Ready to transform artificial intelligence from a buzzword into your personal revenue generator?

HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.

Inside you'll discover:

Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.

Get Your Guide

Last week, we had a rummage through the EU’s tech sovereignty plans, and this week we’re bringing things a bit closer to home: the undeniable urge to run AI on your own PC. Enter Odysseus, launched by none other than PewDiePie (yes, you read that right) in mid-2026. It’s free, open-source, and entirely self-hosted, think of it as the privacy-obsessed cousin of ChatGPT and Claude, but without the monthly bill nibbling away at your wallet. Odysseus is all about 'local-first', so there’s no data telemetry and everything runs on your own hardware. You keep full control over your prompts, data, and workflows, no need to surrender your digital life to the cloud. And it’s not just another chatbot, either. Odysseus is more like a Swiss Army knife for AI: you get autonomous agents, model comparisons, hardware-aware serving, a document editor, and tools for deep research and email management, all under one roof. Not bad for something that won’t cost you a penny.

But first, let’s have a look at what’s been happening around the web.

Leviathan Waking

Microchip PIC32CM GC Curiosity Pro - DK Unboxing

Predicting model behavior before release by simulating deployment

Poor man’s Polaroid camera

On Transparent AI Cyber Protections

GLM-5.2: Built for Long-Horizon Tasks

AI-referred US shoppers browse longer, spend more per visit, data shows

The AI layoff wave is becoming a powder keg

There’s a definite shift afoot in the digital world: people are quietly tiptoeing away from cloud-based AI and towards running things locally. It’s not just a matter of taste, these days, your personal data is worth its weight in gold, and nobody wants to hand it over to the tech giants if they can help it. Digital sovereignty (that’s owning and running your own AI, for those keeping score at home) has gone from being a niche concern to something every professional should care about. Odysseus was born because, let’s be honest, the self-hosted AI world was a bit of a shambles, plenty of clever models, but the user interfaces were about as friendly as a tax return. What started as a developer’s side project has grown into a serious tool for getting work done. It’s all part of a bigger trend: people want a workspace that actually respects their intelligence and doesn’t make them juggle a dozen different apps. Who knew a bit of tinkering could turn into a manifesto for digital independence?

Odysseus is built around a simple idea: your data should stay yours. Privacy isn’t just a nice-to-have here, it’s the whole point. In the usual cloud setup, every time you type something, it’s spirited away to some corporate server, where it might be traded, leaked, or just plain lost. Odysseus turns that on its head. No tracking, no subscriptions, and no need to trust anyone with your secrets. You can feed it your most sensitive data, bank statements, work projects, embarrassing poetry drafts (we’ve all been there), and it all stays safely on your own machine. That’s the real value: keeping your digital life out of the hands of companies who’d love to turn it into their next revenue stream.

Swapping a jumble of apps for one tidy workspace is a big win for anyone’s sanity. Odysseus acts as your minimalist command centre: email, calendar, notes, and even image editing all living together in harmony. The real magic, though, is in its 'agent', think of it as your digital butler, only less likely to judge your file-naming habits (no more 'final_final_v2.docx' shame). This agent can hunt down files across your machines, convert videos, transcribe audio, and generally do the boring bits for you. It even learns to get better at its job over time. For those who fancy a bit of philosophical debate with their workflow, there’s a 'Council' feature, where you can have Socrates, Nietzsche, and friends weigh in on your latest problem. It’s all about making your digital life less fragmented and a lot more clever.

A good AI isn’t just clever, it needs to remember what you actually care about. Odysseus dodges the usual 'text dump' approach by giving you a 'Brain' that stores memories from your chats. There’s also a 'Skills' module, so you can check how well it’s doing (complete with confidence scores, if you’re partial to a bit of number crunching). The 'Deep Research' feature is where things get interesting: it can do proper, multi-step web research and spit out visual reports with citations, not just a wall of text. There’s even a document editor inspired by Claude, so you can write and format as you go. The idea isn’t to replace you with a robot overlord, the AI is there to help you fact-check, spell-check, and polish your work, while you stay firmly in the driver’s seat (no robot uprising scheduled for this week).

Of course, running AI locally means you’ll need to wrestle with your hardware. Setting up your own sovereign workspace does require a bit of technical know-how, Python, Git, and local model providers like Ollama or VLLM are all part of the fun. For those who’d rather not spend their weekend trawling Stack Overflow, Odysseus comes with a 'Cookbook' that scans your hardware and suggests the best models for your setup. No need to worry about confusing jargon like quantisation or parameter counts; you just get a simple score to help you choose. There is a trade-off, though: running big models locally can make your computer sweat (sometimes literally), and things might slow down if you try to do too much at once. On the plus side, there’s a 'sensitive blur' feature that hides your API keys and secrets during screen sharing, because nobody wants to be the person who leaks their credentials on a Zoom call (we’ve all seen the horror stories).

Odysseus isn’t just another pile of productivity tools cobbled together for the sake of it. Think of it more as a rallying cry for anyone tired of Big Tech poking about in their digital sock drawer. Here, you get a private, flexible ecosystem that actually lets you hold the keys to your own data. Everything is free and local, so you don’t have to swap your privacy for a workspace that looks pretty but leaks like a sieve. As more people jump on board, Odysseus is shaping up to be a genuine force for democratising high-level AI. If you’ve got the hardware, you’ve got the power, no need to mortgage your soul to Silicon Valley. The days of quietly handing over your data are numbered; the future belongs to those who keep their digital house in order.

The arrival of Odysseus marks a proper turning point in personal computing. We’re moving away from the era of ‘rented intelligence’, you know, paying someone else to think for you, towards a future where you actually own your digital life. Odysseus, being self-hosted and open-source, gives a polite but firm nudge to the old habit of relying on centralised cloud providers. Instead, you get a single space where chat, research, and automation can all get along nicely, without anyone peering over your shoulder. For the modern professional, this isn’t just a technical curiosity; it’s a strategic weapon for reclaiming that precious first hour of the workday from the clutches of mindless scrolling. By putting your own hardware at the centre of things, Odysseus offers a safe haven for your sensitive intellectual property, keeping your logic well away from third-party training data or the prying eyes of state-level data harvesters.

At the heart of Odysseus is its rather clever dual-engine setup, which means you’ll need to balance the joys of local sovereignty with the raw muscle of the cloud. No single engine ticks every box (yet), so you get to play a bit of a balancing act. On one hand, you’ve got the total data sovereignty of a local engine like Ollama; on the other, the cutting-edge smarts of cloud-based APIs. If you do venture into the cloud, your choice of provider, OpenRouter or Venice, sets your privacy profile. OpenRouter gives you access to a whole zoo of models, while Venice is the privacy purist’s pick, thanks to its strict ‘Zero Storage’ policy. Nothing gets kept on their servers, which is rather reassuring. Meanwhile, if you stick with Ollama, nothing ever leaves your machine, so your proprietary data stays locked up tighter than the biscuits in the staff room.

Of course, all this cleverness depends on what you’ve actually got under the bonnet. Plugging into a cloud provider is as easy as popping in an API key, and it’ll work just as well whether your laptop is brand new or old enough to remember Windows XP. But if you’re running local models with Ollama, your hardware suddenly matters, a lot. If you’ve got an Apple Silicon machine or a beefy NVIDIA GPU, you’ll get lovely, sophisticated answers from models like Qwen 2.5 7B or Gemma 4. If you’re on a bog-standard laptop, expect a bit of a wait and the occasional ‘that’s all you’re getting’ summary. There’s also the age-old question: do you want to pay per use for cloud APIs, or splash out up front for decent hardware and then watch your electricity bill creep up? This engine choice is what lets you mix and match, use the cloud for the heavy lifting, and keep the sensitive stuff local, just the way you like it.

In a high-stakes professional environment, the value of information is determined by synthesis rather than mere retrieval, and Odysseus elevates this process through its Deep Research engine. This feature moves beyond the reactive search functionality of a standard browser to become an active decision-making engine. By automating the extraction and analysis of raw data, the system allows a professional to move from a complex inquiry to a structured, actionable strategic report. This multi-step reasoning process iterates through search rounds, determining when it has accumulated sufficient evidence before synthesising findings into a visual report. These reports, which can be exported as PDF or HTML files for "bedtime reading," transform dozens of disparate URLs into a cohesive intelligence product complete with citations.

Of course, how well this research works depends on how you set up your tools. Choose your internet interface wisely: DuckDuckGo is free and functional, but let’s be honest, it can be a bit ‘janky’ compared to the posh results you get from Brave (though you’ll need an API key for that). The quality of your final report also hangs on the model you pick, frontier models like Minimax M3 or Gemini 1.5 Pro are often needed if you want to avoid shallow summaries or the occasional AI daydream. Get everything set up just right, and you’ll turn one-off research into a regular intelligence cycle, with your AI acting as a diligent investigator, serving up evidence for your next big decision. Once you’ve got this running, it’s only natural to start thinking about automating the whole process in the background.

Moving from reactive to proactive AI is where things get interesting. Enter ‘Scheduled Tasks’, think of it as your productivity bouncer, keeping things running smoothly while you’re off making tea. In the old way of working, you had to start every interaction yourself, but with Odysseus, background automation means your digital assistant gets on with things without needing constant supervision. This is what sets a high-value workspace apart: you can have ‘Morning Intelligence Briefings’ ready and waiting before you’ve even had your first coffee. Just set up your triggers and how often you want things checked, and Odysseus will keep you briefed and ready to go the moment you log in.

Setting up these tasks is a bit like telling a story: you decide what you want, when you want it, and how you’d like it delivered. Unlike the usual suspects like ChatGPT or Claude, which can’t schedule things in the background, Odysseus gives you three ways to get your results: right in the interface, as a system notification, or straight to your inbox. The email option is especially handy, as it means your AI’s handiwork follows you onto your phone, so you can catch up on the go. This kind of autonomy means you get a custom briefing that actually fits your working life, with no need for extra subscriptions or fiddly workarounds. These automated tasks can even handle your most sensitive data, changing the way you deal with your daily deluge of emails.

Let’s face it, most of us are drowning in ‘email fatigue’, the inbox feels more like a public dumping ground than a private filing cabinet. Odysseus tackles this by using standard IMAP and SMTP protocols, so the AI can treat your email client as a proper data asset. Inside your private workspace, the AI can read, sort, flag, and summarise those endless threads, making your relationship with your inbox a lot less fraught. If you’re the privacy-conscious type (and who isn’t these days?), you’ll need to set up a bit of extra security: connecting to Gmail, for example, means using a 16-digit App Password, so your main credentials stay safely tucked away on your own machine.

The real magic here is in the ‘High-Priority Triage’ workflow. Odysseus can scan your inbox every 15 minutes, sorting emails into buckets like important, urgent, or just plain spam, and only bothering you when something actually needs your attention. This means you can get on with the good stuff while the AI drafts replies in your own style. Of course, there’s a privacy decision to make: do you want the speed and shine of a cloud model like Minimax, or the total security of keeping things local with Ollama? When it comes to your private inbox, you’ll have to decide if the cloud’s extra brainpower is worth the risk, or if local intelligence is good enough to keep things ticking over. Once you’ve got this sorted, you’re ready for the next wave of creative and document-focused features.

Odysseus isn’t just about emails and research, it brings all your creative work under one roof, from image editing to long-form writing, and acts as your ‘human-in-the-loop’ partner. The ‘Documents’ feature gives you a collaborative writing space: you bring the ideas, and the AI helps with tone, structure, and turning your bullet points into actual sentences. Meanwhile, the ‘Gallery’ is your local diffusion server, with tools for background removal, and upscaling. The idea is simple: the AI doesn’t replace you, it just helps you polish your work, all within your own secure environment.

Of course, all this creative magic relies on a few behind-the-scenes helpers, think Python libraries and background tools like RMBG for background removal, Playwright for browser automation, and Tmux for keeping tasks ticking along. You’ll also need to be realistic about your hardware: basic text is fine on most machines, but if you want to do serious image processing or in-painting, you’ll need something beefy, like a Mac Studio M3 Ultra or a multi-GPU ‘monster’ setup. If your hardware isn’t up to scratch, expect things to hang or freeze (and possibly a few choice words from you). All these creative outputs eventually feed into the system’s evolving cognitive core, which remembers your work to make things even smoother next time. Odysseus workspace resides in its ability to develop a "Digital Second Brain" through persistent cognition. While most AI interactions are ephemeral, Odysseus utilises a "Brain" and "Skills" architecture to extract and store user preferences, project context, and specialised terminology from unstructured documents. This transforms a generic tool into a specialised assistant that remembers everything from professional formatting styles to personal preferences. The system's memory remains entirely offline, representing the ultimate victory for the privacy advocate by ensuring that a user's life-long context is never uploaded to a corporate server.