Local AI Series
AI is no longer exclusive to data centers and cloud services. With the right software and a decent PC, you can run powerful AI models right from your desktop. This means you can create content, analyze data, and experiment with cutting-edge technology without an internet connection, subscription fees, or privacy concerns.
The future of AI isn’t just in the datacenters or in the cloud; it’s right on your desktop, phones and other endpoint devices. Running AI models locally offers unparalleled privacy, speed, and creative freedom. But with so many options from AMD, Intel, Apple and NVIDIA, how do you choose the best “Local AI PC” for you? Let’s break down the leading architectures and what they bring to the table.
You don’t need a supercomputer to run AI models, but the right software makes all the difference. Articles below talk about Local AI and my own learning journey.
Topics covered in this section include:
Recent Modern EUC Related Posts:
-
Digital Coworkers: The New AI Teammates for Work
By Jorge Pereira • 2026-05-17The workplace is moving beyond chatbots and into a new category of software: digital coworkers. These are AI systems that do more than answer questions; they can execute background tasks, manage files, coordinate workflows, and keep working on complex jobs with less step-by-step prompting. What makes this shift important is simple: instead of asking AI… -
The Rise of the Enterprise Token Broker
By Jorge Pereira • 2026-05-14As enterprises scale their AI operations from experimental “playgrounds” to full-scale agentic workflows, a new bottleneck has emerged: Token Controlling and API Key Chaos. With teams of 6–10 developers or automated agents hitting multiple providers (OpenAI, Anthropic, Gemini) and local servers simultaneously, managing individual accounts is no longer viable. Enter the AI Gateway—the centralized “Token… -
Cloud AI vs Local AI – Cost Comparision
By Jorge Pereira • 2026-05-14Back in 2024 I wrote a blog post: How Much Does It Cost to Operate AI ChatBots? As we move into the new era of the token economy, the conversations, about tokens costs and power are very much part of the story. A useful model must account for real model pricing, utilization, infrastructure, performance, and… -
Windows 11 Taskbar: Now Open to AI Agents
By Jorge Pereira • 2026-05-14Starting with the May 2026 security update, any developer can register AI agents via Windows.UI.Shell.Tasks API to appear alongside Copilot. Users invoke with “@” in search, monitor chain-of-thought via hover. This positions Windows as the OS for agents, not just AI features. In a move bigger than another Copilot tweak, Microsoft is turning Windows 11… -
Small Local Models: Why Tiny AI Is Having a Big Moment
By Jorge Pereira • 2026-05-06The narrative around artificial intelligence has long been dominated by bigger equals better. Models with trillions of parameters, trained on internet-scale datasets, powered by massive GPU clusters—these were the benchmarks of progress. But something interesting is happening at the other end of the spectrum. Small models—measured in billions, not trillions of parameters—are proving they can… -
AI: Don’t Just Chat — Practical Roadmap
By Jorge Pereira • 2026-04-22For the last few months, I’ve been working on a new book series called AI: Don’t Just Chat. I wrote it because the pace of AI has become exhausting for a lot of people. The tools change constantly. The advice shifts. What worked a few months ago can already feel outdated. Even people who are deep… -
Agent Zero FAISS Memory Error: What It Means, What to Keep, and What to Reset
By Jorge Pereira • 2026-04-21To learn more about Local AI topics, check out related posts in the Local AI Series If you are using Agent Zero and suddenly see this error, you are not alone: “`text ValueError: Could not find document for id 6oqX1vwBxV, got ID 6oqX1vwBxV not found. “` This problem can show up repeatedly, sometimes with different document… -
What Kind of Computer Do I need to run Gemma 4 Locally
By Jorge Pereira • 2026-04-05I just posted yesterday Local AI Sovereignty: Deploying Ollama, Gemma 4, OpenWebUI, and n8n and I used Gemma 4 locally on Ollama. Someone asked me a good question: What size computer (Windows PC) do I need tor it? What size computer do you need for Gemma 4? Google’s April 2026 release of Gemma 4 changed… -
Local AI Sovereignty: Deploying Ollama, Gemma 4, OpenWebUI, and n8n
By Jorge Pereira • 2026-04-04With 64GB of RAM and the latest Ryzen AI silicon, you are no longer a mere consumer of AI—you are a host. This setup leverages AMD’s XDNA architecture to run Gemma 4 and / or and Qwen 3.5 locally, ensuring your data never leaves your machine while providing a professional-grade automation suite via Docker. This… -
What Is NemoClaw / OpenClaw and How Are They Different?
By Jorge Pereira • 2026-03-30AI agents are moving fast from demos to real workflows. Two names that keep coming up are OpenClaw and NemoClaw, and they are often mentioned together because NemoClaw is built as a security-focused layer around OpenClaw. What is OpenClaw? OpenClaw is an open-source AI agent framework that runs on your own hardware and connects AI… -
OpenClaw its Good but AgentZero is my choice
By Jorge Pereira • 2026-03-15OpenClaw (formerly known as Clawdbot, Moltbot). It’s remarkable how quickly a tool can become indispensable once you integrate it into your daily routine. I still remember my first week using it—curiously has taken the productivity world by storm, and for good reason! With its impressive range of features, it has captured the attention of countless… -
Streamlining Agent Zero: Skill Errors Back to Basics
By Jorge Pereira • 2026-03-11When I first started experimenting with Agent Zero, I wanted to keep things totally “temporary” and isolated from my own data. My mindset was simple: I didn’t want the Docker container saving any clutter to my computer. I wanted it to start fresh every time. For a few days, this worked well. Agent Zero was… -
My Journey to a Self-Hosted Web Search
By Jorge Pereira • 2026-02-21For months, I’ve been refining my local AI lab. I had the hardware dialed in—my AMD Ryzen AI processor, 128GB of RAM, running Ollama and Open WebUI like a dream. But I kept hitting the “invisible wall.” Every time I asked my local agents for the morning’s technical headlines or a deep dive into a… -
Building Your Own Tech Stock Trading Agent
By Jorge Pereira • 2026-02-15I started an experiment today… We’ve all seen the flashy trading bots promising “1000% returns,” but usually, those come with a hefty subscription fee and a black box where your data goes to die. If you’re already running a local agent (via Ollama, LM Studio, AgentZero or a custom build one), you’re sitting on untapped…
…