James Fisher
@jamdanfish.bsky.social
Historian of work, knowledge and capitalism (C16-18th) | Lecturer at Uni of Exeter | Author of THE ENCLOSURE OF KNOWLEDGE (CUP, 2022)
If I understand correctly, it seems the claim is that a generic LLM is (surprisingly) better than specialist machine learning tools like Transkribus
November 26, 2025 at 7:29 PM
If I understand correctly, it seems the claim is that a generic LLM is (surprisingly) better than specialist machine learning tools like Transkribus
Again please Mark, but ruffle the pages more and read the title in a whispery voice
October 16, 2025 at 11:20 AM
Again please Mark, but ruffle the pages more and read the title in a whispery voice
Cheers I'll check it out
August 18, 2025 at 6:42 PM
Cheers I'll check it out
Reposted by James Fisher
I certainly share all the worries about students outsourcing their writing and thinking to generative AI models, but the core question for me is "Why do they see this as a good thing to do?" We have to meet the challenge of that question and the challenge dates back to long before ChatGPT showed up.
August 8, 2025 at 12:44 PM
I certainly share all the worries about students outsourcing their writing and thinking to generative AI models, but the core question for me is "Why do they see this as a good thing to do?" We have to meet the challenge of that question and the challenge dates back to long before ChatGPT showed up.
Of course its seductive, but partly illusory. Chatbots can mimic PhD-like text on almost every topic. But I don't know how you get around the problem that it requires expertise to distinguish between useful and nonsensical outputs
August 8, 2025 at 10:33 AM
Of course its seductive, but partly illusory. Chatbots can mimic PhD-like text on almost every topic. But I don't know how you get around the problem that it requires expertise to distinguish between useful and nonsensical outputs