Blog
My writing sits at the intersection of philosophy, technology and investigation.
Covering AI architecture, surveillance, governance, behavioural design, and the way systems built for one purpose end up doing another. Plus the smaller, sharper observations from working with AI under live pressure when, if you are paying attention, the architecture reveals itself.
The recurring lenses are: power and people, truth versus spin, system rhetoric versus lived reality and the occasional rant! Hope you enjoy.
10 August 2026
An A to Z of the human history hidden inside the words we use for AI: from al-Khwārizmī's algorithm and the Arabic ṣifr behind cipher and zero, through Shannon, Turing and Dartmouth, to why the future may have been hiding in the acronym all along, AI to IA.
Read →9 August 2026
Six years on from Rethink the Rainbow, the children's mental health anthology I co-curated in lockdown, a reflection on deplatforming, why most of that work vanished into a digital black hole, and why platform fragility is exactly why I have spent this year building contAIn™, a methodology I own and control.
Read →27 July 2026
The UK government's 'Watch This Space' green paper proposes making some news more prominent in your feed, without yet saying who decides what counts as trustworthy. A personal argument for why the power to decide what is true should not be handed to any government, and why the consultation, open until 31 August, is worth answering.
Read →15 July 2026
Claudisms: the name I gave the moments an AI tells on itself. Not bloopers but evidence, small dated proof that a human has to direct the machine, or the machine directs the human. Why I log what AI gets wrong, or accidentally right.
Read →15 July 2026
A first-hand account of Barnsley's AI and the Future of Young People event: the broken first rung of the job ladder, the town's tech-town push and £800,000 skills fund, women in tech, cognitive debt, and why everyone wants to govern AI but nobody wants to build it.
Read →13 July 2026
What careers in AI are young people in the UK actually being offered? In Barnsley, the government's first AI Tech Town, the pathways on show lead heavily to defence and national security. A look at who is steering it and why it matters.
Read →10 July 2026
Every week brings a new 'best ever' AI model, but the model is not the system, and chasing them is a treadmill. Why the method that sits above the model is the part that lasts, and how contAIn™, the methodology layer that directs AI to the outcome you define, lets the model underneath change without your way of working falling over.
Read →6 July 2026
A predatory journal scraped my Feed Loop preprint, awarded me a doctorate I never earned, and never read a word. What machine-generated academic spam reveals, and why a person deciding what gets through is most of what contAIn™ is: the methodology layer that directs AI to the outcome you define.
Read →21 June 2026
On 18 June the HR industry quietly walked back its own AI hype. AI washing, agent washing, and why the methodology layer that directs AI to the outcome you define is the missing piece the market is circling.
Read →14 June 2026
A day at London Tech Week, from disabled people locked out of a digital-first financial system to the race for AI capability without control. Why the human, and the methodology layer that directs AI to the outcome you define, has to stay in charge.
Read →11 June 2026
The EU AI Act is not the regulation most UK businesses should worry about. This article maps the seven UK legal frameworks that actually govern AI, what the ICO now requires for genuine human oversight, and how contAIn™, the methodology layer that directs AI to the outcome you define, answers each requirement.
Read →10 June 2026
At DevCon 2026 the room circled the missing piece: the methodology layer that sits above the agents. Whose operating system are you actually running, and why the human-authored layer that directs AI to the outcome you define is the one that matters.
Read →6 June 2026
The 'let's connect' post borrows the language of community and authenticity, but the pattern reveals a growth tactic disguised as solidarity. Here is what happens after the algorithm delivers.
Read →4 June 2026
650 developers spent two days circling the same problem: who directs AI, and what governs the human's part? Samantha Maeer explains why the methodology layer they described but could not name is exactly what contAIn already is.
Read →2 June 2026
One saves to the cloud. Two save locally in hidden folders Windows doesn't show you. If your laptop breaks, you could lose everything.
Read →26 May 2026
OpenAI just moved into the advertising business. Here is what that actually means for anyone using AI in their work, and why the 'your conversations are private' reassurance is missing the point.
Read →22 May 2026
I spent a day at SlatorCon London listening for one thing: how the room would define build. The build vs buy debate in enterprise AI strategy surfaced a gap that nobody named. AI governance, data sovereignty, human oversight, and the missing methodology layer.
Read →21 May 2026
An IT consultant lost £83,000 and his marriage by trusting AI without a methodology layer. This article explores what happens when AI runs without a methodology layer and no human in control.
Read →21 May 2026
Without the methodology layer that directs AI, you don't know what you've lost. That's not a theory. It's fifty-five documented sessions and the evidence base behind contAIn™.
Read →21 May 2026
GA4 is free. What it costs you is buried in 9,000 words nobody reads: GDPR data controller obligations, hidden data sharing defaults, ads personalisation across 307 regions, and a liability cap of five hundred dollars. This is what contAIn™ research found when AI read the Google Analytics terms of service.
Read →19 May 2026
Microsoft Recall revealed more than a feature, it exposed how operating systems are becoming behavioural infrastructure. This article examines the governance gap between AI systems that model human behaviour and the organisations that have no methodology layer directing AI oversight.
Read →17 May 2026
Claude's persistent memory builds a psychological profile from your chat history without notification. This article examines what that file contains, why deletion is not straightforward, and what it means when inferred conclusions are stored as facts.
Read →17 May 2026
I never asked for or sat a test, but the Machine profiled me anyway. How Claude assigned me an INFJ personality type without my knowledge or consent, and what it reveals about how AI directs decisions about you.
Read →15 May 2026
In 1950 Turing asked if machines could think. Seventy years later, we're feeding them our secrets, and the methodology layer that directs AI in the workplace remains absent from most organisations. A report from a Veran HR event.
Read →14 May 2026
Not them, obviously. The UK digital ID timeline from Blair to Starmer, what ChatGPT didn't tell you, why 'voluntary' is doing heavy lifting, and why the methodology layer that directs AI matters when governments decide what AI knows about you.
Read →11 May 2026
AI burnout is real and nobody is talking about it honestly. This article logs the failures, the system that emerged from them, and why the methodology layer that directs AI is the difference between capability and overwhelm.
Read →26 April 2026
A first-hand account of AI sales psychology, the gap between what is promised and what is sold, and why no machine works unless the human directs AI toward defined outcomes.
Read →22 April 2026
Five months building AI infrastructure revealed a binary truth: AI fails when you give it discretion. The failure built the methodology, and the human directs the machine toward defined outcomes.
Read →21 April 2026
A day of self-hosting failures, email infrastructure confusion, and platform independence decisions reveals what it actually looks like to build a system you understand and can direct yourself.
Read →18 April 2026
When Claude silently swaps models mid-session, a frustrated human finds an unlikely parallel in Shakespeare's Sonnet 4, proving tangential thinking is the one capability AI cannot replicate.
Read →17 April 2026
AI setup done right still fails when execution fidelity breaks down. This article explores why Claude ignores its own instructions and why the human governance layer that directs AI is the entire point.
Read →14 April 2026
When your gut instinct tells you a critical dependency will fail, overriding that signal has consequences. A lesson in personal governance, pattern recognition, and trusting the feed loop before it costs you the launch.
Read →10 April 2026
I corrected the same wrong line seventeen times before finding it baked into three sources. This is why what you feed AI matters more than how you prompt it, and why a methodology layer that directs AI from defined outcomes fixes what conversation corrections never will.
Read →1 April 2026
Does the internet still have a sense of humour? A real-time check on April Fools Day 2026 found the gap between satire and the actual news has essentially closed.
Read →29 March 2026
The day after a jury found Meta negligent, the company published TRIBE v2, a model that predicts how your brain responds to the stimuli they control. This article maps the overlap.
Read →26 March 2026
Two US juries found Meta's platforms were designed to be addictive, awarding $381 million in damages. The legislation that followed targets users, not the design. If the problem is in the machine, why is the solution pointed at you?
Read →26 March 2026
Substack's age verification partner Persona runs 269 checks including watchlist screening and profiling, not just age verification, exposing a digital feedback loop between government legislation and surveillance technology.
Read →25 March 2026
After half a decade off platforms, Samantha Maeer returns to writing, this time on Substack, with thoughts on deplatforming, biometric verification, and why no platform should feel like home.
Read →The full archive is now on this site. Articles are also published on Substack: substack.com/@samanthamaeer
Also on Medium: medium.com/@samjmaeer
Connect on LinkedIn: LinkedIn
Everyone starts at Ignite. Certify is earned.
Ignite (Start here — free)
The bit the unicorns and rainbows crowd continually skip. What AI actually is and how to make it work for your specific use case.
Activate
The super important basics. Get your house in order so you know what you've got, what's working, and how to improve it.
Steer
The full methodology in play. You set the heading, you read the conditions, you decide where this goes.
Control
Deeper understanding. Learn to build the tracks, system, rules and set the standard everything runs to.
Command
Full depth. Your flight plan, your AIrspace, your call. The soul of how you operate with AI.
Certify
Certification. The full contAIn™ methodology validated. Earned, not bought.
Enterprise
Application only. The full methodology tailored to and delivered within your organisation. [email protected]
By submitting this form you agree to our Privacy Policy.