Writing on technological change, scientific authority, and their organisational and social consequences—separating genuine progress from inflated promise.
34 essays
Most companies use AI, but few achieve real impact. In part I of this series I explore why people—not tools—determine AI maturity, and why the next serious organisations will dominate the space between pilots and profit.
AI models mirror WEIRD cultural norms while claiming universality. Alignment methods reinforce this bias, projecting one community’s “common sense” as global truth. Recognising plurality through Wittgenstein’s language-games and Popper’s falsification reframes alignment as provisional and adaptive,
AI models often mirror our beliefs, rewarding us with agreeable but shallow answers. This sycophancy flatters rather than challenges, eroding judgment and candour. To gain true value, leaders must set incentives that favour truth over comfort, design prompts that demand trade-offs, and treat AI as a
Clicks are dying, and with them provenance, nuance, and the economics of ideas. AI summaries reward skim over substance. We need friction, attribution, and long-form thinking—treating summaries as aperitifs, not meals—so leaders decide from sources, not headlines.
Advances in LLMs show promise, but token bias undermines their logical reliability. Small input shifts can distort outputs, posing risks in fields like medicine, law, and policy. Their dependence on pattern recognition over true reasoning demands closer scrutiny and better design.
By employing leaders capable of creating an AI framework — because they are awake and aware to the unintended effects of AI on social well-being, data integrity and privacy, diversity, and governance — organisations seeking to transform into being AI-first are well positioned to engage in trustworth
Because being off the grid is next to impossible in many societies, it is tempting to give up and proclaim 'I have nothing to hide'.
There is the case to be made that this is not only the cost of doing business in the digital age but that, like environmental protections, there are great rewards to be reaped in improved quality of life.
As Apple's privacy respecting posture gives way to the pull of marketing dollars through making user data the product, their third way begins to fall short.
The hope is much, for having gotten this far is to be forewarned and thus forearmed. In that we do well to employ scepticism when listening to a human interlocutor. Because even the best of us are filling in the blanks in our memory.
Ultimately, I do not think the problem is that we are building machines so ‘smart’ that their output is indistinguishable from human compositions. The problem is that we are educating humans whose output is so illiterate, and devoid of experience, it is indistinguishable from computer generated text
You can join me in camp confusion or post using or not using CWs with a certainty that would make a prophet blush.
Little did I know how consistently Firefox would be my window to the World Wide Web, nor how its security, and in time privacy focus, would shape my understanding of what it means to be online.
Whether you have something to hide has far more to do with who is looking than what they can see.
I wonder if there is an alternative reading in which they did not make a conscious decision to be 'evil'?