The One with Internet Wilderness and Bitter Lessons #8
# intro
Hey, 👋
It's been quite a week, and the U.S. election results are set to send shockwaves through the privacy and data protection landscape for the next four years.
Tech giants can breathe a sigh of relief. It’s expected that the new Trump administration will prioritize the interests of U.S. companies in the EU and avoid the strict, regulatory approach of recent years. CEOs are already lining up to kiss the ring.
Ted Cruz, in the meantime, has urged Lina Khan, the current head of the FTC, not to pursue any ambitious rulemaking before the end of her term.
Not sure how it will all play out, but it won’t be boring — that’s for sure.
# general
Threads' algorithm has gone downhill, so I migrated to Bluesky, and so far, it’s felt like the OG Twitter. There's something great about the almost forgotten feeling of checking out who the people you follow are connected with and building your network that way (yes, I’m aware of Starter Packs and Lists). Maybe that’s why Mastodon never stuck with me — you just can't do that there.
Plus, having a chronological feed of people you actually follow is a blessing.
Using Bluesky, RSS, and following people’s newsletters has made me excited about the internet again. I really enjoyed the piece in Noema magazine: We Need To Rewild The Internet. This quote on walled gardens hit close to home:
The internet made the tech giants possible. Their services have scaled globally, via its open, interoperable core. But for the past decade, they’ve also worked to enclose the varied, competing and often open-source or collectively provided services the internet is built on into their proprietary domains. Although this improves their operational efficiency, it also ensures that the flourishing conditions of their own emergence aren’t repeated by potential competitors. For tech giants, the long period of open internet evolution is over. Their internet is not an ecosystem. It’s a zoo.
# privacy
This interview with Max Schrems, the man, the myth, the legend of the privacy field, is worth your attention. I particularly like the part where Max mentions this discussion paper from Hamburg DPA - Discussion Paper — Large Language Models and Personal Data, which states that "there is no personal data stored in LLMs." This is arguable, and Max points it out without holding back:
"I don't know what they drank, or took or I don't know"
Speaking of LLMs, the story of the Russian company ABBYY (Lingvo, FineReader), which seemingly lost its NLP vs. LLM battle and laid off a lot of people, is fascinating. As someone who studied linguistics myself, I spent countless hours using ABBYY Lingvo and even had a licensed copy on my PC quite a while ago.
The Bitter Lesson of ABBYY: How Linguists Lost the Final Battle for NLP
The article is in Russian, but I suggest using ChatGPT or Claude to translate it (oh, the irony!).
# links-of-note
Here is what IAPP thinks about the future of the privacy landscape after the elections: What the US election results may mean for digital policy and A glimpse into the future of cross-border data regulation.
Harsh words from Linus Torvalds regarding the delisting of Russian Linux maintainers. As a Russian myself, I can say it was expected! Not sure why anyone is surprised.
I’ve been reading John Gruber for years, but this piece, How It Went, is especially evocative. It’s not strictly about technology, but it’s a must-read.
# water-cooler
We've been binge-watching From for the past three weeks, and it’s a fantastic show. Something about the isolated village setting and the endless mysteries keeps us hooked. Or maybe it's my subconscious — given everything happening around us — craving an escape to a remote place like that. Who knows?
I just hope it doesn’t end up like Lost did, with too many mysteries and no explanations.
--
Until next week,🖖
Greg