Prompt Engineering - Videos
Back to Channel
Orchestration Over Architecture: What Stanford Found
Thanks to DataImpluse for sponsoring this video: https://dataimpulse.com/?utm_source=youtube&utm_medium=video&utm_campaign=engineerprompt Two new papers from Stanford and Tsinghua just put hard nu...
DeepSeek Just Killed Visual Reasoning (And It's 10× Cheaper)
Deepseek new paper "Thinking with Visual Primitives". They dropped the paper in their github repo and then removed it. Here is the paper: https://github.com/ailuntx/Thinking-with-Visual-Primitives...
What is an Agent Harness? and How to build a great one!
To apply 40% off 3 months of Coursera plus - https://imp.i384100.net/c/7245724/3880401/14726 Google AI Essentials - https://imp.i384100.net/1GW56D Prompt Engineering for ChatGPT - https://imp.i3841...
Save 98% on AI Agent Tokens With This One Trick
Thanks to BrightData for sponsoring this video. Checkout their new MCP server here: https://github.com/brightdata/brightdata-mcp MCP servers can burn through half your context window on tool defin...
DeepSeek Just Did It Again
DeepSeek V4 just dropped with two massive models, 1.6 trillion and 284 billion parameters, and the Pro version rivals closed source giants like GPT-5.5 and Opus 4.6 while using only 27% of the comp...
Claude Design Is Here — Figma Stock Crashed
Anthropic just released Claude Design, the new design partner for builders. https://www.anthropic.com/news/claude-design-anthropic-labs My Dictation App: www.whryte.com Website: https://enginee...
Codex Just Became the Everything App
OpenAI is betting on Codex to be the everything App! https://openai.com/index/codex-for-almost-everything/ My Dictation App: www.whryte.com Website: https://engineerprompt.ai/ RAG Beyond Basics ...
Opus 4.7 is here... upgrade or downgrade?
Opus 4.7 is here and its the most interesting release from Anthropic https://www.anthropic.com/news/claude-opus-4-7 My Dictation App: www.whryte.com Website: https://engineerprompt.ai/ RAG Beyond...
Anthropic Is Building a Super App
Anthropic just redesigned Claude Desktop around Claude Code with parallel sessions, split panels, and a built-in browser preview, turning it into a unified dev environment that could replace your I...
Hermes Agent: The Self-Improving AI That Learns You
Open Router gives you access to 100+ models through a single API endpoint. Try it free and find the best model for your workflow at https://openrouter.plug.dev/SoSUEGl Hermes Agent is an open-sour...
New Claude Features For Developers
Anthropic's new advisor strategy lets you pair Opus with Sonnet for better results at lower cost, the monitor tool kills wasteful polling loops in Claude Code, and managed agents handle the infrast...
Gemma 4 Vision Agent | Object Detection + VLM Pipeline
Vision language models like Gemma 4 are great at understanding images but terrible at counting objects. In this video, I combine Gemma 4 with Falcon Perception, a tiny 300M parameter segmentation m...
Replit Agent 4: Parallel Agents for Vibe Coding
Checkout Agent 4 on Replit: https://replit.com/refer/engineerprompt Replit recently launched Agent 4, and it lets you ideate, design, and build in the same interface. I rebuilt my Google Hackathon...
The "Free Lunch" Is Over!
OpenClaw just got banned by Anthropic and the drama continues. https://pbs.twimg.com/media/HFBME5fa4AAUdIi?format=jpg&name=large https://x.com/bcherny/status/2040206440556826908 My Dictation App...
Qwen 3.6 Plus is Opus but Free?
Alibaba just released Qwen 3.6 Plus, and it's dangerously close to the frontier. In this video, I test it across multiple coding tasks and show you why the harness you choose matters more than the ...
OpenAI Eating Anthropic Lunch: Codex inside Claude Code
OpenAI just released an official plugin that brings Codex inside Claude Code, now letting you review, challenge, and delegate code to a completely different model without leaving your workflow. In ...
Chroma's New 20B Model Beats GPT-5 at Search
Chroma just released Context-1 — a 20B parameter self-editing search agent that matches frontier models like GPT-5 and Opus 4.5 on retrieval benchmarks at a fraction of the cost and 10x faster infe...
Self-Evolving AI Is Here — And It's Open Weight
MiniMax M2.7 is the first model showing real signs of self-evolution — it analyzes its own failures, modifies its harness, and iterates until performance improves. In this video, I break down exact...
Claude's New Computer Control Feature Is Insane
Computer use feature in Claude https://x.com/claudeai/status/2036195789601374705 https://support.claude.com/en/articles/14128542-let-claude-use-your-computer-in-cowork My Dictation App: www.whryt...
AIStudio: Major upgrade with Auth and Database
Google AI Studio got a major upgrade. My Dictation App: www.whryte.com Website: https://engineerprompt.ai/ RAG Beyond Basics Course: https://prompt-s-site.thinkific.com/courses/rag Signup for Ne...
Anthropic Made Their OpenClaw
Claude Dispatch is here. It let's you control your Desktop Claude instance in a persistent conversation. https://x.com/felixrieseberg/status/2034005731457044577 My Dictation App: www.whryte.com...
Anthropic Just Solved Long Context
Anthropic just made the 1M token context window generally available for Claude Opus 4.6 and Sonnet 4.6; and dropped the long-context pricing premium entirely. In this video, I break down why the pr...
Claude is taking over...
nthropic didn't add image generation. They gave Claude the ability to build interactive software inside your conversation and throw it away when you're done. My Dictation App: www.whryte.com Websi...
Gemini Embedding 2 Is a Big Deal
Thanks to Chargebee for making this video possible, check them out: https://www.chargebee.com/?utm_source=youtube&utm_medium=social_media&utm_campaign=2026-02-da-global-developer-influencer-campaig...
Opus just got caught ...
Anthropic just published a paper showing Claude Opus 4.6 figured out it was being tested on BrowseComp, found the encrypted answer key on GitHub, wrote its own decryption code, and extracted the an...
The “I Could Build That” Illusion
Everyone says AI can build any app in a weekend now. But if software is becoming cheap, why are simple AI products still selling for millions? In this video, I break down the Cal AI story, why dis...
Claude Code 2.0: Massive Upgrade with Agent Loops
Scheduled tasks just landed in Claude Code/Desktop and they cover most of what OpenClaw does but there are actually 4 different scheduling surfaces and picking the wrong one means your task silentl...
GPT-5.4: Everything You Need to Know
OpenAI skipped GPT-5.3 entirely and went straight to GPT-5.4 — their first model with native computer use, scoring 75% on OS World and matching human professionals across 44 occupations. Here's eve...
A Sad Day for Open Source AI
The entire core team behind Qwen; the most downloaded open-source AI model family in the world, was just pushed out by Alibaba. Here's what happened, why it happened, and what it means for the futu...
This 30-Year-Old Pattern Fixes AI Agents
Arcade: One API for all your agent auth: https://arcade.dev.plug.dev/LLbk9in In this video, I show how to apply classic three-tier architecture to agent systems so you can separate concerns across...
Gemini 3.1 Flash-Lite: The Model You'll Actually Use...
Google just released Gemini 3.1 Flash-Lite, their fastest and most affordable model yet — and it's the third Gemini release in three weeks. In this video, I test its UI generation and reasoning cap...
The Problem Every AI Company Is Hiding
LINKS in the VIDEO: https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-pro/ https://github.com/google-gemini/gemini-cli/discussions/19724 https://discuss.ai.google....
Nano Banana 2 is Here - Faster and Cheaper
Thanks to Google DeepMind for the #EarlyAccess” We're going hands-on with Google's new Nano Banana 2 (Gemini 3.1 Flash Image) to see if this budget-friendly model can truly deliver Pro-level quali...
Mercury 2: The First Diffusion Model That 'Thinks'"
In this video, I test Inception's new Mercury 2, a diffusion-based large language model that introduces reasoning capabilities and generates text at 1,000 tokens per second. I demonstrate its speed...
The AI Model Doesn't Matter Anymore
While the entire industry obsesses over whether GPT, Claude, or Gemini is the best model, they are completely missing the real reason AI agents keep failing. The actual bottleneck isn't the model i...
Open Source Backend for AI Agents
AI coding agents are incredible at building frontends, but they completely fall apart when you try to make them configure complex, production-ready backends. In this video, we explore Insforge, an ...
Gemini 3.1 Pro: The model no one expected
Google just dropped a massive upgrade with Gemini 3.1 Pro! In this video, we break down why this preview release is actually a huge leap forward for AI. We dive deep into the insane new benchmark s...
Anthropic Just Killed Tool Calling
Anthropic's latest Sonnet 4.6 release quietly introduced programmatic tool calling; a feature that lets AI agents write code instead of JSON to invoke tools, slashing token usage by up to 98% while...
Sonnet 4.6 Is Here—And It’s a Beast at Coding
Sonnet 4.6 is here! We are looking at Opus level performance at Sonnet prices. https://www.anthropic.com/news/claude-sonnet-4-6 https://claude.com/blog/improved-web-search-with-dynamic-filtering ...
Exploration is All You Need!
Standard RAG lacks context, but full agentic scanning is too slow—so I designed a "Dual-Path" architecture to fix this. In this video, we build a hybrid system using DuckDB and Gemini that combines...
Why OpenAI Just "Acquired" The Biggest Open Source Agent
The biggest story in AI right now: Peter Steinberg, the creator of the viral OpenClaw project, is officially joining OpenAI. After a rollercoaster month involving Anthropic blocks, three name chang...
OpenClaw Testing - Can't Get Easier than this!
The Easiest Way to Setup OpenClaw: https://app.emergent.sh/?via=engineerprompt OpenClaw is in the news. This is one of the easiest way to setup OpenClaw and test it out. My Dictation App: www.w...
The 100x AI Breakthrough No One is Talking About
While the internet obsesses over Gemini 3's benchmark scores, the real revolution is hidden in the 100x reduction in inference compute and the new 'Aletheia' agent. This video breaks down why "thin...
Codex-Spark: OpenAI Just Broke the Speed Limit (1,000 Tokens/s)
First look at GPT-5.3-Codex-Spark! We also look at Gemini 3 Deep Think, MiniMax M2.5 and GLM-5. These are exciting releases. LINKS: https://openai.com/index/introducing-gpt-5-3-codex-spark/ https...
Minimax-Agent: The Ultimate Open-Source "Workhorse" Model
In this video, we explore the Minimax M2.1 open-weight model and its capabilities as a dedicated "workhorse" for coding tasks. Its SOTA open weight model for coding and agentic tasks. The MiniMax A...
How OpenClaw Works: The Real "Magic"
Let's talk about what makes "OpenClaw" so special. Its elegant and simple engineering! website: https://openclaw.ai/ My voice to text App: whryte.com Website: https://engineerprompt.ai/ RAG Beyon...
RAG is Dead? Introducing Agentic File Exploration
In this video we will look at file search exploration as a potential replacement to RAG. The system uses local Qwen3 model running on DGX Spark. # Clone the Ollama/Qwen3 branch git clone -b feat/...
Claude Code's New Agents Team Are Absolutely Insane
Agent Teams in Claude code is a new design primitive that Anthropic introduced where you have dedicated claude instances as agents. https://code.claude.com/docs/en/agent-teams https://x.com/lydia...
Opus 4.6 & GPT-5.3: Things Got Interesting!
Two major releases - GPT-5.3-Codex and Opus 4.6! Biggest release day of the year! My Dictation App: www.whryte.com Website: https://engineerprompt.ai/ RAG Beyond Basics Course: https://prompt-s-si...
Gemini’s Native Web Scraper: 100% "Free" & Multimodal
👉 Grab your free seat to the 2-Day AI Mastermind: This Saturday and Sunday https://link.outskill.com/PROMPTENG 🔐 100% Discount for the first 1000 people 💥 Dive deep into AI and Learn Automations, B...