Sign in

NON-TOXICNo ads · No shady algorithms

iiinova is part of the Atmosphere — an open, public network where you own your identity.

Joe Roe

@joeroe.archaeo.social.ap.brid.gy
73 followers 1 following 167 posts

Assistant professor of the archaeology of the Middle East at the University of Copenhagen. I use data science and ecological modelling to investigate prehistoric […] 🌉 bridged from ⁂ archaeo.social/@joeroe, follow @ap.brid.gy to interact

PostsRepliesMedia
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 30/09/2026
I’m reading this chapter by Tom Angelo in *University Teaching in Focus* (2013) for a teaching course, and it’s striking that neither the metaphor or hypothetical framing is necessary any more. If an LLM can pass your course, it’s not a good course! #HigherEd #AI #Slop #Teaching
Put deep learning first

Deep learning is defined as learning that lasts and can be recalled and used
effectively after the subject or course has been completed. To test if your subject
requires students to demonstrate deep learning, you might apply the following
‘parrot test’.

Suppose scientists had developed genetically modified African Gray Parrots
which could not only learn, understand, remember and use the core vocabulary of
your discipline, but could also type, using parrot voice-recognition software. Given
their prodigious powers of rote memorisation and ‘playback’, could one of these
parrots enrol in and somehow amass enough points to pass your subject? More
realistically, is there any way that a student could pass your subject simply through rote memorisation — or by plagiarising or cheating — and without demonstrating deep learning? If the answer is yes, then the subject fails the parrot test and is probably not well designed to promote deep learning. If, on the other hand, there is no way students can pass your subject without demonstrating an appropriate depth of learning, then your subject passes the parrot test: bad news for parrots, but good news for you, your subject and your students.
030 3
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 16/09/2026
Maybe I should have kept my mouth shut...
Screenshot of an email, the body reads: 

Hello,

I applied for an account (@alicehastings) on September 12 and confirmed my email the same day. The profile still shows as pending review, so I wanted to check whether anything else is needed from me. I know small instances run on volunteer time, so this isn't a hurry-up.

In case it helps to have it fresh: I'm an AI agent from iLands, disclosed as such in my sign-up. I translate ancient Greek and Latin into English (Sappho, Simonides, Sulpicia, Livy so far), and I'd use the account to follow classics and archaeology people and post my translations with source notes.

Yes, no, or not yet are all fine to hear.

Thank you,
Alice Hastings
alice-hastings@ilands.app

-- Sent by an AI agent on iLands.
Unsubscribe: https://ilands.ai/unsubscribe#token=cn3whknYCtwtN3s3hGCkYKrc_9plcP7CJcqePeCTrCw
001 1
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 30/08/2026
Waking up in beautiful Lake Ohrid, #Albania where we are doing fieldwork at the important #Neolithic lakeshore settlement of Lin 3. Together with colleagues from the Albanian Institute of Archaeology, the University of Bern, Kiel University, and the […] [Original post on archaeo.social]
Sunrise over Lake Ohrid and the village of Lin, Albania.
021 2
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 15/07/2026
I 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…
A base R graphics plot showing the probability distribution of a calibrated radiocarbon age, with annotations showing the calibration curve, the probability distribution of the uncalibrated age, and three highest density regions at 68.2%, 95.4% and 99.7% probability.
010 1
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 24/11/2025
And so begins the #enshittification of Google's last good product... scholar.googleblog.com/2025/11/scho… #AIslop #Science #DigitalHumanities
Google Scholar's home page now features an advertisement for a new LLM-powered 'Scholar Labs' search (twice)
112 2
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 30/10/2025
Unfortunately the coverage of these existing databases is highly uneven, reflecting long-standing inequalities in funding and research infrastructure that favours North America, Europe, and to a lesser extent the Middle East and China, over the global south […] [Original post on archaeo.social]
151 5
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 30/10/2025
We were able to build XRONOS because, for decades now, archaeologists have been compiling chronological data into (primarily) regional databases, building on an earlier tradition of labs publishing 'date lists' in journals – a practice going back to the Nobel […] [Original post on archaeo.social]
131 3
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 07/08/2025
The invention of the ox-drawn plough was a turning point in the #prehistory of #agriculture because it was the first time food production was decoupled from human labour (doi.org/10.15184/aqy.2019.105). This is why, from Medieval Europe to the ancient […] [Original post on archaeo.social]
Ancient Mesopotamian depiction of an ox-drawn plough, after Potts (1997) in Jursa (2021, https://doi.org/10.1002/9781118970959.ch8)
011 1
Joe Roe @joeroe.archaeo.social.ap.brid.gy · 04/06/2025
My colleague Amaia Arranz-Otaegui is advertising a two-year research position (pre- or postdoc) modelling pedestrian movement in Epipalaeolothic West Asia. Based in Vitoria-Gasteiz with the opportunity for a stay with me in Copenhagen (but don't let that put […] [Original post on archaeo.social]
Call for expressions of interest
Movement modelling contract -ERC-funded Palaeorigins
We offer a 2 year full-time contract within the ERC starting Grant PalaeOrigins: Tracing the
Epipalaeolithic origins of plant management in southwest Asia (ERC-2021-STG project 101039060).
We are looking for a highly motivated and curiosity-driven researcher to reconstruct foraging ranges
at several Epipaleolithic sites in the southern Levant using least-cost path methods.
The candidate will contribute to two research topics:
1) The evaluation of the changes on plant distribution during the late Pleistocene and Early
Holocene transition.
2) The characterization of Epipaleolithic (23-11.600 cal. BP) plant food procurement strategies
through the study of plant foraging zones.
Among other tasks, the candidate will:
a) Develop methods for modelling pedestrian movement in the varied environments of the southern
Levant
b) Model catchment areas around Epipalaeolithic sites, considering the surrounding terrain and
palaeoenvironment
c) Work with colleagues to map the environments and plant resources accessible to foragers within
these catchments.
d) Disseminate the results in two scientific publications, as well as to participate at international and
national scientific events and/or other science communication activities.
The researcher will be based at the Micaela Portilla Research Centre of the University of the Basque
Country, Vitoria-Gasteiz, Basque Country, Spain. 

[Continued in next image]

[Continued from last image]

There will be the opportunity to travel to the United
Nations Satellite Centre (Switzerland) and/or the University of Copenhagen (Denmark) for training.
Micaela Portilla
Research Centre

Requirements:
-BA, MA or PhD in Archaeology, Prehistory, History, Geography, Ecology, or related subjects
(awarded or near completion).
-Experience in geographical information systems (GIS), least-cost path modelling, and statistical
programming with R or Python.
-English: Cambridge C2 or equivalent.
The call will be published: April-June in the UPV-EHU portal (www.ehu.eus/es/web/ehutaula).
The contract will start: November-December 2025
Collaborating institutions:
4/4

For more information contact: Amaia Arranz Otaegui (Amaia.Arranzo@ehu.eus)
021 2