About this show
I'm Tim, a self-taught developer & entrepreneur who brings you educational tech content without the fluff and noise. I went from a broke dropout working at McDonalds to making $100k+/month as a developer, travelling the world and moving to Dubai all before the age of 22. I've since helped thousands of aspiring developers land their first software engineering role and assisted mid-level devs land jobs at companies like google.
Videos come from the channel's public YouTube feed. The artwork is the newest video thumbnail. Last updated . Names, artwork and text belong to their publishers.
Cosmos
Similar by topic and creator
- IGN
Shares the topic Technology. - Theo - t3․gg
Shares the topic Technology. - David Bombal
Shares the topic Technology.
Videos
Python Voice Agent Tutorial: Clone Your Voice & Talk to AI
· 15 min
Get started with AssemblyAI here: https://www.assemblyai.com/dashboard/signup?utm_source=youtube&utm_medium=influencer&utm_campaign=techwithtim&utm_content=tts_voice_agent In this video I clone my own voice and build a Python AI agent that talks back to me in it, in under 100 lines of code. You’ll hear the demo at the start, and the clone sounded close to me on the very first version I tried. 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-nov26 🎞 Video Resources 🎞 Download the code: https://www.aiagentbuilders.co/yt-resources/voice-clone-agent ⏳ Timestamps ⏳ 00:00 | Overview & Demo 00:45 | Architecture 02:09 | Cloning Your Own Voice 04:05 | Project Setup 06:09 | Writing the Code Hashtags #AssemblyAI #Python #VoiceAI UAE Media License Number: 3635141
Not rated yetClaude Fable 5.2 is INSANE
· 10 min
Learn how to make money with AI agents: https://aiagentbuilders.co/pro I gave Claude Fable 5.2 one prompt, went to bed, and woke up to three finished apps. In this video I show you the exact prompt, what it built overnight, and what it actually cost in tokens and usage limits. By the morning it had built a 3D glider game in raw WebGL2, a programming language with its own WebAssembly compiler and IDE, and a multiplayer whiteboard with a custom sync engine and zero npm packages. That is almost 14,000 lines of code in under 4 hours, with no follow-up messages from me. I walk through each app, break down where the 60 million tokens went, and give you my honest verdict on whether Fable 5.2 is worth it. Learn how to make money with AI agents: https://aiagentbuilders.co/pro 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-nov26 ⏳ Timestamps ⏳ 00:00 | Intro 00:27 | The prompt 01:08 | The night in numbers 01:58 | App 1: Emberwake 04:46 | App 2: Nocturne 06:11 | App 3: The whiteboard 08:27 | Where the tokens went 09:03 | Verdict 09:28 | Outro Hashtags #Claude #Fable #AICoding #Anthropic #ClaudeCode UAE Media License Number: 3635141
Not rated yetI Tested the Cheapest Coding Model on 7 Real Builds
· 18 min
🔗 Read more about MiniMax M3.1 Flash: http://minimax.io/blog/minimax-m31f 💻 Download MiniMax Code Desktop: https://agent.minimax.io/download For about two weeks, a mystery AI model called Space Bunny held the #1 spot on OpenRouter, and nobody knew who was behind it. Turns out, it was an early version of MiniMax M3.1 Flash, which has now officially launched with even more impressive capabilities. So I wanted to see what all the hype was about. In this video, we put MiniMax M3.1 Flash to the test, explore its coding capabilities, and see what it can actually build. Want to LEARN how to build useful AI agents that actually get things done? Go here: https://aiagentbuilders.co/pro How to Make Money With Coding instantly: https://techwithtim.net/newsletter 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-nov26 ⏳ Timestamps ⏳ 00:00 | Overview 00:50 | Model Stats 03:21 | Experiment Rules 04:21 | How to Run MiniMax M3.1 Flash 06:04 | Demo 1 - Neon Dash 06:46 | Demo 2 - Chrome Extension 07:41 | Demo 3 - Data Analysis 08:51 | Demo 4 - Kahoot Clone 09:53 | Demo 5 - Adding to Existing Appl 11:09 | Demo 6 - Web Operating System 12:44 | Demo 7 - Minecraft Clone 13:45 | $10 Build - Figma Clone 16:21 | Cost Breakdown Hashtags #MiniMax #MiniMaxM31 #ClaudeCode UAE Media License Number: 3635141
Not rated yetWhy Your AI Agent Needs an IDE (Real Experiment, Not Hype)
· 16 min
Give your AI agent the same tools you use. Try JetBrains IDEs free 👉 https://jb.gg/IDEs-JetBrains 📖 Why we still need an IDE (JetBrains blog): https://jb.gg/6d1jwb Some developers say AI killed the IDE. The agent writes the code and runs the tests, so why not just live in the terminal? On small projects that works fine. But in the terminal your agent is mostly coding blind: it searches through text and hopes it lands on the right thing. Your IDE actually understands the code. So I tested what happens when the agent gets the same tools you do, first on a small Python project and then on a real production codebase of about 220,000 lines. 🧠 Want to LEARN how to build AI agents that actually get things done? Start here: https://aiagentbuilders.co/pro 🎞 Video Resources 🎞 JetBrains IDEs: https://jb.gg/IDEs-JetBrains Blog post – Why We Still Need an IDE: https://jb.gg/6d1jwb 🚀 Tools I Use (10% off with code techwithtim) OpenClaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (best AI dictation): https://ref.wisprflow.ai/TechWithTim-nov26 ⏳ Timestamps ⏳ 00:00 | Overview 00:29 | The Experiment 01:14 | Arguments Against an IDE 02:15 | What an IDE Knows 04:43 | Giving your Agent an IDE 07:47 | Why You Need an IDE 09:49 | Production Demo 14:21 | The Human Review Layer Hashtags #JetBrains #PyCharm #ClaudeCode UAE Media License Number: 3635141
Not rated yetThe Most Autonomous AI Agent
· 42 sec
Is Paperclip worth it? It's one of the more out-there agent frameworks, and probably the most autonomous on the list. Set it up right and you've got a full team of agents running on a budget all day, completing work for you, organizing themselves, even hiring new agents, like a whole company running in the background. The catch: it's genuinely hard to get working effectively because so much is happening at once. But the demos people have built, entire companies run on this framework, are impressive, especially when the budget isn't tight. #techwithtim #ai #aiengineer
Not rated yetAn AI That Runs a Whole Company
· 58 sec
What if an AI didn't do the work itself, but managed a whole team of agents to do it? That's Paperclip. It runs agents like a company: I set up one called Agent Builders HQ with a single goal, and it hired a small team with a real org chart, a CEO, a founding engineer, and a researcher, all working toward that goal. It burns through a lot of tokens because it handles many tasks at once, and you can build a complex org chart of agents working together over a long period. Each one kicks off automatically on a heartbeat, say every 30 minutes or every hour, and starts completing tasks. #techwithtim #ai #aiengineer
Not rated yetThe AI Coding Agent for Your Ticket Backlog
· 42 sec
Is OpenHands the right AI coding agent for you? The upside: it gives the best end-to-end coding performance of any agent here, and nothing leaves your server except model calls, so it's private and works with almost any model. The downsides: it's only useful for coding, it's heavier because of everything it runs (you'll want a decent KVM VPS plan to host it), and it works best on scoped tickets in an existing codebase rather than building from scratch. Use it if you've got a backlog of tickets or issues to fix, skip it if you'd rather code side by side with AI in your own editor. #techwithtim #ai #aiengineer
Not rated yetLearn 99% of Codex in One Video
· 47 min
Deploy a Hostinger VPS and connect Hostinger connector to deploy the apps you build with Codex and use code TECHWITHTIM for an additional 10% off: https://www.hostg.xyz/SHKK6 Codex has changed a lot over the past few months, and most people open it, type a prompt, and never touch the features that actually make it good. In this video I walk you through everything in the order you should learn it, and by the end you’ll have Codex properly configured, you’ll know which models and effort levels to use, and you’ll understand the advanced features most people don’t know exist. Want to LEARN how to build useful AI agents that actually get things done? Go here: https://aiagentbuilders.co/pro How to Make Money With Coding instantly: https://techwithtim.net/newsletter 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-aug26 🎞 Video Resources 🎞 Download Codex: https://openai.com/codex/ ⏳ Timestamps ⏳ 00:00 | Overview 01:04 | ChatGPT vs Codex 02:21 | How/Where to Use Codex 03:16 | Pricing & Plans 03:45 | Layout & UI 05:37 | Space/Spaces 06:42 | Other Secondary Features 07:38 | Sessions & Projects 09:43 | Codex Composer 11:36 | Session Sidebar UI 13:02 | Model Selection 16:27 | Effort Levels & Fast Mode 17:48 | Permission Modes 18:35 | Usage & Credits 19:42 | Creating a Plan 22:19 | Connecting Tools (Plugins/Connectors) 26:59 | Building an App 29:53 | AGENTS.md 32:59 | Browser & Computer Use 35:28 | Deploying the App 36:13 | Session & Context Management 39:52 | Skills 43:25 | Subagents & Multi-Agent 45:28 | Goal Mode 46:16 | Automations Hashtags #Codex #OpenAI #ChatGPT UAE Media License Number: 3635141
Not rated yetThe Open Source AI Software Engineer
· 55 sec
What can an AI software engineer actually do? OpenHands (formerly OpenDevin) is the one agent built specifically for coding, and it works right inside your existing GitHub repo. Hand it tickets or issues and it comes back with a completed pull request. In the demo, I connected a small to-do list API and asked it to move everything into a SQLite database and add tags and search, and it produced a ready-to-submit PR with all the changes laid out. It's fully code-focused, with views for code changes, a planner, the full terminal of what it ran, and a browser to watch it work. #techwithtim #ai #aiengineer
Not rated yetThe Most Transparent AI Agent
· 49 sec
Should you use Agent Zero? Reach for it if you want a flexible agent with access to its own computer, and skip it if you just want something running quietly in the background on scheduled tasks. It's one of the most flexible on the list: its own isolated, sandboxed machine means it can tackle complex tasks and write code or run bash commands without touching the rest of your server. It's also very transparent, you can see exactly what it's doing, down to the sub-agent calls. The cons: it can wander off on vague tasks, and it burns a lot of tokens since it writes code for basically everything. #techwithtim #ai #aiengineer
Not rated yetThe New Job of a Software Engineer (It's Not Writing Code)
· 11 min
Try Air in JetBrains IDEs: https://jb.gg/uf1rym Google says about 75% of its new code is AI-generated. Anthropic says more than 90% of its code is written by Claude Code. So why are US software engineering job postings up about 18% compared to this time last year? Because the job itself has changed, and writing code is now a very small part of it. In this video I break down what software engineers actually spend their time doing in 2026, which skills matter far more than they used to, and exactly what you should be practicing if you want to be good at this job over the next few years. Want to LEARN how to build useful AI agents that actually get things done? Go here: https://aiagentbuilders.co/yt How to Make Money With Coding instantly: https://techwithtim.net/newsletter 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-nov26 ⏳ Timestamps ⏳ 00:00 | New Era of SWE 01:06 | The Stats 02:24 | Directing, Not Typing 03:34 | Breaking Work Up 05:42 | Reviewing Bottleneck 06:59 | Owning the Context 08:11 | Being Accountable 09:18 | What to Focus On Hashtags #SoftwareEngineering #AICoding #JetBrains UAE Media License Number: 3635141
Not rated yetThe AI Agent That Writes Its Own Tools
· 59 sec
What makes Agent Zero different from other AI agents? It gets a full Linux machine inside Docker all to itself, and instead of shipping with a fixed list of tools, it writes its own code and runs it in its own terminal when it needs something. It's very coding-centric. In the demo, I asked it to build a research report on what developers said about AI on Hacker News over 30 days, and told it to work as a team, so it spun up three sub-agents: a data agent to pull the info, an analysis agent to parse it, and a writer agent to produce the report, then brought it all back to the main chat. #techwithtim #ai #aiengineer
Not rated yetI Built an Entire AI Ad Campaign in One Chat Window
· 1 min
Normally, creating an ad campaign means bouncing between chat, image generation tools, video editors, and downloads. But by connecting Runway directly to Claude, I can generate product images, turn them into video clips, and refine the results through a single conversation. So this is a look at how the workflow works, how to set it up, and how to create a complete ad campaign without constantly switching tools. Want to try it yourself? Use code TIM50 for 50% off your first month of Runway: https://runwayml.com/ #techwithtim #ai #MadeWithRunway
Not rated yetI Looked Inside my Agents Context Window...
· 18 min
Get started with BrightData's Scraper Studio here: https://brdta.com/twt_context The context window is the single most important thing in how well your AI agent works. It decides how much you pay, how fast it runs, and how accurate your answers are, and most people have no idea what’s inside it or how to look. In this video I show you what fills up your agent’s context, what happens when it gets full, and then run a real test: the same price-shopping job across 5 different tools. Want to LEARN how to build useful AI agents that actually get things done? Go here: https://aiagentbuilders.co/pro How to Make Money With Coding instantly: https://techwithtim.net/newsletter 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-nov26 ⏳ Timestamps ⏳ 00:00 | Overview 00:44 | What's in the Context Window 01:52 | What Happens when Context Gets Larger 02:55 | A Practical Test 05:03 | What we are Scraping 05:56 | Run #1 07:05 | Run #2 08:27 | Run #3 09:23 | Run #4 10:33 | Run #5 11:27 | Scraper Studio Setup & Usage 13:05 | Scoreboard 15:05 | Total Context Usage 16:06 | Self Healing Web Scraping 17:37 | Audit Your Tools Hashtags #ContextWindow #AIAgents #ClaudeCode UAE Media License Number: 3635141
Not rated yetThis AI Can Teach Itself Wrong
· 23 sec
What's the downside of a self-learning AI agent like Hermes? It can teach itself the wrong skills, and then it keeps repeating those mistakes, so it needs a bit more managing than something like Open Claw. It also takes more setup. The rule of thumb: use Hermes if you have recurring workflows you want it to get really good at over time. Skip it if you just want a quick, zero-setup chat assistant or a few cron jobs running on a schedule. #techwithtim #ai #aiengineer
Not rated yetThis AI Agent Gets Smarter Over Time
· 43 sec
Should you use Hermes Agent? Go for it if you'll chat with it a lot and have workflows you want it to learn, because the more you use it, the smarter it gets. It's more complex than Open Claw and needs more interaction, but that's the point: it works great as a personal assistant running your workflows, and as you change things over time it keeps up. After running it for several months, the skills it builds on its own are genuinely impressive and save a huge amount of time. #techwithtim #ai #aiengineer
Not rated yetAI can do more than just generate text
· 44 sec
AI can generate text all day, but what if you could make it actually make structured decisions? Jev lets you build AI-powered decision systems with Python, using Choice to select between options and Score to evaluate them. So this is a quick breakdown of how Jev works, how to send your first request, and how to put everything together in a working example, so you can start building AI systems that do more than just generate text. #techwithtim #python #ai
Not rated yetJev Python Tutorial: AI Decisions in 10 Minutes
· 12 min
Download "The AI Stack I Run My Business On" in collaboration with HubSpot: https://clickhubspot.com/92149a What if you could let AI make structured decisions instead of just generating text? In this tutorial, I’ll show you how to use Jev with Python to build AI-powered decision systems in just 10 minutes. We’ll start with the basics, go through the Jev cheatsheet, and send our first request. Then, we’ll explore two key decision types: Choice and Score. Finally, we’ll put everything together in a complete example with multiple questions so you can see how it works in practice. Want to LEARN how to build useful AI agents that actually get things done? Go here: https://aiagentbuilders.co/yt How to Make Money With Coding instantly: https://techwithtim.net/newsletter 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-oct26 ⏳ Timestamps ⏳ 00:00 Overview 00:42 HubSpot 01:30 Jev Cheatsheet 02:51 Sending a Request to Jev 05:14 Choice Decision Type 07:11 Score Decision Type 10:23 Full Example with Multiple Questions Hashtags #Python #AIDecisions #Jev UAE Media License Number: 3635141
Not rated yetThe AI Agent That Teaches Itself
· 48 sec
What makes Hermes Agent different from other AI agents? It learns. On the surface it looks like Open Claw, but the difference is the core: Open Claw is a messaging gateway with an agent inside, while Hermes is a learning agent with a messaging gateway on top. After it finishes a task, it reviews what worked and saves it as a skill, just a markdown file you can read and edit. Ask it for something like a weekly YouTube idea brief, and it'll research, build the brief, then write its own skill so it remembers how next time and saves tokens. It's basically a self-evolving agent. #techwithtim #ai #aiengineer
Not rated yetHow to Build RAG From Scratch in 50 Lines of Python - Full Tutorial
· 26 min
This video is sponsored by TripleTen, a professional education platform designed to help tech professionals and career changers move into higher-paying tech and AI roles. Explore two TripleTen programs for experienced engineers: AI Systems Engineering (22 weeks) https://tripleten.com/systems-engineer/?utm_source=youtube&utm_medium=video&utm_campaign=techwithtim AI and Machine Learning (20 weeks) https://tripleten.com/ai-machine-learning/?utm_source=youtube&utm_medium=video&utm_campaign=techwithtim Both are for engineers with 2+ years of recent, hands-on production experience and require a short skills audit before you can enroll. Both also include career coaching and mentorship to help you stand out in the job market. Ask a normal AI model “what’s the alarm code at my cafe?” and it has no idea, because your handbook, notes, and PDFs were never part of its training data. It either refuses or makes something up. RAG fixes that, and it’s a lot simpler than most people make it sound. In this video I build a complete RAG system from scratch in plain Python and under 50 lines of code. No LangChain, no pre-built vector database, no closed-source models, and everything runs locally on your own computer, so it’s free and completely private. 🚀 Tools I Use Get 10% off with code techwithtim Openclaw setup: https://www.hostinger.com/techwithtim VPS setup: https://www.hostinger.com/techwithtim10 Wispr Flow (Best AI Dictation): https://ref.wisprflow.ai/TechWithTim-oct26 🎞 Video Resources 🎞 Get all of the code in this video: https://www.aiagentbuilders.co/yt-resources/rag-from-scratch ⏳ Timestamps ⏳ 00:00 | Overview 00:26 | The Problem 01:18 | What is RAG 02:03 | How it Works 03:01 | TripleTen 04:48 | Step 0 - Models 13:14 | Step 1 - Chunking 16:25 | Step 2 - Embedding 18:31 | Step 3 - Searching 21:47 | Step 4 - Calling the LLM 25:13 | Next Steps Hashtags #RAG #Python #Ollama UAE Media License Number: 3635141
Not rated yet