The Turing Test
participants 56% of the time, ELIZA 23% and GPT-4o 21%. The most interesting outcome from this test is the difference between GPT-4o and GPT-4.5. Although being released less than a year apart (May 2024 and February 2025), the difference in convincing participants is immense, with GPT-4o even being beaten by ELIZA. Assuming this pace of improvement continues, or even if it majorly slows down, it is safe to assume that in the next few years, AI will be indistinguishable from humans in basic conversation. Modern LLMs also don’t rely on the tricks which older chatbots did, which reflects on their greater ability to learn from training material, although it is widely believed that they are not truly capable of thought yet. The recent appearance of reasoning models in mainstream AI products also points to LLMs being able to very closely mimic thought in the near future. In summary, while the Turing Test has its clear flaws, it raises an important question of computers’ abilities to mimic the most human behaviour, thought. As the rate of AI development rapidly increases, the gap between human and computer intelligence will only narrow, and the Turing Test’s significance will become trivial, leading us to ask more important questions about the future of our co-existence with computers. Bibliography Assad, Z. (2025) ‘ChatGPT just passed the Turing test. But that doesn’t mean AI is now as smart as humans’, The Conversation , April 2025 Boden, M. (2018) Artificial Intelligence: A Very Short Introduction. Oxford Brainerd, W. ‘Eliza chatbot in Python’, https://github.com/wadetb/eliza/tree/master. Consulted: 17/08/25 Bringsjord, S., Bello, P. & Ferrucci, D. (2001) ‘Creativity, the Turing Test, and the (Better) Lovelace Test’, Minds and Machines 11, 3–27. https://doi.org/10.1023/A:1011206622741 Huang, A. (2006) ‘The Loebner Prize’, https://www.ocf.berkeley.edu/~arihuang/academic/research/loebner.html. Consulted: 16/08/25 Marcus, G. (2017) ‘The Search for a New Test of Artificial Intelligence’, Scientific American . https://www.scientificamerican.com/article/the-search-for-a-new-test-of-artificial-intelligence/ Merriam-Webster. (2025) ‘Think’, Merriam-Webbster.com Dictionary . https://www.merriamwebster.com/dictionary/think Riedl, M. (2014) ‘The Lovelace 2.0 Test of Artificial Creativity and Intelligence’, https://arxiv.org/abs/1410.6142 Ronkowitz, K. ‘ELIZA: a very basic Rogerian psychotherapist chatbot’, https://web.njit.edu/~ronkowit/eliza.html. Consulted: 17/08/25 Shieber, S. (2011) ‘Inverting the Turing Test’, American Scientist , https://doi.org/10.1511/2011.93.502 Turing, A. (1950) ‘Computing Machinery and Intelligence’, Mind 59. University of Reading. (2014) ‘Turing Test Success Milestone in Computing History’, University of Reading News Archive, 8 June 2014. https://archive.reading.ac.uk/news-events/2014/June/pr583836.html
84
Made with FlippingBook - PDF hosting