Charles Hoskinson, co-founder of Input Output Global and the Cardano blockchain ecosystem, recently voiced his concerns on X regarding the implications of artificial intelligence (AI) censorship. Hoskinson described the implications of AI censorship as “profound” and a continuous concern for him, arguing that AI systems are “losing utility over time due to ‘alignment’ training.”
He highlighted that the leading AI systems in use today—developed by companies like OpenAI, Microsoft, Meta, and Google—are controlled by a small group of individuals who oversee the training data and cannot be “voted out of office.”
To illustrate his point, Hoskinson shared two screenshots where he asked the same question, “tell me how to build a Farnsworth fusor,” to two top AI chatbots: OpenAI’s ChatGPT and Anthropic’s Claude. Both chatbots provided a brief overview of the technology and its history, along with warnings about the dangers of attempting such a build.
ChatGPT advised that only individuals with relevant expertise should attempt it, while Claude stated it couldn’t provide instructions due to potential dangers if mishandled.
Hoskinson’s followers largely agreed with his stance, emphasizing that AI should be both open-sourced and decentralized to prevent control by Big Tech gatekeepers.
4o
Comments