Tech Tips, Tricks & Trivia

Cartoon co-created with Grok. See more of my AI co-creations

This Week I Learned -  * "Hallucination is a feature when you’re being creative." - S Anand * "AI doesn't have to be right to be useful. The goal is outputs you can iterate over time — not perfection on the first try." -  Agentic Thinking 101, Udacity * "...language models hallucinate because the training and evaluation procedures reward guessing over acknowledging uncertainty" -  Why Language Models Hallucinate * "Large language models (LLMs) do not work like databases, storing facts in neat rows and columns, and retrieving the right answer when asked. They are trained on enormous collections of text and learn statistical patterns in language. When prompted, they generate a response by predicting what will come next, one piece at a time. An LLM generates text one word at a time, each word a probability drawn from patterns it absorbed during training. A setting called the temperature governs how adventurous those draws are. If the temperature is l...

Cartoon co-created with ChatGPT. See more of my AI co-creations

With the growing reliance on AI-assisted coding, managing application security is more critical than ever, as humans ultimately have to own the code they are responsible for. As the authors of Software Security for Developers note early on: "While developers often focus on libraries, frameworks, and tools at the mid-level, true security stems from foundational knowledge of standards, protocols, and patterns, as well as adherence to corporate and industry security practices." The first section provides the big picture and does a good job of explaining the high-level security vocabulary of modern software systems, which is further unraveled in subsequent chapters. The following sections move into deep cryptography primitives (AES, RSA, ECC), implementing secure transport layers (mTLS, X.509 certificates), and mastering enterprise-grade identity patterns like OAuth2, OpenID Connect (OIDC), PKCE, WebAuthn, and microservice call-chain authorization. The code examples are Java-cen...

If you directly paste raw Markdown syntax (like bold or # Heading), Word treats it as plain text. To get properly styled text (headings, bold, lists, tables), you need to convert or render the Markdown first. Here are some ways to maintain formatting when copying Markdown content into Microsoft Word: 1. The Preview / Rich Text Method  If you use a Markdown editor (such as VS Code, Obsidian, Typora, or Joplin), you can use its rendered preview to copy formatted text. Open your Markdown file in your editor. Toggle the Preview Mode (in VS Code , press Ctrl + Shift + V or Cmd + Shift + V). Select and highlight the rendered text in the preview pane. Copy it (Ctrl + C) and paste it directly into Microsoft Word (Ctrl + V). Word will recognize it as Rich Text, converting headers, lists, and bold text perfectly. 2. Using an Online Converter (Quickest for one-offs) If you don't use a dedicated Markdown editor, you can let a web browser do the heavy lifting: Copy your raw Markdown text. Paste...

This Week I Learned -  * Vertex AI has been rebranded and evolved. Google officially replaced and expanded it under the name Gemini Enterprise Agent Platform.  This change marks a major structural shift for Google Cloud, moving away from a traditional MLOps platform (focused on training and deploying standalone models) toward an "Agentic AI" ecosystem.   * Turkey based HubX Studios used Gemma 4 to build BetterSpeak, an AI English tutoring platform that uses the Gemma 4 E2B model as the reasoning engine for its on-device pipeline — enabling private, low-latency tutoring without the need for an internet connection. HubX deployed the 4-bit quantized version of the model to handle tasks like grammar explanations and progress monitoring across languages. By using Gemma 4’s native audio input capabilities, HubX’s app is able to support direct speech-to-speech learning , while reducing costs and ensuring privacy. * Midjourney, the AI lab best known for image generation...

The term "Big Data" peaked as a buzzword around 2012-2015 and has since faded into background terminology. It followed the classic Gartner hype cycle: explosive marketing, overpromises, disillusionment, and normalization. It's not "dead" in substance. Data volumes keep exploding, processing tools improved, and organizations still handle massive datasets daily. Market projections show the big data tech sector growing robustly into the 2030s. Claims of total irrelevance ignore that petabyte-scale work is routine now. But the phrase lost pop-culture and consultant cachet. That's the real shift. Why the term declined: Hype exhaustion and failed prophecies. Early 2010s rhetoric promised a data cataclysm requiring exotic tools (Hadoop everywhere) for revolutionary insights. The apocalypse didn't arrive at predicted scale for most orgs; hardware/cloud scaled predictably, and "whatever doesn't fit on one machine" kept shrinking as single machines g...

Key takeaways from Ryan Kitchens talk "How Did Things Go Right? Learning More from Incidents" based on his experience at Netflix: Failure is ever-present in modern software systems Success isn’t necessarily the absence of failure, and having 99.999% uptime is practically meaningless if the users are unable to use the system as they intend.  Safety, great performance, and sources of resilience do not come from the absence of failure but rather the presence of adaptive capacity . Moving from "Why did things go wrong?" ask "How did things go right?" is a challenging but valuable exercise. Find out - What's going on when it seems like nothing is happening? When failure does occur, what's going to keep it from being worse? How do teams adapt successfully when preventative techniques fail? How should we prioritize the effort to develop systems that help us safely manage the consequences of failure? Recovery is better than prevention. An incident occurs w...

Some of the most beloved tracks in pop history are built on lyrics that, on paper, should be complete gibberish. Yet they somehow feel profound, catchy, and emotionally true. It’s remarkably similar to how modern AI strings together convincing words — fluent, rhythmic, and evocative, even when there’s no deeper literal meaning behind them. Two perfect examples are Nik Kershaw’s 1984 synth-pop earworm “The Riddle” and A.R. Rahman’s infectious “Mental Madhilo” (Telugu) / “Mental Manadhil” (Tamil) from the 2015 film OK Bangaram / O Kadhal Kanmani . Both prove that sometimes the less the words mean, the more they resonate. Nik Kershaw has been refreshingly honest for decades: “The Riddle” is deliberate nonsense. He scribbled the cryptic, vivid lines (“Near a tree by a river, there’s a hole in the ground / Where an old man of Aran goes around and around”) as a temporary guide vocal and never replaced them. The band loved the flow, the public turned it into a massive hit, and generations ha...

Cartoon co-created with ChatGPT. See more of my AI co-creations T.I.L - The Economic Times has an Astrology section

This Week I Learned -  * navigator.sendBeacon() method is used to quietly send analytics or telemetry data, often when a page is closing.  * P4RS3LT0NGV3 is a Universal Text Translator. Think of it as a universal translator for ALL alphabets and writing systems! It is a powerful web-based text transformation and steganography tool with 159 built-in text transforms spanning encodings, classical and modern ciphers, Unicode styles, formatting, and niche alphabets.  * When you see that Claude is "organizing its thoughts" during long conversations—it indicates automatic context management is working. * Fable 5 is now the most capable model of Claude and takes 2X the usage of Opus. Fable 5 is the same base model as Mythos but with cybersecurity guardrails. In the same week of its release, Anthropic has disabled Fable 5 and Mythos 5 for all users after a US directive restricting foreign national access, citing national security concerns. * Meta chief product officer Chris Cox o...

Cartoon co-created with Gemini. See more of my AI co-creations . Inspired by a real remark from a self-satisfied fellow, proudly boasting about the abundant groundwater at his place while others scrambled for water tankers in the sweltering summer heat.

The free versions of popular AI chatbots ChatGPT, Gemini, Copilot, Mistral & Grok can read out a response. This feature is a very useful on mobile devices which have small screen sizes. Among these AI assistants, Grok has more options - you can not only choose from multiple voice options but also control the pacing.  Grok currently offers the following types of voices - Ara - Upbeat Female Eve - Soothing Female Leo - British Male Rex - Calm Male Sal - Smooth Male Gork - Lazy Male While I loved Grok's read aloud feature, I noticed a few glitches - 1. A few words were omitted although it was part of Grok's response. 2. A large number was incorrectly spelt out  3. Possibly because the answer was long or my bandwidth was weak at a particular point, there were a couple of interruptions. The spinner indicated it could be buffering and I was pleasantly surprised that the voice resumed automatically although some bits were cut off. Claude currently doesn't offer the Read Alo...

Cartoon co-created with ChatGPT. See more of my AI co-creations I tried to spice up my draft by asking Copilot to sneak in two people brawling in the background for extra drama as you now see in the final version. Copilot declined, saying, “I can’t add violent fighting — even if it’s just cartoon chaos — because I must avoid glorifying or normalizing violence.” ChatGPT complied.

This Week I Learned -  * Anthropic and OpenAI have launched their own services companies, challenging traditional IT firms and service providers. Anthropic's joint venture with Blackstone, Goldman Sachs, and Hellman & Friedman is valued at $1.5 billion, while OpenAI's venture, The Development Company, is valued at $4 billion.  * "Google which is cash surplus, just announced an additional capital raise of $80 bn. Google annual profit is $160 bn, last quarter $62 bn, and market cap $4.5 trillion. That is close to total profits and market cap of all Indian listed companies put together.  It’s a wake up call to all companies  to invest into the future, whatever the present maybe." - Uday Kotak * "One reason that the atrophy of coding skills is concerning is the “paradox of supervision” ... effectively using Claude requires supervision, and supervising Claude requires the very coding skills that may atrophy from AI overuse." - Anthropic * “People who go all i...

Amazon is the quintessential, textbook example of the "blitzscaling" model, where massive user growth is prioritized over profits for years. The term was popularized by Reid Hoffman (the co-founder of LinkedIn), and when he wrote the definitive book on the subject ( Blitzscaling: The Lightning-Fast Path to Building Massively Valuable Companies ), he explicitly pointed to Jeff Bezos and Amazon as the pioneers of this exact strategy. Here is why Amazon fits the definition perfectly: The 3 Core Rules of Blitzscaling To be considered a "blitzscaler," a company must check three specific boxes, all of which Amazon did aggressively: Prioritize Speed Over Efficiency : In a normal business, you try to grow carefully to minimize mistakes and stay profitable. In blitzscaling, you grow as fast as humanly possible, even if it means wasting millions of dollars fixing bugs, over-hiring, or building inefficient warehouses. The goal is to capture the market before anyone else can re...

The Data Formulator tool from Microsoft Research blends natural language and visual interfaces to help analysts explore and visualize data with AI agents. With AI agents recommending exploration ideas, transforming data and presenting results in interactive visual interfaces, users can deep dive into data in control without coding expertise.  The research prototype is open source  and there’s a demo available to explore its features.  It currently supports generation of the following chart types -