artificial intelligence

AI summaries produce ‘confident responses in a detached, authoritative tone, without clear context, and with little way to fact-check any claims. There is no author of the information, typically no citations [and] no transparency on what information the AI models have been trained upon’ (Jackson et al. 2025: 657). Thus, no guarantees exist that the information is correct or unbiased. It is therefore necessary to contrast machine-driven AI (with its opacity, centralised decision-making, monolithic knowledge, and tendency to hallucinate) to human-driven collective intelligence (with its collaborative decision-making, dialogic process, and error correction through deliberation, transparency and accountability) as performed in Wikipedia (Tarkowski et al. 2026).