Seasoned.tech with That Mike Hamilton

What are LLM tokens, and why do they affect context windows, AI costs, memory, and even spelling mistakes? This plain-English guide explains tokenization, token IDs, embeddings, next-token prediction, and why any of it matters when you use ChatGPT, Claude, Gemini, or another large language model. 🌐 Website: https://thatmikehamilton.com An AI token is not exactly a word, character, or thought. It is one of the reusable pieces of text a large language model processes. In this episode, Mike follows those pieces from your prompt through the tokenizer and into the model—then connects them to the practical things people actually notice: context limits, usage costs, forgotten instructions, strange spelling mistakes, and confident AI hallucinations. No hype. No graduate degree required. Just a normal-human explanation for people who remember dial-up. ## Chapters 00:00 What are AI tokens? 01:03 Seasoned.Tech intro 01:13 Why tokens matter 02:09 A token is not a word 05:14 How tokenization works 08:04 From token IDs to meaning 11:19 The next-token machine 14:24 The context window 16:31 Tokens, cost, and speed 18:06 Why tokens behave strangely 20:11 How to use tokens in real life 22:40 Tokens are the pieces—not the meaning 24:19 Outro ## More from Mike ▶️ YouTube: https://www.youtube.com/@thatmikehamilton ✉️ Free weekday newsletter: https://www.thatmikehamilton.com/newsletter 🧠 Plain-English AI guides: https://www.thatmikehamilton.com/answers/ 🎤 Speaking: https://www.thatmikehamilton.com/speaking.html 🛠️ Consulting and AI workflows: https://www.thatmikehamilton.com/consulting.html 📘 Facebook: https://www.facebook.com/mikeh222x 💼 LinkedIn: https://www.linkedin.com/in/michael-hamilton-b46bab60/ 📝 Full transcript: [ADD PUBLIC TRANSCRIPT URL AFTER UPLOAD] Subscribe for practical AI and tech explained for people who remember dial-up. #AITokens #LLM #AIExplained

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Artificial intelligence didn’t take over overnight. This Seasoned Tech video traces the modern AI story—from NVIDIA’s gaming chips and Google DeepMind’s AlphaGo to ChatGPT, Claude, Llama, DeepSeek, AI agents, deepfakes, and today’s copyright battles. Get practical AI and technology advice for people who remember dial-up: https://thatmikehamilton.com A few years ago, most people encountered AI through recommendations, fraud detection, autocorrect, and customer-service chatbots. Then ChatGPT arrived and made artificial intelligence personal. But the technology didn’t begin with ChatGPT. In this video, That Mike Hamilton explores: • How NVIDIA became essential to the AI boom • Google DeepMind, AlphaGo, and the transformer breakthrough • Why OpenAI and ChatGPT captured the public’s attention • Microsoft’s investment in OpenAI and the rise of Copilot • Anthropic, Claude, and the AI-safety conversation • Meta’s Llama models and the open-model movement • Midjourney, DALL·E, Stable Diffusion, voice cloning, and AI video • Hallucinations, deepfakes, privacy, security, and copyright • How DeepSeek disrupted assumptions about the global AI race • Why experience and human judgment may become more valuable • What AI agents, private models, and connected tools could bring next The goal isn’t to celebrate every AI announcement—or panic over every prediction. It’s to understand what worked, what went wrong, and where this technology may actually belong. LEAVE YOUR ANSWER IN THE COMMENTS: "What part of the AI story do you think people are getting wrong right now—and which AI company or tool should I break down next?" UX.AI is a one-day summit for founders, creators, marketers, and working professionals who want practical help using artificial intelligence. September 12, 2026 Waco, Texas Details and tickets: https://ux.live Subscribe to the daily newsletter and learn more about Mike’s book, Make the Show: https://thatmikehamilton.com CHAPTERS 00:00 How AI went from invisible to everywhere 01:17 The invisible years 01:44 NVIDIA sets the table 02:16 Google, DeepMind, and AlphaGo 03:05 The transformer breakthrough 03:54 OpenAI and the rise of ChatGPT 05:23 The hallucination problem 06:02 Microsoft joins the AI race 07:30 Anthropic, Claude, and AI safety 08:35 From chatbots to AI agents 08:58 Meta, Llama, and open models 10:02 UX.AI 10:58 AI images, voices, video, and deepfakes 12:10 The unresolved copyright fight 13:28 DeepSeek and the global AI race 14:24 What AI got right 15:11 What AI got wrong 15:55 What happens next 17:21 The advantage humans still have What part of the AI story do you think people are getting wrong? Leave your answer in the comments. I’m That Mike Hamilton. This is Seasoned Tech—AI and technology for people who remember dial-up. Transcript: [Add public transcript URL after publishing] #ArtificialIntelligence #AI #SeasonedTech

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Welcome to OVERTIME with Rusty, the after-hours side of Seasoned.Tech — where the conversation keeps going after the main show ends.

Mike Hamilton and Rusty dig into AI, tech, media, old-school internet memories, creator tools, and whatever else deserves a few extra minutes on the clock.

Watch the full video: https://youtu.be/FszBOGPIjx0

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

AI agents are replacing the chatbot era. Instead of only answering prompts, they can research, prepare, organize, and—if you let them—take action across your work. In this episode, I explain what AI agents actually do, why “prompt engineering” is giving way to delegation, and how to use them without handing the intern the keys to the building. 🌐 Website: https://thatmikehamilton.com📰 Free weekday newsletter: https://www.thatmikehamilton.com/newsletter📚 Plain-English AI guides: https://www.thatmikehamilton.com/answers/📘 Facebook: https://www.facebook.com/mikeh222x💼 LinkedIn: https://www.linkedin.com/in/michael-hamilton-b46bab60/▶️ Subscribe: https://www.youtube.com/@thatmikehamilton?sub_confirmation=1 CHAPTERS 00:00 The chatbot era is over01:14 AI’s Windows 95 moment03:49 Asking vs. delegating10:34 The real risk with AI agents13:55 The Agent Ladder16:41 How to use agents without losing control AI agents can save real time—but only if you understand what they can access, what they can change, and where human approval belongs. Subscribe for practical AI and tech explained for people who remember dial-up. Transcript: add public transcript link here once hosted #AIAgents #ArtificialIntelligence #ChatGPT #AIForBusiness #Productivity #SeasonedTech

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

AI agents, AI hallucinations, reasoning models, multimodal AI, deepfakes, and AI copyright—explained in plain English.

If every new AI announcement sounds important but somehow leaves you less sure what it means, this is for you. No hype. No doom. Just what these AI terms actually mean, where the tools help, where they break, and what deserves your attention.

You do not need to become an AI person. You need to know when AI is useful, when it is confidently wrong, and when somebody is trying to sell you a toaster with a chatbot glued to the side.

Subscribe for practical AI and tech explained for people who remember dial-up.

🌐 Website + daily newsletter: https://thatmikehamilton.com

📝 Full transcript: **[PASTE PUBLIC TRANSCRIPT LINK HERE]**

Today’s episode is brought to you by Positive Vibes.

Get a little encouragement in your feed and inbox: https://getpositivevibes.com

00:00 Intro

00:10 Why AI language is so confusing

01:34 What AI agents actually do

04:51 AI “reasoning” explained

05:28 Sponsor: Positive Vibes

06:24 Reasoning is not judgment

08:35 AI hallucinations: confident but wrong

11:12 What multimodal AI means

13:20 Deepfakes and verifying what you see

16:00 AI and copyright

19:06 The thing that actually matters

21:26 The Seasoned Take

21:59 Outro

#AI #ArtificialIntelligence #TechExplained

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Welcome to OVERTIME with Rusty, the after-hours side of Seasoned.Tech — where the conversation keeps going after the main show ends.

Mike Hamilton and Rusty dig into AI, tech, media, old-school internet memories, creator tools, and whatever else deserves a few extra minutes on the clock.

If you remember dial-up, AOL discs, floppy disks, or waiting way too long for a webpage to load, you’re in the right place.

Watch the video: https://youtu.be/OADm3SiWD_Q

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.