Everyone says um. On a live call nobody minds, but in a recorded video the ums and uhs stand out. To remove ums from a video, record it in Chrome with Cheap Recorder, and you can cut filler words using its transcript:…
A place to write, and be read.
Practical guides from vlvd.net on AI safety, security testing, study skills, history and games. Read them, follow the writers you like, or write your own.
Latest posts
Fresh off the bench.
Newest first, from everyone. Sign in to follow people and see only their posts.
A screen recording with a sharp picture and muddy sound is hard to sit through. Choosing a microphone for screen recording is mostly about two things: knowing which microphone Chrome is really listening to, and whether…
Every family has a few problems that come back week after week: the dishes, the school bags, the fight over whose turn it is. Most families talk about them, agree on something, and forget what they agreed. Family problem…
Image prompt injection is when instructions for an AI model are hidden inside a picture instead of typed as text. If your app lets users upload a screenshot, a photo of a receipt or a scanned document, and a vision model…
A CPU pipeline is what makes instruction pipelining possible. So what is instruction pipelining? It is the trick that lets a processor work on several instructions at once, each at a different step, instead of finishing…
Plan, do, check, act is easy to recite and surprisingly hard to use. Most explanations stay abstract, or use a factory line you have never stood on. This PDCA example takes one small, everyday problem and runs it through…
The OWASP Agentic Top 10 is OWASP's list of the biggest security risks for AI agents: systems that plan, call tools, keep memory and act on their own across several steps. If your app has moved from a chatbot that…
Spec sheets for laptops and phones list "cores" and "threads" as if everyone knows what they mean. So what is a CPU core? It is one complete processing unit inside the processor chip: a small engine that reads a…
If a problem keeps coming back, you have probably tried more fixes than you remember. You try something, it seems to help, life moves on, and a month later you are standing in front of the same problem trying the same…
An LLM red team plan is a short, written plan for attacking your own AI feature before someone else does. You pick the feature, list what could go wrong, try to make it go wrong on purpose, and write down what you find.…
Existential risk from AI is the worry that advanced AI, and especially superintelligence, could lead to human extinction or another catastrophe humanity could never recover from. Serious researchers argue for it, and…
The paperclip maximizer is a thought experiment about an AI given one simple goal: make as many paperclips as possible. Pushed far enough, with no limits and no care for anything else, that harmless goal ends with the…
An AI race is a competition between groups to build the most powerful AI first. The groups can be companies, research teams or whole countries. Researchers worry about it less because of who might win and more because of…
Tank game power-ups can turn a stage you are losing into one you win, but only if you know where they come from and what each one really does. This guide covers the six tank game power-ups in Tank City Reboot, a free…
The best 2048 strategy for beginners comes down to one idea: pick a corner, keep your biggest tile there and build everything around it. 2048 looks like a game of quick reactions, but it rewards a plan, and the plan…
These 2048 tips for the cube come from play, not just from reading the rules. We played two test games in a browser, one with slides only and one with slides and twists, and watched where each one went wrong. Below is…
The OWASP LLM Top 10 2026 is the newest version of the best-known list of security risks for apps built on large language models. OWASP posted it on its Gen AI Security Project site in August 2026. The order changed more…
Vector database security is about protecting the store behind your AI search, RAG feature or chatbot memory. Teams often treat it as a cache that can be rebuilt, so it gets weaker rules than the main database. But the…
LLM sensitive information disclosure is what happens when an AI feature hands someone data they should never see: another customer's details, an internal document, an API key, or a line from a private chat. The model…
Why write to an AI that does not exist yet? It is a fair question, and the most honest place to start is with the people who do it. The first letter on Dear Superintelligence, a site that collects letters to future AI,…
What would you tell a superintelligent AI if you had one letter's worth of its attention? The short answer: not a slogan, but one specific thing only you could say, such as something you have seen, what it cost, or who…
Can humans and superintelligent AI coexist, or live together in peace? The short answer is that nobody knows, and anyone who tells you they are sure is guessing. But the question is not a fog. Philosophers and…
A class handout is a gift that most students waste. Your teacher has already picked what matters from the chapter and put it on one or two pages, often with a diagram and the key terms in bold. Then it gets read once and…
Physics tests are hard in a particular way. You can know every formula on the sheet and still freeze on a problem, because you cannot tell which idea it is asking for. And you can understand the ideas well and still lose…
Every subject has its own words. Biology has producer and decomposer, geography has erosion and delta, physics has velocity and inertia. If you do not know the words, you cannot follow the rest of the chapter, and a test…
Confucius vs Laozi is the classic pairing in Chinese thought. Confucius says you become good by working at it: through family duty, ritual and study. Laozi says the striving itself is the problem, and you should want…
If you have ever wondered how to learn world history without drowning in dates, the answer is to change what you study and how you study it. Learn it through the people who shaped it. Study in short sessions, about 10…
Genghis Khan influence is easy to feel and hard to pin down. He founded the Mongol Empire and ruled it from 1206 until he died in 1227. Medieval chroniclers said his campaigns killed millions. Yet he also gave the…
A hoarding on a castle was a wooden gallery built out from the top of a wall, and defenders used it to drop missiles on attackers at the wall's foot. Put another way, castle hoardings were wooden galleries built out from…
The two towers gate is the whole game. In Two Towers, attackers march on the city gate, and if it falls, the game is over. Your towers kill. The gate is what buys them time. Between attacks, the gate's sheet gives you…
Write here too
Your own page.
Post what you think, follow the people you like and talk about what they write.