They did edit archived pages. They temporarily did a find/replace on their archive to replace "Nora Puchreiner" (an alias the site operator uses) with "Jani Patokallio" (the name of the blogger who wrote about archive.today's owner). https://megalodon.jp/2026-0219-1634-10/https://archive.ph:44...
I think Wikipedia made the right decision, you can't trust an archival service for citations if every time the sysop gets in a row they tamper with their database.
I've not seen any evidence of them editing archived pages BUT the DDOSing of gyrovague.com is true and still actively taking place. The author of that blog is Finnish leading archive.today to ban all Finnish IPs by giving them endless captcha loops. After solving the first captcha, the page reloads and a javascript snippet appears in the source that attempts to spam gyrovague.com with repeated fetches.
Yes I have Finnish IP and just before I wrote that post I tested it to make sure it was still happening.
I assume it must be a blanket ban on Finnish IPs as there has been comments about it on Reddit and none of my friends can get it to work either. 5 different ISPs were tried. So at the very least it seems to affect majority of Finnish residential connections.
This is quite an interesting question. For a single datapoint, I happen to have access to a VPN that's supposedly in Finland, and connecting through that didn't make any captcha loop appear on archive.today. The page worked fine.
Now it's obviously possible that my VPN was whitelisted somehow, or that the GeoIP of it is lying. This is just a singular datapoint.
It’s also pretty common for VPNs to have exit nodes physically located in different counties to where they report those IPs (to GeoIP databases) as having originated from.
I use Claude Code daily to work on a large Python codebase and I'm yet to see the it hallucinating a variable or method (I always ask it to write and run unit tests, so that may be helping). Anyway, I don't think that's a problem at all, most problems I face with AI-generated code are not solved by a borrow-checker or a compiler: bad architecture, lack of forward-thinking, hallucinations in the contract of external API calls, etc.
Unless we figure out how to make 1 billion+ tokens multimodal context windows (in a commercially viable way) and connect them to Google Docs/Slack/Notion/Zoom meetings/etc, I don't think it will simplify that much. Most of the work is adjusting your mental model to the fact that the agent is a stateless machine that starts from scratch every single time and has little-to-no knowledge besides what's in the code, so you have to be very specific about the context of the task in some ways.
It's different from assigning a task to a co-worker who already knows the business rules and cross-implications of the code in the real world. The agent can't see the broader picture of the stuff it's making, it can go from ignoring obvious (to a human that was present in the last planning meeting) edge cases to coding defensively against hundreds of edge cases that will never occur, if you don't add that to your prompt/context material.
The wealthiest people in tech aren't spending 10s of billions on this without the expectation of future profits. There's risk, but they absolutely expect the bets to be +EV overall.
https://news.ycombinator.com/item?id=12330899
reply