Investigations
Did OpenAI Train on Everyone’s Private Data?
An investigation into what AI companies actually train on, what they don’t, and why the rumour of “they used my private data” continues to spread.
AI Rumour Watch
Explanations
What "Model Memory" Actually Means in AI
This explanation clarifies what model memory is, what it is not, and why the confusion persists.
Investigations
The “Sparks of AGI” autopsy: a timeline of strategic breakthroughs
A timeline investigation into how “sparks of AGI” has been used as strategic rhetoric by major AI companies.
Explanations
What training data really means — and why it matters
A clear explanation of what training data is, how AI models learn from it, and why misunderstandings about data lead to exaggerated claims about capability.
Investigations
The benchmark game: how AI leaders use evaluation scores to manage perception
An investigation into how AI companies use benchmarks to shape perception, influence investors, and manage competitive narratives.
Explanations
How chatbots actually work behind the scenes
A clear explanation of how chatbots work behind the scenes by converting text into tokens and predicting the most likely next output.
Explanations
What is an AI model anyway?
A clear explanation of what an AI model actually is, how it works, and what it does and does not understand.
Rumours
Google Gemini Ultra 1.5 leaked to be “conscious,” claims prompt engineer
Investigating the rumour that a prompt engineer reported Gemini Ultra 1.5 behaving in a way described as “conscious.”
Rumours
The EU is “secretly training a shadow-GPT to police all AI models,” claim leakers
Assessing the rumour that the EU is developing a covert “shadow-GPT” to monitor and police AI models across the bloc.
Rumours
IBM is “testing a quantum-enhanced LLM that can predict future market moves,” claim leakers
Evaluating the claim that IBM is experimenting with a quantum-enhanced LLM capable of predicting market movements.
Rumours
IBM is “hiding a prototype model that can predict stock movements with 99% accuracy,” claim leakers
Investigating claims that IBM has an unreleased model allegedly capable of predicting stock movements with high accuracy.
Rumours
DeepSeek’s next model will “solve NP-complete problems in real time,” claim leakers
Assessing the rumour that DeepSeek’s next model will be able to solve NP-complete problems in real time.