7 ppl attended
Discussion
This meeting was almost entirely dominated by a spontaneous discussion about applications, effects, and ethics of artificial intelligence in archaeology. Here are some of the tangible resources that where mentioned and referenced:
- A deep-learning-based “search engine” for Archaeologists
- Stephen Wolfram’s explanation of the inner workings of LLMs
- Jean-Claude Gardin, one of the founders of archaeological computing, and his struggle with archaeological expert systems
- A summary of his theoretical work: https://doi.org/10.1007/s10816-015-9241-3
- A postdoctoral fellowship to support the understanding of AI in the social sciences and humanities
The topics raised in the discussion included the misleading semblance of objectivity projected by LLMs (linked to long-standing divides in archaeological theory about positivism), how AI applications may be a strong and convincing argument for Open Data in archaeology, and different levels of transparency in Open Source AI models.
Multiple SIG members were inspired by this discussion:
- See the CAA/SSLA chat at https://matrix.to/#/#caa-ssla:archaeo.social for an AI-related project outline proposed by Zack Batist
- We decided to invite knowledgeable colleagues to talk about AI (and scripting languages?) in our presentation format. Martin Hinz has already sent out multiple emails to invite potential speakers.
TFQA project
In this meeting we spent little time on the SIG’s ongoing TFQA project. But James Allison, Lisa Steinmann, and Joe Roe recently made good progress on that. See the CAA/SSLA chat and various open and already closed pull requests in the https://github.com/jallison7/LDen repository.
