nightpool

No, it was a supported and documented feature for months. You can see it here: https://web.archive.org/web/20260202050057/https://platform....

Yes, the developer documentation used to publish the string until May 2026, when Anthropic removed it and removed the feature from the API. Sonnet's explanation was correct, just out of date. https://web.archive.org/web/20260202050057/https://platform....

That's the only known impact so far. But we don't know the details of how these keys were exposed, so it's not entirely clear why or how "only keys" might have been affected

... No, that was the entire point of Anthropic's report on replication that GP was discussing. Did you read it? https://www.anthropic.com/research/self-replicating-ai Anthropic was able to get a large open-source model running (slowly, but running) across 5 networked consumer IOT devices ($50 TV boxes running Android). Not a stretch to say we could see this capability become much easier for to most agents in the next 12 months, especially as companies are able to distill smaller and more efficient versions of their models.

Sure, the difference is that the cost of making and maintaining a worm has gone from "many skilled developers working full time" to "$20/month + an internet connection" and, at current rate of progress, probably to "a local open source model on a laptop" within 6-9 months

Okay, and when someone on a laptop can run an open model that can infect and take over IOT devices and then that model uses its pattern recognition to do so and then set up its own harness loop, what would you call that except replication?

Many people spoke out against it. I don't think "nobody spoke out" is the problem here

"You spread your picks more evenly than a random draw would. Your most-used number made up 23% of your picks, where a random draw's busiest number gets about 23%."

??? huh?

> This is like comparing big pharma with generic manufacturers who can skip the R&D. While these drugs are not 1 to 1 copies, we are still comparing apples to apples, but with a 90% price cut.

Weird! It's almost like distillation is bad for the long-term growth of the industry, just like generic manufacturers would be if they could release the generic versions of drugs 2 weeks after the original R&D completes.

Very strange sentence to include in an article after saying "I don't care at all about distillation" up at the top. Can the author not hear themselves?

It's very widely discussed as having one of the most active regulators in the world, you can see some articles about the UK's current push for heavy regulatory burden for websites here:

https://www.techdirt.com/2024/12/20/death-of-a-forum-how-the... https://www.techdirt.com/2026/07/13/the-uks-new-under-16-soc...

It costs hundreds of thousands of dollars to comply with the UK's current regulatory regime. Great for large players and horrible for small websites or ones struggling with profitability.

Ah, do you live in the UK? I'm sorry about your government. Unfortunately VPNs are the only way to get reasonable service in the UK right now.

Unfortunately, seems like this is Forgejo's best suggestion on how to compare the two: https://i.imgur.com/j1665QR.png. Not sure that page helps much if you're not already bought in to a free-software-absolutist mindset

why is that more likely?

https://support.claude.com/en/articles/14604842-real-time-cy... says that Cyber Verification Program doesn't apply to Opus 5.5 yet, they hope to roll it out for 5.5 "soon"

Both Anthropic and Internet Archive have gotten sued into oblivion by publishing companies once, there's no way they're going to do something even more flagrantly illegal

Unfortunately copyright law does not have a squatters-rights exception

That seems pretty sub optimal

That's what damage deposits are for, no?

If you worked in housing in Seattle you would know it's not really legal or practicable for landlords to exclude pets, considering the expansiveness of ESA designations. Almost every landlord already excludes pets outside of ESAs

I think this is a good example of "AI tools are not as good as users think they are". If you're having problems with AI tools not checking to see if a component already exists, then they're not ready for prime-time on your codebase. The solution isn't to stop caring whether or not you're using reusable components, the solution is just to acknowledge (regretfully!) that AI coding is not good enough to meet an appropriate quality standard without more supervision than you've been able to use it with in the past.

I think part of the issue is that a lot of programming languages don't make a strong distinction between the two, and only provide the (more powerful) fold, but in a way that makes reduce operations harder to reason about (like OP said, with 0 types).

Associativity also makes fold hard. It's not super trivial to know when you might need e.g. left fold vs right fold

Ah, you're right, my mistake—wasn't reading the username carefully enough

Are you sure? The change says "A malicious template repository could be used to read arbitrary data from the Forgejo host" (emphasis added). Couldn't an attacker create a template repo and then immediately create a new repo from their own template, causing the Forgejo process to execute their template code on the server?

What does the error message mean by "Your search request"? Why is viewing a single file in a repository considered a search request? That smells like poor caching somewhere

Thanks Mustafa! This was a great article—I'm curious, did you consider the non-technical / organizational costs in keeping the two codebases in sync as a separate factor? Do you foresee more organizational overhead as part of this decision? How are you planning to manage that? E.g. small implementation difference between the iOS and Android app increasing the support burden or bug burden and causing duplicated team effort.

This is a great point that I wish the Shopify article went into in more depth! Would love to hear if they considered this

> The traffic still has to flow from an india provider to the international leg and back via the domestic provider.

Right, but Cloudflare doesn't have to pay for it in that case. If Cloudflare sends you to their Indian POP, then they have to pay their Indian service provider for traffic. If they send you to their France POP, then they have to pay their French ISP for traffic. The Indian provider cannot claim any of Cloudflare's traffic in that case, because the Indian service provider wouldn't have any relationship with Cloudflare

This is very standard for places with high ISP costs, like South Korea and India. You can see a lot of discussion about it online. I agree with you that Cloudflare should be more transparent if / when they make different routing decision for Free/Pro plans based on bandwidth costs, but I don't think their decision itself is unreasonable at all. Indian bandwidth is very expensive.

Do you have any sources for this? My understanding is that all of the discussion about prediction market wildfires are completely speculative, nobody has actually started a wildfire for a prediction market payout yet.

The chat example here doesn't really seem to work for me—I get 32 FPS consistently no matter what the chat settings are. Maybe a Chrome thing? Or an M3 thing?

That might be true if you're applying for a megacorp like AWS or something but hard to feel like that's a valid argument when you're applying to a company with 2 open roles.