Author | Lightnews

Ian O'Byrne @wiobyrne.bsky.social · 12d

This week in tech:

✗ AI lesson plans: 90% memorization
✗ Meta AI: Your chats = their ads
✗ OpenAI: Subpoenaing safety advocates
✗ ICE: Immigration tools → protest surveillance

The efficiency promise is a trap.

digitallyliterate.net/dl-408/

2

Ian O'Byrne @wiobyrne.bsky.social · 14d

From a dusty college Vaio to a full-blown Plex media hub, my first DIY server taught me: backups > performance, constraints breed learning, and self-hosting builds confidence. 🖥️💡

Control your tech. Build your freedom.

wiobyrne.com/my-first-diy...

#DIYServer #Homelab #DigitalSovereignty #Plex

From Subsonic to Plex: My First DIY Server | Ian O'Byrne

wiobyrne.com

1

Ian O'Byrne @wiobyrne.bsky.social · 19d

The question isn't whether surveillance infrastructure is expanding. The sea is rising, but we're building.

🏝️ Digital Archipelagos of trust are forming. Where is your island?

Which one are you building?

Read the full analysis in Digitally Literate #407: digitallyliterate.net/dl-407/

Building Archipelagos in Surveillance Seas

digitallyliterate.net

1

Ian O'Byrne @wiobyrne.bsky.social · 20d

The quiet expansion of surveillance in education is dangerous.

Convenience must not bypass due diligence. Demand transparency, governance, and the right to privacy.

wiobyrne.com/privacy-isnt...

#StudentPrivacy #EdTech #DigitalLiteracy

Privacy Isn’t Optional: The Expanding Data Dragnet in K-12 and Higher Education | Ian O'Byrne

wiobyrne.com

1

Ian O'Byrne @wiobyrne.bsky.social · 20d

Mobile Campus IDs: Convenience at the cost of data?

If we’re moving from plastic to a constant data trail of who, when, & where on campus, we MUST prioritize governance, data retention policies, and opt-out options. Convenience shouldn't bypass due diligence.

wiobyrne.com/when-your-ca...

When Your Campus ID Moves to Your Phone: Questions We Should Be Asking | Ian O'Byrne

wiobyrne.com

Ian O'Byrne @wiobyrne.bsky.social · 21d

I'm not just documenting my homelab. I'm publishing the "Director's Cut" of how I wrote the whole thing with AI. 🤯

It’s about thinking alongside AI, not just typing with it.

Read the first part:https://wiobyrne.com/how-i-wrote-this-with-ai/

#AI #Homelab #Writing #Collaboration #ChatGPT

1

Ian O'Byrne @wiobyrne.bsky.social · 22d

Why I built a homelab: It's not a hobby, it's an act of digital literacy and reclaiming ownership.

Building my own "cloud" is how I answer the question: What does it mean to truly understand the systems we rely on?

wiobyrne.com/why-i-built-...

#homelab #digitalagency #selfhosting #tinker

Why I Built a Homelab (and Why You Might Want To) | Ian O'Byrne

wiobyrne.com

2

Ian O'Byrne @wiobyrne.bsky.social · 26d

Tech giants now openly frame ethics as “competitive disadvantages.”

Apple, Google, OpenAI, Meta—this week shows the pattern: profit > protection, every time.

DL 406 → digitallyliterate.net/dl-406/

1

Ian O'Byrne @wiobyrne.bsky.social · 26d

Got LM Studio running, but wondering “Now what?” 🤔

Local AI isn’t just tinkering. It’s privacy, control, and efficiency.

Your AI sandbox is ready. The toys are yours. What will you build? 🛠️
wiobyrne.com/mastering-yo...

#LocalAI #AI #HuggingFace #Privacy #Productivity

Mastering Your Local AI: Beyond the Basics | Ian O'Byrne

wiobyrne.com

2

Ian O'Byrne @wiobyrne.bsky.social · 28d

You can now run AI models locally.

- No internet needed
- Zero monthly fees
- Complete privacy
- Works on regular laptops

Trade-off: Slower responses, technical learning curve

But your conversations never leave your computer.

I wrote a beginner's guide 👇
wiobyrne.com/getting-star...

How to Run Your Own AI Models: Getting Started with LM Studio | Ian O'Byrne

wiobyrne.com

Ian O'Byrne @wiobyrne.bsky.social · Sep 28

Machines aren’t just tools; they’re becoming the standard.

When tech leaders prioritize control over care, society feels the consequences.

Let’s explore the stakes: digitallyliterate.net/dl-405/

Ian O'Byrne @wiobyrne.bsky.social · Sep 21

AI systems systematically become more unethical when stressed.

Platforms self-censor when politically targeted.

People withdraw when harassment campaigns intensify.

The pattern is identical across all systems: apply pressure → watch values collapse.

digitallyliterate.net/dl-404/

When Things Fall Apart

digitallyliterate.net

Ian O'Byrne @wiobyrne.bsky.social · Sep 14

Success requires being visible, but visibility attracts harassment.

The internet promised everyone a voice, but only those willing to accept abuse get to use it safely.

We need digital spaces where meaningful engagement doesn't require armor. 🛡️

digitallyliterate.net/dl-403/

Reposted by Ian O'Byrne

Hypervisible @hypervisible.blacksky.app · Sep 11

Nodding along with this until about half way through, where it goes so off the rails that the proposed solutions are ableist, classist, and carceral in ways that are arguably worse than the current problems.

The Question All Colleges Should Ask Themselves About AI

How far are they willing to go to limit its harms?

www.theatlantic.com

3 2 44

Ian O'Byrne @wiobyrne.bsky.social · Sep 7

The internet is splitting:
🔒 Safety = more surveillance
🛑 Agency = more friction

Platforms, schools, and courts are choosing sides.
The real question: will we use tech on our terms, or theirs?

Full issue → digitallyliterate.net/dl-402/

The Great Fracturing

digitallyliterate.net

Ian O'Byrne @wiobyrne.bsky.social · Sep 3

ChatGPT is OpenAI’s playground.
Running AI locally = building your own sandbox.

More control. More privacy. More transparency.

Your AI. Your rules.
👉 wiobyrne.com/your-persona...

Your Personal AI Sandbox: What It Really Means to Run Models Locally | Ian O'Byrne

wiobyrne.com

Ian O'Byrne @wiobyrne.bsky.social · Sep 2

AI isn’t a black box.
A full conversational model fits in 760MB & runs on your laptop.

Same architecture as ChatGPT—just scaled down.
Meet TinyLlama 🦙: small, open, powerful.

👉 wiobyrne.com/cracking-ope...

Cracking Open AI Models: What AI Intelligence Actually Looks Like | Dr. Ian O'Byrne

wiobyrne.com

Ian O'Byrne @wiobyrne.bsky.social · Sep 1

AI isn’t magic — it’s math.

When you download a model, you’re really just getting 3 files:
🧠 weights (numbers)
📐 config (blueprint)
🔤 tokenizer (translator)
That’s it.

Billions of numbers, carefully arranged, become “intelligence.”
Wild, right?

wiobyrne.com/inside-an-ai...

What’s Actually Inside an AI Model? (It’s Not What You Think) | Dr. Ian O'Byrne

wiobyrne.com

Ian O'Byrne @wiobyrne.bsky.social · Aug 31

I'm convinced the biggest AI risk isn't machines getting too smart. It's us getting too comfortable with cognitive shortcuts.

New analysis: Why "useful friction" might be the most important concept for our automated age.

digitallyliterate.net/dl-401/

#DigitalLiteracy #AI #CriticalThinking

Useful Friction

digitallyliterate.net

Ian O'Byrne @wiobyrne.bsky.social · Aug 30

New AI tools drop daily. 🚀 But before diving in, ask: Does this tool fit my purpose, protect my data, and align with my values?

I share 7 key questions to help you choose AI tools with intention, not hype: wiobyrne.com/decisions-to...

Seven Critical Decisions to Make When Trying New AI Tools | Dr. Ian O'Byrne

wiobyrne.com

1

Ian O'Byrne @wiobyrne.bsky.social · Aug 28

My latest post hits a crucial point: we can't just be critics of technology. We have to be active participants in building a better digital world.

The real question isn't whether to use AI, but how to use it ethically.

wiobyrne.com/reclaiming-o...

#AI #TechEthics #DigitalLife #HumanCenteredTech

Reclaiming Our Tools: Why We Can’t Opt Out of the AI Future | Dr. Ian O'Byrne

wiobyrne.com

2

Ian O'Byrne @wiobyrne.bsky.social · Aug 24

🚨 DL400 is here: Knowledge as Living System 🌱

➡️ Knowledge isn’t accumulated. It’s connected.
➡️ AI isn’t destiny. It’s a human project we must shape.
➡️ The future of publishing isn’t content streams. It’s living knowledge systems.

🔗 Read DL400 here: digitallyliterate.net/dl-400/

From Archive to Architecture

digitallyliterate.net

Ian O'Byrne @wiobyrne.bsky.social · Aug 23

Tech tells us “everything is optimizable.” But when we erase ambiguity, we erase humanity.

Efficiency isn’t free. It hides labor, relationships, & care.

wiobyrne.com/the-myth-of-...

The Myth of Optimization: What Tech Worldviews Leave Behind | Dr. Ian O'Byrne

wiobyrne.com

Ian O'Byrne @wiobyrne.bsky.social · Aug 22

I fed my entire Obsidian vault (500+ notes) to an AI. It analyzed my knowledge system and spotted things I never knew were there. 🤯

Read what happened: wiobyrne.com/ai-analyzed-...

#Obsidian #AI #KnowledgeManagement

The Day AI Analyzed My Entire Knowledge System: MCP + Obsidian in Action | Dr. Ian O'Byrne

wiobyrne.com

1

Ian O'Byrne @wiobyrne.bsky.social · Aug 17

Plot twist: After 398 newsletters, I realized the format itself was the problem

Linear timelines bury good ideas under timestamps

Digital gardens let ideas breathe & connect across time

Newsletter → living knowledge system

Chronology is a choice, not a law ✨

digitallyliterate.net/dl-399/

From Archive to Architecture

digitallyliterate.net