interaction/normans-model-of-interaction"><strong>Norman’s Model of Interaction</strong></a></p><h3>2. Cognitive Psychology &amp; Affordances</h3><p>Affordances mean that an object should <strong>naturally suggest its function</strong> just by looking at it. For example, a door with a <strong>handle suggests pulling</strong>, while a <strong>flat plate suggests pushing</strong>. If a product’s design follows natural expectations, users can figure it out without instructions.</p><p><a href="https://oecs.mit.edu/pub/984ungzu/release/1#:~:text=The%20concept%20affordance%20has%20been%20central%20to%20the,across%20the%20perceptual%20and%20cognitive%20sciences%20more%20broadly.">Affordances · Open Encyclopedia of Cognitive Science</a></p><h3>3. Fitts’ Law &amp; Hick’s Law</h3><ul><li><strong>Fitts’ Law</strong>: The closer and bigger a button is, the easier it is to click. That’s why <strong>important buttons (like “Buy Now”) are large and easy to reach</strong>.</li><li><strong>Hick’s Law</strong>: The more choices a user has, the <strong>slower their decision-making</strong>. Simple and clear options improve usability.</li></ul><p><a href="https://iq.opengenus.org/hicks-law-and-fitts-law-in-ux/#:~:text=Hick%27s%20Law%20and%20Fitt%27s%20Law%20are%20two%20fundamental,are%20intuitive%2C%20easy%20to%20use%2C%20and%20aesthetically%20appealing.">Hick’s Law and Fitt’s Law in UX/CSS</a></p><p><a href="https://medium.com/@rashmibarman99/10-ux-laws-analysis-with-examples-cdaea246af40">“10 UX LAWS ANALYSIS with examples” | by Rashmibarman | Medium</a></p><h3>4. Jakob’s Law of Web UX</h3><p>People spend most of their time on other websites and apps, so they expect <strong>yours to work similarly</strong>. If you make a website or app that looks and behaves like familiar ones, users will adopt it quickly.</p><h3>5. Aesthetic-Usability Effect</h3><p>If something <strong>looks good, people assume it works well</strong> — even if it doesn’t. A clean, minimal design can make a product feel easier to use, even when the functionality is unchanged. Apple products use this principle effectively.</p><p><a href="https://lawsofux.com/aesthetic-usability-effect/">Aesthetic-Usability Effect | Laws of UX</a></p><p><a href="https://www.nngroup.com/articles/aesthetic-usability-effect/">The Aesthetic-Usability Effect</a></p><h3>6. Minimum Viable Product (MVP) &amp; Lean Startup</h3><p>Instead of spending years perfecting a product, release a <strong>basic but functional version</strong> first. Then, improve it based on real user feedback. This way, you don’t waste time building things users don’t need.</p><p><a href="https://blog.logrocket.com/product-management/minimum-viable-features-mvf/">How to identify the minimum viable features (MVF) for your product — LogRocket Blog</a></p><p><a href="https://www.atlassian.com/agile/product-management/minimum-viable-product">Minimum Viable Product (MVP): What is it &amp; Why it Matters</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f0f92811031a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[You are not the one Single Identity, here is one more perspective to add to your thinking]]></title>
            <link>https://medium.com/@priyamjainofficial/you-are-not-the-one-single-identity-here-is-one-more-perspective-to-add-to-your-thinking-b1b49fdfae08?source=rss-74a05d726fd9------2</link>
            <guid isPermaLink="false">https://medium.com/p/b1b49fdfae08</guid>
            <dc:creator><![CDATA[Priyam Jain]]></dc:creator>
            <pubDate>Sun, 09 Feb 2025 15:48:55 GMT</pubDate>
            <atom:updated>2025-02-09T15:48:55.686Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/255/0*MMrzdAK09Br17E38.jpg" /></figure><h3>Who Are We? A Brain’s Perspective</h3><p>Have you ever stopped to wonder, <em>Who am I?</em> From a brain’s perspective, the answer might surprise you. It turns out, we’re not a single, unified identity. Instead, we’re a complex interplay of multiple systems and processes, each contributing to what we perceive as “self.” These systems sometimes align, but often, they conflict — leading to the internal struggles we all experience.</p><p>Two key players in this internal orchestra are the <strong>limbic system</strong> and the <strong>prefrontal cortex</strong>. The limbic system is emotional and focused on immediate gratification, driving us toward desires and comfort. In contrast, the prefrontal cortex is logical and oriented toward long-term goals, helping us plan, stay disciplined, and work hard.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*uV4mo1A-qHEfAsIk" /></figure><p>These systems often compete, creating internal conflict. When you procrastinate for days and then suddenly feel a surge of motivation before an exam, it’s your limbic system yielding to the prefrontal cortex under pressure. This dynamic interplay of emotions, logic, and awareness makes us who we are — a complex and fascinating union of different parts working together to shape our consciousness and actions.</p><p>This inner conflict isn’t just a lack of willpower; it stems from the way our brain is wired. Different parts of the brain have competing priorities, creating a tug-of-war that affects our decisions and actions.</p><h4>A Group of Many, Not One</h4><p>What makes us who we are is not one cohesive personality but a collection of parts in our brain working together (or against each other). Consider two primary systems:</p><ol><li><strong>The Limbic System (The Emotional Brain)</strong></li></ol><ul><li>The limbic system is the brain’s emotional and reward center. It’s responsible for processing emotions, desires, and immediate gratification.</li><li>It seeks pleasure and avoids discomfort, pushing us toward activities that are easy and enjoyable in the short term, like watching TV or scrolling through social media.</li><li>This system plays a key role in procrastination because it prioritizes immediate rewards over long-term benefits.</li></ul><p><strong>2. The Prefrontal Cortex (The Rational Brain)</strong></p><ul><li>The prefrontal cortex, located in the frontal lobe, is responsible for logical thinking, planning, and self-control.</li><li>It helps us focus on long-term goals, prioritize tasks, and resist distractions.</li><li>This part of the brain drives hard work, consistency, and progress.</li></ul><h4>Why Does the Limbic System Often Win?</h4><p>The limbic system operates automatically, responding to impulses and emotions without much effort. On the other hand, the prefrontal cortex requires energy and focus to function. When we’re tired, stressed, or overwhelmed, the limbic system takes over because it’s easier to follow its lead. This is why we tend to procrastinate, even when we know it’s not in our best interest.</p><h4>The Exam Week Example</h4><p>Consider a situation where you have five days to prepare for an important exam. Despite having ample time, you only start studying seriously on the last day. Why does this happen?</p><ul><li><strong>Days 1–4:</strong> The limbic system dominates, convincing you that relaxing now is better and that you can always study later. It underestimates the urgency and overvalues comfort.</li><li><strong>Day 5:</strong> The pressure of the approaching deadline activates the prefrontal cortex. Suddenly, the importance of the exam becomes clear, and you start working hard. Stress can act as a trigger that shifts control to the rational brain.</li></ul><h4>What Makes Us Conscious?</h4><p>If we’re made up of these distinct systems, what creates the sense of “I” — our consciousness?</p><ul><li><strong>The Prefrontal Cortex:</strong> While it’s not the sole source of consciousness, this part of the brain integrates information, plans actions, and reflects on our decisions. It provides a sense of continuity and purpose.</li><li><strong>The Limbic System:</strong> Although emotional and instinct-driven, it contributes to the richness of our experiences, shaping our motivations and reactions.</li><li><strong>The Interplay:</strong> Consciousness emerges not from one area but from the seamless integration of various brain functions. It’s like an orchestra where each instrument plays a part, creating the symphony we call “self-awareness.”</li></ul><h4>The Bigger Picture</h4><p>Procrastination and hard work are outcomes of the dynamic interplay between different parts of the brain. Understanding this inner conflict can help you make better choices. By strengthening your prefrontal cortex and managing the impulses of your limbic system, you can align your actions with your long-term goals and create a more productive, fulfilling life.</p><h3>References</h3><ul><li>McGonigal, K. (2012). <em>The Willpower Instinct: How Self-Control Works, Why It Matters, and What You Can Do to Get More of It.</em> Avery.</li><li>Thinking in Black, White, and Gray. (2021, November). <a href="https://www.psychologytoday.com/us/blog/thinking-in-black-white-and-gray/202111/two-parts-the-brain-govern-much-mental-life"><em>Two Parts of the Brain Govern Much of Mental Life</em></a><em>.</em> Psychology Today.</li><li>OpenAI. (2025). <em>ChatGPT</em></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b1b49fdfae08" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>