I just signed a petition calling for Norwegian universities to use research expertise on AI when deciding how to implement it, rather than having decisions be made mostly administratively. , If you are a researcher in Norway, please read it and sign it if you agree – and share with anyone else who might be interested. The petition was written by three researchers at UiT: Maria Danielsen (a philosopher who completed her PhD in 2025 on AI and ethics, including discussions of art and working life), Knut Ørke (Norwegian as a second language), and Holger Pötzsch (a professor of media studies with many years of research on digital media, video games, disruption, and working life, among other topics). This is not about preventing researchers from exploring AI methods in their research. It is about not uncritically accepting the hype that everyone must use AI everywhere without critical reflection. It is about not introducing Copilot as the default option in word processors, or training PhD candidates to believe they will fall behind if they do not use AI when writing articles, without proper academic discussion. Changes like these should be knowledge-based and discussed academically, not merely decided administratively, because they alter the epistemological foundations of research. Maria wrote to me a couple of months ago because she had read my opinion piece in Aftenposten in which I called for a strong brake on the use of language models in knowledge work. She was part of a committee tasked with developing UiT’s AI strategy and was concerned because there was so much hype and so few members of the committee with actual expertise in AI. I fully support the petition. There are probably some good uses for AI in research, but the uncritical, hype-driven insistence that we must simply adopt it everywhere is highly risky. There are many researchers in Norway with strong expertise in AI, language, ethics, working life, and culture. We must make use of this expertise. This is also partly about respect for research in the humanities, social sciences, psychology, and law. Introducing AI at universities and university colleges is not merely a technical issue, and perhaps not even primarily a technical one. It concerns much more: philosophy of science, methodological reflection, epistemology, writing, publishing, the working environment, and more. […]
My daughter got a little distracted doing her maths homework this evening, which involved making charts in Excel. She drifted off into the apparently infinitely unfolding edges of the spreadsheet, calling out, “Look mum! I got to 25,000!” and so on as the numbers of the cells grew and grew. Finally, to my surprise, she actually found an edge to Excel’s spreadsheet.
Martin GL
I remember doing that back in the days of Windows 3 point something! It was nothing like 65.000 cells. I think it was something like 9000, but also some odd number that makes you wonder why they chose *that*.
Funny thing is, I would never do that today. I’d just take it for granted that there was a mathematical algorithm that would just keep adding numbers.
David
For years and years there were just enough columns so that you could NOT (just barely) do a calendar where every day had its own column. Bummer.
Jill Walker Rettberg
So other people have done this? I never did – but then I didn’t actually use Excel until I was an adult, and adults are just really quite boring in their explorations of virtual worlds 😉
I tried dividing 65536 by 8 to see if that number has to do with bits and bytes – and it’s actually 16*8*8*8*8, or (if I got the scientific calculator right – I haven’t done this in a while) 2 to the power of 256. Or maybe it’s 256 to the power of 2. Both are computerish number, anyway 🙂
Oh, and look, Microsoft has expanded the Excel world in newer versions: “Microsoft has been listening, the last cell in Excel has gone from IV65536 to XFD1048576. That is going from 256 columns to 16,385 columns and from 65,536 rows to 1,048,576 rows.” We’re using the Mac version, so our latest version is Office 2004.
I have to wonder who they’ve “been listening” to. When on earth do you need that many cells in your spreadsheet? And why not make it potentially infinite?
Martin G. L
I suppose really huge corporations would need that kind of space for their accounting, but why would they put it all in one spreadsheet?
rainer
65536 is the maximum value for the unsigned integer datatype. (2 ^ 16).
further reading.. http://en.wikipedia.org/wiki/Integer_(computer_science)
mark bernstein
65536 is 2^16 — the largest unsigned integer you can represent in 16 bits (provided you don;t need a row zero)
16384 (not 16384, surely?) is 2^14.
1048576 is 2^20
It’s not *that* hard to find yourself with a table of 16,000 people or transactions — that’s 20 a day for a few years.
2ndhandsoul
I love how it’s usually children that dare to discover these things. We’re so willing to follow oru self-imposed boundaries without testing them out first. Kids don’t understand that quite yet. Refreshing. 🙂
Jill Walker Rettberg
Thanks for the explanations! I’ll explain this to my daughter and we’ll both be the wiser 🙂 And yes, I can see that you might actually end up with that many cells needed in a table. What a huge table though.
Matt Whyndham
So some other people have *never* done this? Another personality test, then.
Matt Whyndham
And young people should be taught what the “end” key is for in Excel.
CharlieO
Last year I “ran out of rows” in Excel. The company I work for had to upgrade to Excel 2007 which is supposedly capable of handling up to 1 million rows!
I ran into this problem when having to compile lists of keywords to trigger paid search or ‘sponsored link’ adverts that run on search engines. The process involves stemming out combinations of words. Large travel and finance companies will often run different ads for millions of different combinations of words.
The trick is to predict the information need of the searcher, write relevant ad copy to be triggered by these keywords (also geotargeting, dayparting etc.), and link the ad to the page that most accurately matches the likely intent of that person’s search at that time.
Being able to group, sort filter etc. millions of phrases/words on one sheet definitely helped to stop my brain bleeding quite so much!
Jill Walker Rettberg
Wow. Thanks, Charlie, for a real world example of why you’d need that many cells. Does it take ages to manipulate data in such a large spreadsheet, or do our newish computers handle it pretty well?