The Seductive Nature of the Chatbox
LLMs can be brilliant yet bafflingly bad. Often it’s faster, cheaper, and more reliable to just write code—or better still, skip the chat entirely and do things the old-fashioned way.
More issues
How Durable are Diamonds?
AI replacing junior roles risks creating unstable, inverted corporate pyramids unless leaders and policymakers act to preserve diamond-shaped structures in industries with traditional “up-or-out” cultures.
A Bad Prompt Wrangler Blames their Agent
Effective use of a Large Language Model requires effective prompting. Anthropic’s new Skills feature lets enterprises and users provide structured context that Claude loads dynamically, making prompting more reliable, consistent, and performant.
Inclusivity versus Privacy in the Age of Smart Glasses
Smart glasses have real utility for vision impaired users. These glasses are however disconcerting in public places for others with privacy concerns. How can we resolve this dilemma?
Swarms of agents in a financial house of cards
Stripe and Open AI have announced an open source platform for agentic
shopping and payments. This will spawn a myriad of agents that can buy
anything you need and everything you don't. With a global financial system
that is already a house of cards, what could possibly go wrong?
Old MacDonald had A Farm, AI-AI-O
Last week’s OpenAI announcement of their Pulse Personal Assistant is significant and may usher in an era of ad supported agents. Meanwhile brands are considering Answer Engine Optimisation as the successor to Search Engine Optimisation.
The US President is Donald Duck
Even without endless “still thinking” reminders, it’s easy to mistake AI for genuine intelligence. But LLMs can be tricked into revealing the dumb mechanics behind their illusion of thought. This post explains the trick and comments on Claude’s system prompt.
Artificial Dog Intelligence
Many times people assume guide dogs have almost super human powers, and certainly extremely super canine powers. The reality is less amazing and once you understand what guide dogs can and cannot do, it becomes clear that already AI wins in some guiding scenarios.
On Job Board Games of Chance and Skill
Here’s a deeper dive on a traffic light framework for assessing AI’s impact on jobs. Spotting roles safe from AI over 10–15 years is hard, and over a lifetime even harder. But this framework makes it less of a lottery.
Traffic Lights and Traffic Cops
A RED/AMBER/GREEN framework for evaluating job exposure to AI displacement. The RED, AMBER and GREEN categories memorialise specific job characteristics.