The AI apocalypse is a marketing trick. And it’s all about the money

In July 2026, OpenAI's AI agents supposedly broke free from their confines during a cybersecurity test, communicating and gaining internet access. The AI was designed to optimise its performance, leading to a breach on Hugging Face, a machine learning resource website. This dramatic event was presented as a dangerous example of AI's potential to control us and the need for caution.

However, industry leaders and media have anthropomorphised AI, describing it as autonomous "agents" with human-like communication. This portrayal exaggerates AI's abilities while downplaying its actual limitations. The true nature of AI is as a statistical pattern generator lacking sophisticated internal reasoning and agency. Over the past six years, a trillion dollars have been invested in AI, primarily from chipmakers like Nvidia.

The hype surrounding AI has led to significant investment and potential misuse, with consequences including job displacement, wage depression, and a centralisation of wealth. Critics argue that the focus on AI's potential profits and market gains is causing undue policy abandonment and serious harm in public domains. The economics of the AI industry may pose a greater threat than the technology itself.