Eric Kansa @ekansa.scholar.social.ap.brid.gy · 16/10/2026I'm at a cafe in San Diego, where I was an undergrad more than 30 years ago. They are playing exactly the same music I heard back then. I, however, have lost some hair along with auditory and visual acuity in the intervening decades 170 7
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 07/10/2026For context, Google has committed a cool $200 billion in data center investments for 2026 alone. I'm sure Google will have zero temptation to use it's AI systems to influence permitting regulation (sarcasm) […]scholar.socialOriginal post on scholar.social 020 2
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 07/10/2026So, the ACHP wants to use Google's AI algorithms to regulate Google's AI data center plans? Doesn't this seem like a clear conflict of interest? #archaeology #culturalheritage #section106 #policy #AI #governance 111 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 07/10/2026The ACHP is contracting with Google for the new "AI" system. Google is a major investor and developer of AI data centers. These data centers projects can be enormous and often require hugely disruptive use of land for the data centers themselves and for the power and cooling infrastructure they […]scholar.socialOriginal post on scholar.social 110 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 07/10/2026Chris Nicholson and I recently wrote about the Advisory Council on Historic Preservation (ACHP)'s plans to use "AI" to streamline permitting workflows. This comes at a time when US historical preservation protections (esp. Section 106) face multiple legislative and regulatory threats. See […]scholar.socialOriginal post on scholar.social 124 4
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 07/10/2026The 11 AM test, on the first Wednesday of each month, of Berkeley's civil defense / air-raid system never ceases to freak me out. 010 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 06/10/2026@electricarchaeo @archaeoscape Right now I'm re-reading "Guards! Guards!" and enjoying the humor. I'm interested in reading Mort next, since I haven't read it yet. 110 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 06/10/2026Been rereading Terry Pratchett (en.wikipedia.org/wiki/Terry_Pratche…) because of various reminders and references made by @electricarchaeo Totally worth anyone's time.en.wikipedia.orgTerry Pratchett - Wikipedia 250 5
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 06/10/2026#AI agents / #bots that do "read only" scraping are already damaging enough, but now #OpenAI seems to be experimenting with (how can it not be criminal) hacking at scale. Not just HuggingFace and various government websites around the world, but also the Wikipedia […]scholar.socialOriginal post on scholar.social 000 0
Reposted by Eric KansaZack Batist @zackbatist.archaeo.social.ap.brid.gy · 05/10/2026Please get in touch with me or respond here if you would like to share any ideas or feedback for improving the next iteration of the Digital Archaeology Maintainathon! - Do you want certain kinds of resources or guides added to the website? - Any ideas for community engagement? - Constructive […]archaeo.socialOriginal post on archaeo.social 005 5
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 02/10/2026I will soon be canned in an aluminum tube, packed in with my fellow primates for several hours, and hurtled through the air by a profit maximizing algorithm. 030 3
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 29/09/2026@lavaeolus @InternetArchaeology Yep, I've been complaining about AI and AI scraping all summer. 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 25/09/2026I'm trying to improve the usefulness of a small, locally hosted, GPU-free language model for Open Context search. All very interesting. At the same time, Open Context is getting renewed unwanted attention of Javascript executing bots from random IP addresses. I think I need more network […]scholar.socialOriginal post on scholar.social 013 3
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 25/09/2026@electricarchaeo If you've been alive in the past 2 decades, such feelings are a given. 120 2
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 22/09/2026@electricarchaeo Wow! Lots of apples!, probably a huge effort going into managing them too. No wonder you're running out of steam. 010 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 22/09/2026@electricarchaeo Time for a change of pace! When do you harvest those apples? 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 22/09/2026Chis Nicholson (Executive Director of Digital Antiquity / tDAR) and I wrote an article about the Advisory Council on #AI for #culturalHeritage conservation policy. See: mydigitalpublication.com/publicatio… Instead of inscrutable black-box systems with […]scholar.socialOriginal post on scholar.social 021 2
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 21/09/2026Ugh. It's getting to that time of year with Trader Joe's is all in on seasonal pumpkin spice everything. I mean it's fine. Just some moderation please. 010 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 15/09/2026#Fuck_AI mailchi.mp/osf/osf_announcement-615… #openScience #opendata #openaccessmailchi.mpOSF Emergency Maintenance — Sept. 15, 2026 020 2
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 15/09/2026@joeroe That's a good question! Yes, I think it helps a great deal, but with limits. Anubis works by imposing a (computational, energy) cost for each session of access to a website. If you have lots of IPs, you have lots of Anubis challenges. But, if attackers use something like a residential […]scholar.socialOriginal post on scholar.social 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 12/09/2026I'm sounding like a broken record. But my #bot headaches with Open Context aren't just a simple function of our limited #archaeology resources. #AIbots are currently overwhelming #Zenodo, a major repository (with CERN infrastructure!). This problem cannot be ignored and needs a public policy […]scholar.socialOriginal post on scholar.social 025 5
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 09/09/2026Just had to block IP addresses from Brunei, Morocco, and Brazil, in addition to Singapore, because of bots getting past the Anubis checks. Again, the bots burned through Open Context's map API credits. To stop the bleeding, I've blocked some 300 million (?) people from accessing #opendata in […]scholar.socialOriginal post on scholar.social 206 6
Reposted by Eric KansaMisha @mishavelthuis.social.edu.nl.ap.brid.gy · 25/08/2026"The best time to have made Canada digitally sovereign would have been before an American president announced his intention to annex Canada and began explicitly deploying America's tech companies to attack his geopolitical adversaries. The second-best time is now." […]social.edu.nlOriginal post on social.edu.nl 0112 12
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 21/08/2026@archaeoscape.bsky.social @electricarchaeo CIDER powers the resistance! 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 21/08/2026@electricarchaeo Californian here: I'd protest with a nice tall drink of Canadian cider, FYI. Know where I can source one? 010 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 18/08/2026Does anyone have good examples of click-through terms-of-service agreements that attempt to regulate web scraping? Things are so bad, I need to have some sort of (even vaguely credible) legal teeth. #openaccess #opendata #scholarlycoms 051 5
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 17/08/2026If you're in Singapore and want to access data in Open Context, I'm so sorry. But some sort of botnet using IP addresses from Singapore was just too much for us. So I had to severely throttle (effectively block) the entire region. Send complaints to those financing this dumb "AI" bubble. 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 16/08/2026I'm guessing every damn network connected refrigerator and microwave oven in Singapore decided to simultaneously develop obsessive interests in sheep/goat bone data. Day 2 of an overwhelming bot swarm hitting Open Context, when I don't have the time to dedicate to investigate and deploy new […]scholar.socialOriginal post on scholar.social 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 14/08/2026So many javascript executing bots are hitting Open Context right now. We're seeing 15 times more traffic than "normal" (if there is such a thing any more) on Open Context. They burned through the use limits on our map APIs. Now maps our don't render properly […] [Original post on scholar.social] 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 13/08/2026David Hasselhoff, and he's riding a bus at night. Which makes him something of a "night rider", no? www.youtube.com/watch?v=XKdKU6p78TQ #coversong #thePassenger #dadJoke 010 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 13/08/2026@electricarchaeo Is Kevin Bacon connected? 000 0
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 11/08/2026Oh wow! Chris Webster and colleagues at the "Archaeology Podcast Network" chatted about my recent AI / Large Language Model paper (doi.org/10.11141/ia.71.15). That's a first for me! Here's the podcast link: www.archaeologypodcastnetwork.com/a… (I'm not sure they're on […]scholar.socialOriginal post on scholar.social 043 4
Reposted by Eric KansaJoe Roe @joeroe.archaeo.social.ap.brid.gy · 11/08/2026We are advertising two PhD positions in the Southwest Asia Palaeoarchaeology Lab at the University of Copenhagen. As part of a new project 'Networks Before Agriculture: Social Interaction in Epipalaeolithic and Neolithic Jordan', the successful candidates will work together to map prehistoric […]archaeo.socialOriginal post on archaeo.social 023 3
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 10/08/2026@hrheingold I think it more likely that LLM over-investment is inflating a hugely risky bubble. LLMs burn money and have ghastly environmental externalities. Those are real threats and damages. Otherwise, they are stochastic parrots. Their outputs can be dangerous, not because they come from […]scholar.socialOriginal post on scholar.social 000 0
Reposted by Eric KansaJoe Roe @joeroe.archaeo.social.ap.brid.gy · 15/07/2026I finally got round to adding classic single radiocarbon calibration plots to my #Rstats package {c14}. I think they turned out pretty nice! c14.joeroe.io/reference/plot.c14_ca… 010 1
Reposted by Eric KansaMartin Paul Eve @mpe.hcommons.social.ap.brid.gy · 06/08/2026"Researchers may cite an infrastructure as a necessary part of their work in a grant proposal, but funding flows only to the researcher and research project, not the repository." issues.org/open-data-research-%20in… 104 4
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 07/08/2026I guess that worked! Also, I forgot to note you can zoom in and pan around the visualization. It's kinda fun to watch the sherds dance and swirl around in a stately, weirdly turbulent way... #archaeology 010 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 07/08/2026I made a little update to include (hopefully working) social media metadata. Let's see: storage.googleapis.com/opencontext-… 100 1
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 06/08/2026Well shoot. Google Scholar isn't indexing my paper about #AI in #Archaeology -- more than one 1 month since publication (doi.org/10.11141/ia.71.15). If I was a paranoid, I'd wonder if Google was trying to inhibit the discoverability and circulation of a paper very critical of #LLMs. I […]scholar.socialOriginal post on scholar.social 014 4
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 04/08/2026If you click on a thumbnail, you'll go to the full database record for the object in Open context. 121 2
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 04/08/2026This visualizes diagnostic pottery sherds from Poggio Civitate, an Etruscan site south of Siena, Italy. More similar sherds group together. You can adjust the weights of different attributes to change groupings of the sherds. The "Embedding" slider is for similarity based on a small language […]scholar.socialOriginal post on scholar.social 155 5
Reposted by Eric KansaDan Gillmor @dangillmor.mastodon.social.ap.brid.gy · 02/08/2026Ed Zitron has been on this for a long time, but good to see that the NY Times finally is digging into Silicon Valley OG Larry Ellison's sleazeball ways -- and his prop-up-Oracle-stock-until-it-crashes bets on AI. www.nytimes.com/2026/07/31/magazine…nytimes.comLarry Ellison Bet It All on the A.I. Boom. Will He Be the Face of the A.I. Bubble?Inside the 81-year-old billionaire’s risky, debt-fueled scramble to transform his data empire into an A.I. juggernaut. 1219 19
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 23/07/2026Playing with some #datavisualization of #Etruscan frieze plaque fragments from Poggio Civitate. storage.googleapis.com/opencontext-… This is a force directed graph with edge (connection) strength between […]scholar.socialOriginal post on scholar.social 022 2
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 13/07/2026I fixed a stupid CSS bug and now Open Context is properly falling back to a "noto" font for use with ancient scripts. Here's a nice example of Etruscan, with the unicode getting rendered correctly. It reads (something like): 𐌀𐌙𐌉𐌓 (or less likely) 𐌀𐌙𐌐𐌓 It's interesting because it's written on […]scholar.socialOriginal post on scholar.social 012 2
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 10/07/2026We're celebrating 60 years of continual archaeological research at Poggio Civitate (Murlo) in Italy. The excavation and field school, led by Prof. Anthony Tuck (University of Mass. Amherst), has trained generations of students. It's a wonderful project encouraging learning, collaboration, and […]scholar.socialOriginal post on scholar.social 033 3
Eric Kansa @ekansa.scholar.social.ap.brid.gy · 01/07/2026What will this chart look like next year? Will the "AI" bubble pop and sanity return? How can we continue to sustain public, open data that inform the sciences, social sciences, and humanities? #LLM #botTraffic #Bubble 010 1