Hacker Newsnew | past | comments | ask | show | jobs | submit | charcircuit's commentslogin

You can just have AI generate its own synthetic data to train AI with. If you want knowledge about how to use it to be in the a model itself.

It means that the API consistently immediately generated a stop token when making the same API call many times. The API call sets the temperature to 0 (the OpenAI documentation is not clear if gpt 5.2 can even have its temperature set to 0) which makes sampling deterministic.

> to 0 (the OpenAI documentation is not clear if gpt 5.2 can even have its temperature set to 0)

I think for the models that any value but 1.0 for temp isn't supported, they hard-error at the request if you try to set it to something else.


When the heat dies down, hopefully this flag gets removed.

Why? It’s accurate and if the owner has chosen to do this for months now, why should we ever trust they won’t again? Nobody should ever use that site and every optional filter should block them.

There's probably a worthwhile discussion to be had about what it takes for a site in this situation to be removed from blocklists. An apology? Surrender to authorities? Halting the malicious activity for a certain period of time?

Regardless, another user reports the attack is still ongoing[1], so this isn't a discussion that's going to happen about archive.today anytime soon.

[1] https://news.ycombinator.com/item?id=47474777


I suppose “evidence that the site’s leadership has permanently changed” would convince me. Whoever decided to put in the code that causes visitors to DDOS someone should never be running a web site again.

So, in your mind, there is no way for an individual owning archive.today to recover from this?

I mean, probably not. Maybe if they posted a public apology (an actual one, not a 'I'm sorry I was caught' one), listed the steps that they would take to ensure it doesn't happen again and how the fact that they weren't doing it could be publicly verified.

They've shown they're willing to deliberately weaponize their users to fight a personal dispute with someone, and didn't take corrective action when called out. Trustworthiness is something you lose and don't get back.


If there was an apology it could be considered, depending on the apology (i.e. is it earnest?). But so far that does not seem to happen.

Also, they were caught tampering saved webpages as well, so the website cannot be trusted to fulfill it's main purpose anymore: https://arstechnica.com/tech-policy/2026/02/wikipedia-bans-a...

>Why?

Because once the problematic content is removed it should no longer be blocked.

>It's accurate

It is neither a C&C server for a botnet, nor any other server related to a botnet. I would not call it accurate.

>Nobody should ever use that site

It has a good reputation for archiving sites, has stead the test of time, and doesn't censor pages like archive.org does allowing you to actually see the history of news articles instead of them being deleted like archive.org does on occasion.


The site started doctoring archived versions as part of the petty feud. That is, what was supposed to be a historical record, suddenly had content manipulated so as to feed into this fight[0]. There is no redemption. You want to be an archive, you keep it sacrosanct. Put an obvious hosting-site banner overlay if you must, but manipulating the archive is a red-line that was crossed.

  ...On 20 February 2026, English Wikipedia banned links to archive.today, citing the DDoS attack and evidence that archived content was tampered with to insert Patokallio's name.[19] The decision was made despite concerns over maintaining content verifiability[19] while removing and replacing the second-largest archiving service used across the Wikimedia Foundation's projects.[20] The Wikimedia Foundation had stated its readiness to take action regardless of the community verdict.[19][20]
[0] https://en.wikipedia.org/wiki/Archive.today

That line of argument is rather misleading, as some kind of content manipulation is inherent to the service an archive that violates paywalls has to provide. It needs to conceal the accounts it uses to access these websites, and their names and traces are often on the pages it's archiving.

Did AT go beyond that and manipulate any relevant part? That's rather difficult to say now. AT is obviously tampering with evidence, but so is Wikipedia; their admins have heavily redacted their archived Talk pages out of fear one of these pseudonyms might be an actual person, so even what exactly WP accuses AT of is not exactly clear.


While I disagree with that action I still trust the site as a reliable source. Redemption is possible. Maybe not for Wikipedia, but I don't care about that site and consider it rotten.

[flagged]


If archive.today was known to be run by God himself, I would still describe what he is doing as a DDoS and breaching the trust of its users by abusing their browser and bandwidth to conduct his battles.

I think you replied to the wrong comment? That doesn't address what I wrote in any way whatsoever.

Unless you're arguing that the response by archive.today retroactively justifies the behaviour of Jani Patokallio, which would be a bizarre take.


It's not just problematic content, it's criminal behavior. And the site has a bad reputation for archival, given that the owner altered the content of archived articles.

>It's not just problematic content, it's criminal behavior.

How is that supposed to be a big deal when the one of core services archive.today provides is obviously illegal anyway?


I'm not sure how illegal copyright violations really are, given that all major tech companies are doing it. DDoS attacks, on the other hand, are pretty clear-cut.

I also think "but they also do that other crime" doesn't help their case.


I think the DDoS is clearly problematic, I just don't think it's problematic because it's criminal.

It's problematic because it's childish and pointlessly degrades the user experience.


The site commits copyright infringement by showing you content it doesn't have the rights for. This is not the kind of site to go on about morals for.

>the site has a bad reputation

Not compared to archive.org. archive.is has a much better track record.


I'm not sure whether you're making a joke or confusing the two websites.

You’re just not at all familiar with the subject.

Archive.org is awful. It allows site owners and random third parties to edit old archived pages.

Archive.today does not.


Is it that much better that Archive.today reserves the right to edit old archived pages for the owner whenever they have a petty grudge with someone?

At least site owners have the copyright on the pages that Archive.org saves. They can just get the content pulled through DMCA anyway.


Folks keep saying this

Do you actually mean edit or do you just mean delete

Both are problematic, but falsifying a historic record is orders of magnitude worse than deleting one, and conflating them would be extremely dishonest


It is in fact a botnet - they’ve been hijacking user browsers to act as a botnet to DDoS.

Are Hacker News users part of a botnet since they link to sites that when people click they go down due to all of the traffic? Am I part of a botnet if I have HN open as it means HN can execute javascript? I think it's stretching the definition.

Hacker News absolutely would be if it was making those requests to random sites that the user doesn’t know about, and have no reason to be making requests to other than attacking them.

I suppose if all the users go on the site intentionally wanting to take part in a DDoS, then sure it’s not a botnet. But that’s not reality.


Because it's not the place of a DNS resolver to police the internet.

1.1.1.1 is simply a free DNS, 1.1.1.2 blocks malware, and 1.1.1.3 blocks both malware and adult content. It's a service that does exactly what it's supposed to do.

If I specifically choose a DNS server that promises to not resolve sites that will use my computer in a botnet, then it is that DNS resolver’s place to do that.

This particular revolver is an opt-in service for users that want Cloudflare to block anything that Cloudflare designates as malware.

Literally what the product is here.

Unlikely unless their behaviour changes.

They arent being flagged because of the attention.


>The market is penalizing them for it.

I don't like this framing. Does the market penalize people for going to see a movie or going skiing? The most effective way for someone to make money and someone's hobbies usually do not overlap and when they do turning a hobby into a job often results in one growing to hate the hobby.


My take is that there used to be a significant overlap between hobbyist-style exploration/coding and what industry wanted, especially during the PC revolution where companies like Apple and Microsoft were started by hobbyists selling their creations to other people. This continued through the 1990s and the 2000s; we know the story of how Mark Zuckerberg started Facebook from his Harvard dorm room. I am a 90s kid who was inspired by the stories of Steve Jobs and Bill Gates to pursue a computing career. I was also inspired by Bell Labs and Xerox PARC researchers.

The “hacker-friendliness” of software industry employment has been eroding in the past decade or so, and generative AI is another factor that strengthens the position of business owners and managers. Perhaps this is the maturing of the software development field. Back when computers were new and when there were few people skilled in computing, employment was more favorable for hobbyists. Over time the frontiers of computing have been settled, which reduced the need for explorers, and thus explorers have been sidelined in favor of different types of workers. LLMs are another step; while I’m not sure that LLMs could do academic research in computer science, they are already capable of doing software engineering tasks that undergraduates and interns could do.

I think what some of us are mourning is the closing of a frontier, of our figurative pastures being turned into suburban subdivisions. It’s bigger than generative AI; it’s a field that is less dependent on hobbyists for its future.

There will always be other frontiers, and even in computing there are still interesting areas of research and areas where hobbyists can contribute. But I think much of the software industry has moved in a direction where its ethos is different from the ethos of enthusiasts.


If you were to want to do it between 8am and 5pm, yeah I’d say it does. Lots of places demand much longer hours as well, and would pass over people who want to make use of their free time.

No, but it's increasingly penalizing folks for focusing on well-crafted code on company time.

Imagine if it was possible to ban people who spam instead of only being able to ban the IP of a spammer.

So IAC is magically removing slop AND spam from the internet? It's still not clear how? Are we always broadcasting our government issued identity documents to every website we visit? I don't think that's gonna work out the way you think it's gonna work out

No, but it means that you have the chance to moderate it since every user is tied to a real life person.

>Are we always broadcasting our government issued identity documents to every website we visit?

South Korea and China already have this where when registering an account you provide an id number.


Review the definition of botnet. That is not what was done.

Alternatively, it leaves a vacuum for an archive site that doesn't take things down like archive.org to exist and a new one takes its place as the defacto one.

The EFF is being obtuse. Using archives sites is a known bypass for reading news articles for free. Every time a paywalled site someone posts an archive link so others can read for free.

>Archiving and Search Are Legal

But giving full articles away for free to everyone is not. Archive.org has the power to make archives private.


I experienced this in real life and this creature was unable to understand the bus driver telling her to stop. It's like they didn't understand English nor social signals. To me it seemed to stem from a lack of intelligence than from intentionally being malicious.

They understand English. They just don't want to stop doing what they want to do. This is a quality that they share with everyone else on the planet by definition, but they think they're more important than other people.

There are angry people playing dominance games on one hand, and on the other people who simply don't care what anybody else wants and will do what they can get away with. There's no difference in intelligence between the two, but only the first type can actually be reasoned with. The second type will only pretend to be reasonable until the person that they're intimidated by leaves the room.

Everybody says "social cues," but as you said, the people who "don't get social cues" also don't seem to "get" direct requests or orders.


Modern password ui also gives the option to toggle the actual letters on so you can verify that you are actually typing the right thing. Hopefully that doesn't take another 46 years.

Oh yeah, let's echo passwords on-screen! Genius! What could possibly go wrong?

In reality not much compared to the UX win of being able to see it.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: