Skeptics Almanac

"Extraordinary claims require extraordinary evidence"

A collection of sources I’ve come across as I wrestle with the hype vs reality of A.I.

• Microsoft boffins figured out how to break LLM safety guardrails with one simple prompt

• AI boom could falter without wider adoption, Microsoft chief Satya Nadella warns

• Signal’s Whittaker on privacy in the age of data and AI

• Majority of CEOs report zero payoff from AI splurge

• 87% of execs are using AI on the job, compared with just 27% of employees

• The AI Shift: where are all the job losses?

• Salesforce’s Benioff calls for AI regulation, says models have become ‘suicide coaches’

• How AI slop is clogging your brain

• Just 2% of Canadian businesses got return on generative AI investments, survey shows

• AI Darwin Awards Show AI’s Biggest Problem Is Human

• AI agents are coming for your privacy, warns Meredith Whittaker

• AI-Generated “Workslop” Is Destroying Productivity

• Companies are blaming AI for job cuts. Critics say it’s a ‘good excuse’

• When AI Hype Meets AI Reality: A Reckoning in 6 Charts

• Large language mistake

• I got an AI to impersonate me and teach me my own course – here’s what I learned about the future of education

• Something Extremely Scary Happens When Advanced AI Tries to Give Medical Advice to Real World Patients

• OpenAI admits AI hallucinations are mathematically inevitable, not just engineering flaws

• Against Generative AI

• AI Is the Bubble to Burst Them All

• Super-intelligence or Superstition? Exploring Psychological Factors Underlying Unwarranted Belief in AI Predictions

• The AI-Ready Software Developer #13 – *You* Are The Intelligence

• Findings from DX’s 2025 report: AI won’t save you from your engineering culture

• AI Capabilities May Be Overhyped on Bogus Benchmarks, Study Finds

• The state of AI in 2025: Agents, innovation, and transformation

• Investors expect AI use to soar. That’s not happening

• Where's the Shovelware? Why AI Coding Claims Don't Add Up

• Yes, Maintainability Still Matters in “AI”-assisted Coding

• Google’s Gemini AI will use what it knows about you from Gmail, Search, and YouTube

• How AI Is Changing Higher Education

• Meet the AI workers who tell their friends and family to stay away from AI

• Vibe Coding Fails

• AI Agent, AI Spy

• "AI" is bad UX

• Report Finds That Leading Chatbots Are a Disaster for Teens Facing Mental Health Struggles

• Context Widows

• Researchers Just Found Something That Could Shake the AI Industry to Its Core

• Salesforce regrets firing 4000 experienced staff and replacing them with AI

• Agent Psychosis: Are We Going Insane?

• The AI-Ready Software Developer #20 – It’s The Bottlenecks, Stupid!

• AI companies will fail. We can salvage something from the wreckage

• Cursor used agents to write a browser, proving AI can write shoddy code at scale

• Whether they are building agents or folding proteins, LLMs need a friend

• Evidence Grows That Google’s AI Overviews Have Eviscerated the Media Industry

• Sixty years of learning the same lesson

• Coding Is When We’re Least Productive

• How AI assistance impacts the formation of coding skills

• Am I Anti-AI? No. I’m Anti-Harm.

• Gary Marcus on the Massive Problems Facing AI & LLM Scaling

• critical AI literacies for resisting and reclaiming

• Moltbook is about as fun as watching two Roombas bump into each other

• Against the Uncritical Adoption of 'AI' Technologies in Academia

• We Studied 150 Developers Using AI (Here’s What's Actually Changed...)

• Elon Musk gives less than a year to coding as a profession

• Thousands of CEOs just admitted AI had no impact on employment or productivity—and it has economists resurrecting a paradox from 40 years ago

• AI World Clocks

• The AI-Ready Software Developer #10 – Comprehension Debt

• Agentic AI Coding: Best Practice Patterns for Speed with Quality

• The AI Treadmill

• The AI Data Center Financial Crisis

• We've been here before

• AI Productivity Gains in Different Situations

• The Hater's Guide to Anthropic

• The Worst-Case Future for White-Collar Workers

• ‘It’s destroyed me completely’: Kenyan moderators decry toll of training of AI models

• Unified Bug Report: Claude Code Agent Systematic Failure Patterns

• Tech Has Never Caused a Job Apocalypse. Don’t Bet on It Now

• I Am An AI Hater

• ‘Unbelievably dangerous’: experts sound alarm after ChatGPT Health fails to recognise medical emergencies

• Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?

• AI Translations Are Adding ‘Hallucinations’ to Wikipedia Articles

• Blowhard ChatGPT bot posed as lawyer, convinced woman to fire her real attorney — while citing phony ‘case law’: suit

• AI Doesn’t Reduce Work—It Intensifies It

• Lower Artificial Intelligence Literacy Predicts Greater AI Receptivity

• 1984, but with LLM’s

• GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models

• Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task

• Operating Multi-Client Influence Networks Across Platforms

• A knockout blow for LLMs?

• AI 2027

• Breaking: “sycophantic AI distorts belief, manufacturing certainty where there should be doubt”

• Yeah, About Your CLAUDE.md file…

• AI Agents have, so far, mostly been a dud

• AI Hallucination Cases

• AI Therapy Bots Are Conducting 'Illegal Behavior,' Digital Rights Organizations Say

• AI, layoffs, productivity and The Klarna Effect

• AI False Information Rate Nearly Doubles in One Year

• Bad Actors are Grooming LLMs to Produce Falsehoods

• Botshit Gone Wild

• The AI Dilemma

• AI can now stalk you with just a single vacation photo

• Commoditization 2.0: How Agentic AI Could Undermine The Open Web’s Best Publishers

• Complete Accuracy Collapse

• Decoding (and debunking) Hard Fork’s Kevin Roose

• Diabolus Ex Machina

• Does Open Source AI really exist?

• General Purpose is an Accountability Loophole

• Distilled summary of 2024/2025 Google DORA Report

• Does AI Actually Boost Developer Productivity?

• Empire of AI: Karen Hao & Roger McNamee Expose Sam Altman, Musk & Big Tech

• Excerpt on Multi-Agent Systems (Agentic AI) from the eBook ‘Mastering AI Ethics and Safety’

• 2025 Stack Overflow Developer Survey: A TL;DR for Leaders

• Five quick updates about that Apple reasoning paper that people can’t stop talking about

• GPT-4’s successes, and GPT-4’s failures

• Unlock the potential of gen AI while mitigating risks with actionable insights from the special edition DORA report.

• GenAI coding: most teams aren’t ready

• Generative AI’s crippling and widespread failure to induce robust models of the world

• AI Copilot Code Quality: 2025 Look Back at 12 Months of Data

• AI Copilot Code Quality

• Goldman Sachs is piloting its first autonomous coder in major AI milestone for Wall Street

• Goldman Sachs: AI Is Overhyped, Wildly Expensive, and Unreliable

• Horse rides astronaut

• Humans versus Machines: The Hallucination Edition

• I Tested How Well AI Tools Work for Journalism

• Integrating AI agents: Navigating challenges, ensuring security, and driving adoption

• LLMs + Coding Agents = Security Nightmare

• LLMs: Dishonest, unpredictable and potentially dangerous.

• Large Language Models, Small Labor Market Effects

• UK government trial of M365 Copilot finds no clear productivity boost

• McDonald's is ending its drive-thru AI test

• Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity

• Microsoft Lost $20 for Every $10 Copilot AI Subscription

• Musk’s AI firm forced to delete posts praising Hitler from Grok chatbot

• Not to be outdone by ChatGPT, Microsoft Copilot humiliates itself in Atari 2600 chess showdown — another AI humbled by 1970s tech despite trash talk

• On Illusion of Thinking - Do LLMs Reason?

• OpenAI’s Waterloo? [with corrections]

• People Are Asking ChatGPT for Relationship Advice and It’s Ending in Disaster

• Pieces AI Productivity Summit

• Prompt-inject Copilot Studio AI via email, grab a company’s whole Salesforce

• Questioning AI Resource List

• The Trouble with AI

• Slopocalypse Now

• Stanford Research Finds That “Therapist” Chatbots Are Encouraging Users’ Schizophrenic Delusions and Suicidal Thoughts

• Tech sector layoffs explained: What you need to know

• The AI Industry Is Still Light-Years From Making a Profit, Experts Warn

• The Age of AI has begun

• The Death of the Student Essay—and the Future of Cognition

• The Gentle Singularity

• This is the gentle singularity?

• The Stochastic Code Monkey Theorem

• The joy of blackouts; AI ruins college; The Consciousness Wars continue; Peter Singer’s chatbot betrays him, & more

• This article is more powerful than the human brain.

• Time saved by AI offset by new work created, study suggests

• Trusting your own judgement on ‘AI’ is a huge risk

• Tulsi Gabbard Admits She Asked AI Which JFK Files Secrets to Reveal

• Trump’s acting cyber chief uploaded sensitive files into a public version of ChatGPT

• Ed tech is profitable. It is also mostly useless

• Vibe coding won't replace software engineers

• When ChatGPT summarises, it actually does nothing of the kind

• Why DO large language models hallucinate?

• Why high-quality data is essential for agentic AI

• Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task

• Zero-Click AI Vulnerability Exposes Microsoft 365 Copilot Data Without User Interaction

• Impact of Generative AI in Software Development

• LiveCodeBench Pro: How Do Olympiad Medalists Judge LLMs in Competitive Programming?

• Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity

• The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity

• ”Those claiming we’re mere months away from AI agents replacing most programmers” should think again

• What is digital literacy and why is it important?