recorded a Windows 95 full disk defrag to soothe your timeline.

Is there anything more satisfying than sending out final edits on an article or a book chapter? I know, I lead a sad life.

Can you steal game gold pieces? The Court of Appeal says yes It's been a while since we had a proper digital property and virtual gold story here at Llama Towers, I have to admit that it's been mostly AI for the last few years. But a recent decision in the Court of Appeal (Criminal division) in…

My timeline.

Wait, doesn't everyone start their classes with "hey chat", and end with "please like and subscribe", or is it just me?

Oooh! A good old-fashioned digital property and gold ruling! It's 2006 all over again. www.rockpapershotgun.com/a-judge-has-...

Oh no, the three most fearful words in the English language: Bus replacement service.

God, I hate long footnotes.

Exclusive reveal of the Board of Peace headquarters.

I'm sensing a vibe shift in the copyright AI debate. People were promised that copyright would destroy generative AI, yet it's now 2026 and that is not going to happen. I sense something akin to the end of the P2P wars, some sort of recognition that the landscape has changed.

Why We Need Tech Lawyers to Shine Again Text from my editorial in IIC. n courtrooms across the globe, a quiet crisis is brewing. As the number of artificial intelligence copyright litigations increases, judges are being asked to decide the fate of technologies that often operate in dimensions the…

I've never felt proprietary about my writing, once it's out there it's no longer mine, which is perhaps one of the reasons why I’m comfortable with my work being used to train AI. It feels good knowing a small part of me lives in the training data.

The Greenland distraction had the desired effect: Nobody is talking about Epstein.

Insert a weary "I'm starting to think that DJT didn't deserve the inaugural FIFA Peace Prize" joke here.

I'm looking forward to the less interesting times.

Inadvertently coming across some timelines.

"Oh wow, your Spanish is so good! You almost sound Costa Rican!"

This new paper is very cool. But takeaway isn't "LLMs constantly leak books.' It’s "with the right (adversarial) procedure, you can extract memorized text." Prevalence in the wild is a separate question!! arxiv.org/abs/2601.02671

Why are Americans so loud? I've never understood why, last night there was a couple at the bar and you could follow their conversation over the music and the rest of the people. Is it a learned behaviour? Is it a defence mechanism to scare off predators?

Me typing "This is just a gentle reminder..."

Has anyone else noticed that the date of the invasion coincides with the capture of Manuel Antonio Noriega? January 3, 1990.

Happy New Year.

Happy Public Domain Day to all who celebrate. Remember to leave Steamboat Willie a plate of biscuits and a free beer, so he’ll release your favourite works into the commons.

Last night I dreamt that because of a weird legal loophole, I became both president of the United States and Costa Rica, and I was sworn in on a train in Spanish. Even stranger, I sounded like Manousos Oviedo from Pluribus. What should be my first act? Invade all the in-between countries?

We’re living through a truly 21st-century technological revolution, but we’re still relying on legal frameworks rooted in the 18th century and repeatedly stretched to fit 19th- and 20th-century media.

The only booming creative industry at the moment is copyright lawyers.

Finally joined Letterboxed! letterboxd.com/anduin13/

Visiting family and friends, I’ve been surprised by how much AI has escaped online discourse and permeated so many conversations this holiday season, and how overwhelmingly positive those conversations have been. I’ve also been surprised by how popular AI slop has become.

I often think about this masterpiece.

"I hate The Clash". That noise you heard was the collective gasp of millions of Gen-Xers.

Currently reading "Project Hail Mary", really enjoying it, good book for the beach. It was published in 2021, and the one aspect that aged badly is how stupid the onboard computer is. Even November '22 ChatGPT would have been smarter, it's amazing how fast things have moved.

CHIEF: "Conan, what is best in life?" CONAN: "To crush your enemies, see them driven before you, and to hear the lamentation of their women." The enemy:

I'm officially on vacation and yet I spent the day working.

Big news out of Germany. The appeal in Kneschke v LAION has been rejected, LAION wins again. grunecker.de/en/insights/...

Thinking of a new AI policy in essays. If the student used AI to write it, I will mark it with AI and give feedback with AI. Machines marking machines.

I thought about writing a post about the Disney - OpenAI deal here, but seeing the level of responses I might as well not.

Disney and OpenAI just made a deal that has surprised onlookers - and I'd wager shocked people within both companies. For @newscientist.com I tried to answer what happens next? www.newscientist.com/article/2508...

Interestingly, I'm starting to see more questions that seem to come directly from the disparity between what the text of a case says, and what chatbots say about a case. I just had to tell a student to read the case because the answer was there.

It's crazy that Europe should be castigated by a corrupt plutocracy with no balance of power, but here we are.

One of those days when I’m tempted to quit and open a bar on a Costa Rican beach. I love academia, but some days the sheer level of snobbery really drags me down.

We need to talk about the EU TDM exception and AI training Following the recent decision in GEMA v OpenAI, and last year's Kneschke v LAION, we now have two German courts grappling with the applicability of the text and data mining exceptions to AI training. Despite arriving at different outcomes,…

We need to talk about the EU TDM exception and AI training Following the recent decision in GEMA v OpenAI, and last year's Kneschke v LAION, we now have two German courts grappling with the applicability of the text and data mining exceptions to AI training. Despite arriving at different outcomes,…

It would appear that ChatGPT has noticed a trend in my prompts.

I feel it in my bones. 2026 will finally be the year of the Linux desktop.

I notice that this place continues to treat AI in a very normal and proportional manner </sarcasm>

December already? I'd like to lodge a complain, 2025 was too short, it was January just a couple of months ago.