David Stott
banner
davstott.bsky.social
David Stott
@davstott.bsky.social

Paleotopographer. Asking how people changed landscapes and landscapes changed people.

GIS, remote sensing, archaeology.

Tinkerer. Aspiring bodger.

Glaswegian in Jutland.

Archaeologist at Moesgaard Museum. Personal account. All views my own etc. .. more

History 18%
Political science 15%
Pinned
There's enough of us now to do one of these!

go.bsky.app/M9FwG2K

🏺

At least they've got lakes. May Iowans for maximal irony?

Does that include slightly peevish junior academics?

Who gets to be the sea peoples?
The Bronze Age vs The AI Age
1/ Can archaeology predict the future of Big Tech?

Economists usually look back 50-150 years; archaeologists look back further.
When you look at the Bronze Age Collapse (c. 1200 BCE), are there any parallels to our current AI trajectory?
#AI #TechHistory #Archaeology
A wonderful Reddit post asking for songs in minority languages/small languages. Complete with spreadsheet and playlists.

www.reddit.com/r/AskEurope/...
From the AskEurope community on Reddit: Are there any songs in minority languages from your country that you like listening to?
Explore this post and more from the AskEurope community
www.reddit.com

IMO the whole stepless gear thing is a gimmick anyway- Found myself using the ends of the range and a couple of spots in the middle, so going from that to a 5 speed isn't much of an adjustment...

Hopefully the Nexus internal gear hub proves more robust than the Enviolo ones. Even if it doesn't it's much simpler to work on, has more user serviceable parts, and if it comes to it is much cheaper to replace.

Wheel is now on the bike and working well after chilly test rides and much faffing about getting the chain line right.

Now just need to bleed the brakes and figure out what to do for a chain guard.

Happy with the choice of 38T chainring to 30T sprocket- seems about right for the hub

#cargobike
The Bronze Age vs The AI Age
1/ Can archaeology predict the future of Big Tech?

Economists usually look back 50-150 years; archaeologists look back further.
When you look at the Bronze Age Collapse (c. 1200 BCE), are there any parallels to our current AI trajectory?
#AI #TechHistory #Archaeology
Today is Septidi the 27th of Pluviôse in the year 234.
Pluviôse is the month of rain.
Today we celebrate hazel.#JacobinDay

More information on hazel
Yes! What if we treated extremely normal human variation as something other than a "disorder"???

www.positive.news/society/yout...
Rethinking ADHD as ‘hypercuriosity’
What if ADHD isn’t a deficit of attention, but an intensified curiosity? A new study explores how reframing ADHD could transform education
www.positive.news

Reposted by David Stott

On for #ValentinesDay Double-heart of Stacked Stones (traditional Chinese: 七美雙心石滬; simplified Chinese: 七美双心石沪; pinyin: Qīměi Shuāng Xīn Shí Hù) or the Twin-Heart Fish Trap is a stone fishing weir located on the north side of #Cimei Township, #Penghu #County, #Taiwan.

I don't think many people understand what anarchisim is or are scared off by the appellation.

Reposted by David Stott

Happy Galena-tines!

Rhodochrosite with Galena, Sphalerite and Chalcopyrite.

When hit, Galena will break along three directions of weakness that meet at right angles, creating smaller cubes. Galena’s hardness is similar to that of a fingernail.

Incredible book!

Reposted by David Stott

One of the most powerful books ever written, highly recommend, and my subject for next week’s episode of Past Lives

The RTK is integrated in the drone but the sensor has a good IMU so it's complicated. Maybe I could just dangle the entire drone? I've still got the big lifting kite...

Hooray! I used to have a nice little KAP side gig until drones ruined everything. I wonder if i could dangle our lidar sensor?

Reposted by David Stott

Have you ever seen a Roman plumbing device that is both genius and gorgeous? Just look at this bronze drain strainer, found in the Roman baths at Weißenburg, Germany, dating2nd/3rd century AD!

📷 me

🏺 #archaeology #findsfriday

here's another nice false positive showing how the handwritten palimpsest is a challenge

Here a false positive beside a correctly identified barrow partially obscured by text

Here's some in a different style- one that I like to call 'green hairballs'

Reposted by David Stott

#FindsFriday
The Cairns Character!
Discovered in a pit at the #IronAge #Broch site of The Cairns on the island of South Ronaldsay, #Orkney. He or She was carved from a beach pebble around 2,000 years ago.

📸 mine

Those wind turbines look awfully close together
Palestine Action have WON their judicial review in the High Court!
A massive victory for free speech, freedom of association and the right to protest.
A massive defeat for the draconian attempts by Keir Starmer's government to curtail our fundamental rights.
‘I used to be a mountain’ by Maarten Inghels onboards.be/nl/product/m...
📷 Maarten Inghels/Onboards Biennale
#urbangeology

Reposted by David Stott

Beautiful line drawing of a Scottish gold torc (Netherurd, @tessmachling.bsky.social ??) by Keith Henderson in Piggott’s 1958 Scotland Before History (thanks to @gjmichaelson.bsky.social for reminding me of this book!)
#FindsFriday

... I'd say it's worth giving it a go without re-training first and see what it finds in your dataset. Then it would be possible augment the model with new training data if necessary.

The code is adaptable if temperamental and poorly documented (something I'll sort before we publish)

... There's also a lot going on with the text in the fields (taxable yields & parcel numbers) - this gets erased, redrawn, over written so we end up with a bit of mess, and seems to be the origin of a lot of the false positives ...

We've got quite a few styles in the training data- there's ~11k maps and 20 odd different surveyors listed in the metadata, but this seems to translate to 7 or 8 different 'hands'. Success is variable, as some of them are rather smudgy...

Mask-R CNN used to segment barrows in historical mapping. These hand-drawn maps are challenging as there's a lot going on.

There's also a lot of work involved in evaluating these results- how do we winnow the informational wheat from the chaff to get the delicious kernels of knowledge?

🏺 #GIS