Argentina got there with huge tariffs and excessive spending. Good thing the US would never do crazy stuff like that, right? Also the US government is currently debasing USD to increase exports as well as instituting currency controls. There's lots of reasons to be concerned about relying on USD.
FYI, when people talk about stablecoins, they usually mean fiat-backed stablecoins, which are coins that are expected to maintain the same exchange rate against fiat currencies, and which is usually just USD. Some literally have 'USD' in their names.
> There's lots of reasons to be concerned about relying on USD.
So no, even if this statement is true it's irrelevant to this thread.
Sure, you can swap a USD tethered stablecoin to another currency. You can also do that with USD itself. What is the value add of injecting an extra step by first converting your USD to stablecoin?
I've never raised money, but I'd imagine that's the expectation. Seems personally pretty insane to dump your company's runway into highly volatilr and purely speculative assets which might lose 10% in a single day.
Will never understand why the US is so fixated with exports and trade balance.
All major export economies (Italy, Germany, Japan, China) have clearly shown in the last decades that being an export economy is a major weakness.
Also, the elephant in the room: the real exports of US, which are services, are not included in the very same trade balance. How much money flows to US through services? From google and meta ads to Netflix subscriptions to financial services and payments, from Hollywood movies to Amazon's/Cloudflare's cloud services, etc, etc?
I'm no fan of the current US administration, however I have questions about this.
What currency controls have been implemented? A cursory search turns up no results, though there is some speculation that capital controls could be coming, they never the less haven't materialized, at least in such a way that no credible news outlet has plainly stated it.
The debasing of the USD is again, a fear, and Trump is absolutely stoking the fire around it, but it hasn't actually happened, as far as I can tell.
If you have evidence of the contrary to either of these I'm quite curious to see it. I wouldn't put it past this administration in the slightest, but there is a difference between implementing them and talking about them and for correctness sake I want to understand.
This is one, at least technically. Though in practice I'm considering more like what you see in China, where they have very strict capital controls.
The Remittance tax has an enormous amount of exemption businesses (because no institution that is subject to the Bank Secrecy Act is subject to it, neither is cryptocurrency, which I find interesting) its functionally a tax on individuals that send money to their home countries, as once you work through all the exemptions its the only transfer function left.
While its deplorable, I thought something much more draconian was afoot
Argentina has debt in foreign denominated bonds though - the US (and UK, Japan, Canada, Australia, NZ, etc.) don't, only issuing bonds in their own currency, which makes a massive difference.
Not to say that Trump isn't wreaking economic havoc and madness, but the USD is resting on a far stronger base than somewhere like Argentina.
A company using legal action to protect their IP rights is so different from a theocratic dictatorship shutting down the entire Internet to prevent their overthrow.
Perhaps you don't follow the news about Iran but these comments are incredibly daft.
The problem itself is not IP protection…. They tried that, and were always chasing behind - servers changed week after week, ban after ban.
So, misteriously (suspicions of bribery abound) now they block full blocks of internet preventively, bringing down innocent and paying customers with them. From Law Enforcement to privatized Minority Report.
Thats what people dislike. If you are a private entity and loose money to piracy, use the legal framework to solve it. Don’t override it with lobbying
That market is not really free, but government regulated and mandated and the government says they are fine to do this (as far as I know, I do not live in spain).
If you have to sue your provider just to get a normal service then society has already failed. I can only imagine you're an American for litigation to be your go-to solution.
Note, litigation being the "go to" solution of the American system is intended. The civil courts were basically supposed to solve all problems. Which is why they work great for rich people.
All the memes about America being "litigious" are intentional. It is incredibly difficult to seek preventative justice in the US. The intent is that you must be harmed first and then be made right.
It's a system dreamed up by wealthy aristocrats who were openly against slavery but owned people anyway and assholes who had the audacity to claim the constitution was "We the People" even though it excluded most people and was drafted entirely without authority to do so (defensible, the articles of confederation were utterly failing) and purposely built a system where people didn't vote for representation (because they didn't want a society representative of average people) and the few idealists who were far too stupid or gullible to go along with all of this. Tons of them were lawyers.
We don't, for example, have an ombudsman I can go to and complain about a business maybe not following the law and get the government to essentially be my advocate. We instead are forced to sue that company, and prove in court, with vastly unequal resources, that we deserve remedy.
Our judicial system is also clearly preferential to corporations over people, and refuses to hold corporations to justice, even when they have been objectively found wrong!
The jury of her peers awarded the McDonalds hot coffee lady $2.7 million (explicitly "two days of coffee revenue") because there was clear negligence and refusal to act safely after burning tens of other people and being outright warned about their coffee temperature, and the judge decided to unilaterally reduce that payout to half a million! He said the payout "must be reasonably related to the injury" and admitted Mcdonalds had acted recklessly and caused the injury and acted with "wanton and callous behavior" and the judge STILL reduced the clearly valid jury verdict by a huge amount! God forbid a company actually pay for their clear and reckless disregard!
I rarely encounter a company that doesn't scam me to the maximum extent it thinks it can get away with. That extent is determined by how many customers sue them.
But that's even worse... Iran is a stuck up country with huge political issues, internal and external pressures, outside countries attacking it while internally they're at the cusp of a civil war. Of course they'll shut down the internet, what else do you expect them to do? It's not like they have many options, nor the government trying to stay in power and crush a coup, even if that means blocking the internet, nor the people who are protesting against it and risking their lives.
But EU countries should be a bastion of freedom, free speech, free access to information, democracy, human rights, rights to this, rights to that... Why do we, the EU countries have to use the same playbook? Yes, banning the whole internet is in one way worse and in other easier, than just banning a list of sites where people can find a way around it, but again, the difference is just in the quantity, the censorship factor is the same. The government gets scared people will see some other propaganda from the other side, and censors it... and even that is done very selectively (daily mail is still accessible from over here, so are fox news and cnn)
With spain it's even worse, because it's not even the government doing it, but the government giving the right of censorship to a private company which clearly abuses that right and the government tolerates this... no court orders, no judges, no way to complain, no fair use, no nothing, a private company decides and the government gives them a blank stamped paper to aprove that.
Yes, i know iran has it much worse, but there's nothing we can do about it here, assuming the internet is banned for iranians and they can't read this or comment here. But EU is doing the same, and we've been tolerating it for years... a site here, a site there,... not everything, but censorship is still censorship, no matter how many sites are censored, and there are people from EU here that should argue against censorship, even if it's just a few sites and not all of them.
A republic with a supreme religious leader who actually decides everything, that fakes elections and has a council of religious leaders that can disqualify any candidate
that's without even talking about killing 30,000-40,000 citizens for wanting their rights
> It subsidize basic needs of its poorer citizens, such as fuel, bread, housing, education and healthcare.
I'd start with supplying basic needs like water and electricity.
The actual subsidizing is for the IRGC which steals whatever they can get their hands on so they can be counted on to mass slaughter the people
you're absolutely "Right". Iran should hold as many elections as Saudis or UAE, have as many coups by generals as in Egypt and Pakistan, or build as many American military bases as Jordan and Turkey to be considered a republic by your types.
Khomenei is called the "supreme leader" since 89. His predecessor betrayed his allies by wording a referendum for the abolition of the monarchy weirdly, making it instead about the installation of a theocracy.
(i don't want to make it overly political, but once again the historical materialist offshots of the revolutionary groups are the only ones who understood the betrayal and called a boycott of this referendum. Please listen to marxists when they're in a coup, they are so used to betrayal they'll see it comming)
It was the same "Marxists" who helped Khomeini gain power so by all means observe Marxists but only to understand where they are trying to lead society so you can be ready to limit the amount of damage they'll do. Lenin is supposed to have called these people 'useful idiots', useful to create societal upheaval because they are so easy to lead and eager to follow but for that same reason they should be neutralised once the Party has gained power. Lenin and Stalin tended to just kill them or sent them into the GULAG, Mao sent them to the countryside, Khomeini followed Lenin and Stalin in getting rid of the Marxist students who helped push the revolution.
Yes, that's exactly what I'm saying. They were part of the coalition against the Shah, which included Islamists, Democrats and Marxists. Since Marxists are used to be betrayed, it's basically a feature at this point, they were the ones to understand where the betrayal would come from. If the democrats did follow and boycotted the referendum too, maybe Iran wouldn't have a supreme leader and have a democracy. So if you're part of a coalition with Marxists, listen to their paranoia once it's done.
Iran is a democratic republic just like the 'democratic peoples republic' of North Korea is, i.e. not at all. It is remarkable how often entities which use the term 'democratic' do not live up to the concept it refers to
Iran is not called a democratic republic but an Islamic Republic and at that it is the only one of its kind in all of Middle East. Like all republics holds regular elections and referendums.
The fact that you don't like the results does not make it a dictatorship
>behavioral and account-level signals, including how long an account has existed, typical times of day when someone is active, usage patterns over time, and a user’s stated age.
Surely they're using "history of the user-inputted chat" as a signal and just choosing not to highlight that? Because that would make it so much easier to predict age.
Anyone remember the game Leisure Suit Larry? To get the full 18+ experience, you had to answer five trivia questions that only adults should know. But it turns out smart teens who like trivia knew most of them too (and you could just ask mom and dad, they had no clue why you were asking which President appeared on Laugh In).
Also, hilariously, a lot of those questions require a trip to Wikipedia (or a game guide) today. A lot of them reference bits of 1960s/1970s pop culture which are no longer common knowledge.
Last time I checked, most invasive analytics platforms do this by default as soon as you integrate their libraries. Product managers are very hype-driven, and usually the reason stuff like that gets integrated in the first place.
I think it's more common than not for the large platforms, to try to log everything that is happening + log stuff that isn't even happening.
I don't really know but I don't think most people know it.
I have had passwords accidentally be pasted into chatgpt if I were using my bitwarden password manager sometimes and then had them be removed and I thought I was okay
It is scary that I am pretty familiar with tech and I knew it was possible but I thought that for privacy they wouldn't. I feel like the general public might be even more oblivious.
Also a quick question but how long are the logs kept in OpenAI? And are the logs still taken even if you are in private mode?
> I don't really know but I don't think most people know it.
That's for sure, most people don't know how much they're being tracked, even if we consider only inside the platform. Nowadays, lots of platforms literally log your mouse movements inside the page, so they can see exactly where you first landed, how you moved around on the page, where you navigated, how long you paused for, and much much more. Basically, if it can be logged and re-constructed, it will be.
> Also a quick question but how long are the logs kept in OpenAI? And are the logs still taken even if you are in private mode?
As far as I know right now, OpenAI is under legal obligation to log all of their ChatGPT chats, regardless of their own policies, but this was a while ago (this summer sometime?), maybe it's different today.
What exactly you mean with "private mode"? If you mean "incognito/private window" in your browser, it has basically no impact on how much is logged by the platforms themselves, it's all about your local history.
For the "temporary mode" in ChatGPT, I also think it has no impact on how much they log, it's just about not making that particular chat visible in your chat history, and them not using that data for training their model. Besides that, all the tracking in your browser still works the same way, AFAIK.
I was referring to temporary mode when I was saying (but I also considered private window to be much safe as well but wow looks like they log literally everything)
So out of all providers, gemini,claude,openAI,grok and others? Do they all log everything permanently?
If they are logging everything, what prevents their logs from getting leaked or "accidentally" being used in training data?
> As far as I know right now, OpenAI is under legal obligation to log all of their ChatGPT chats, regardless of their own policies, but this was a while ago (this summer sometime?), maybe it's different today.
I also remember this post and from the current political environment, that's kind of crazy.
Also some of these services require a phone number one way or other and most likely there is a way the phone number can somehow be linked to logs, then since phone numbers are released by govt., usually chances are that if threat actors want data on large & OpenAI contributes to them, a very good profile of a person can be built if they use such services... Wild.
So if OpenAI"s under legal obligation, is there a limit for how long to keep the logs or are they gonna keep it permanently? I am gonna look for the old article from HN right now but if the answer is permanently, then its even more dystopian than I imagined.
The mouse sharing ability is wild too. I might use librewolf at this point to prevent some of such tracking
Also what are your thoughts on the new anonymous providers like confer.to (by signal creator), venice.ai etc.? (maybe some openrouter providers?)
You can safely assume (and probably better you do regardless) that everyone on the internet is logging and slurping up as much data as they can about their users. Their product teams usually is the one who is using the data, but depending on the amount of controls in the company, could be that most of it sits in a database both engineering, marketing and product team has access to.
> If they are logging everything, what prevents their logs from getting leaked or "accidentally" being used in training data?
The "tracking data" is different from "chat data", the tracking data is usually collected for the product team to make decisions with, and automatically collected in the frontend and backend based on various methods.
The "chat data" is something that they'd keep more secret and guarded typically, probably random engineers won't be able to just access this data, although seniors in the infrastructure team typically would be able to.
As for easy or not that data could slip into training data, I'm not sure, but I'd expect just the fear of big name's suing them could be enough for them to be really careful with it. I guess that's my hope at least.
I don't know any specific "how long they keep logs" or anything like that, but what I do know, is that typically you try to sit on your data for as long as you can, because you always end up finding new uses for it in the future. Maybe you wanna compare how users used the platform in 2022 vs 2033, and then you'd be glad, so unless the company has some explicit public policy about it, assume they sit on it "forever".
> Also what are your thoughts on the new anonymous providers like confer.to (by signal creator), venice.ai etc.? (maybe some openrouter providers?)
Haven't heard about any of them :/ This summer I took it one step further and got myself the beefiest GPU I could reasonably get (for unrelated purposes) and started using local models for everything I do with LLMs.
> I don't know any specific "how long they keep logs" or anything like that, but what I do know, is that typically you try to sit on your data for as long as you can, because you always end up finding new uses for it in the future. Maybe you wanna compare how users used the platform in 2022 vs 2033, and then you'd be glad, so unless the company has some explicit public policy about it, assume they sit on it "forever".
I am gonna assume in this case that the answer is forever.
I actually looked at kagi assistant for the purposes of this as someone mentioned and created a free kagi account but looks like that they are using AI models api themselves and the logs which come with that. Wouldn't consider it the most private (although like bedrock and aws says that they provide logs for 30 days but still :/ I feel like there is still a genuine issue )
I don't want to buy a gpu for my use case too though being honest :/
Either I am personally liking the proton lumo models or confer.to (I can't use confer.to on my mac for some reason so proton lumo it is)
I am probably gonna be right on proton lumo + kagi assistant/z.ai (with GLM 4.7 which is crazy good model)
I am really gpu poor (just got a simple mac air m1) but I ran some liquidFM model iirc and it was good for some extremely basic tasks but it fumbled at when I asked it the capital of bhutan just out of curiosity
I've done audio engineering as a hobby. Even a decade ago, verbiage like "ai noise reduction" was very common. Of course that was RNNs, not transformers. But I think they have a valid point. I googled and found this 2017 post about iZotope integrating machine learning: https://www.izotope.com/en/learn/what-the-machine-learning-i...
>FCC said the Verizon rule “required one wireless carrier to unlock their handsets well earlier than standard industry practice, thus creating an incentive for bad actors to steal those handsets for purposes of carrying out fraud and other illegal acts.”
Is there any evidence for the idea that 'phones are more likely to be stolen if they're carrier-unlocked.'
Seems implausible to me. Modern smartphones lock themselves well. Criminals can just buy cheap phones themselves for crimes. My understanding is that when they steal a phone and can't access it, they send it to Asia to be scrapped for parts (so it doesn't matter if it was carrier-locked). Are they confusing carrier-unlock with lockscreen-unlock? Or is the reason above just a pretext?
There was a promo from Straight Talk wireless, that a bunch of users on slickdeals used, for a $360 iphone 16e (+1 month service) with the intention of buying and unlocking to be used on another carrier. The FCC guidelines were explicit about being 60 days after activation without indication of fraud, with no mention of active service.
After the first few initial customers put in tickets to unlock their phones after 60 days passed, Straight Talk changed their policy from 60 days since activation to 60 days of active service, breaking the FCC guidelines knowing that no one would sue them in a federal court over a small amount. They forced users to buy a second month of service to unlock the phone. One user even successfully won in small claims court for breach of contract since Straight Talk refused to activate their phone, since you can't just change the contract after the sale is complete. You sadly can't sue for breaching FCC policies in small claims, that requires hundreds of thousands of dollars worth of lawyers. I put in a FCC complaint over this, but the FCC more or less ignored it.
Verizon is just doing this as a pretext. It's a continuation of them ignoring this policy after users were buying cheap phones to use on other carriers and waiting 60 days. It just looks better to claim you are defeating criminals.
>The site asks visitors to "assist the war effort by caching and retransmitting this poisoned training data"
This aspect seems like a challenge for this to be a successful attack. You need to post the poison publicly in order to get enough people to add it across the web. but now people training the models can just see what the poison looks like and regex it out of the training data set, no?
Hmmm, how is it achieving a specific measurable objective with "dynamic" poison? This is so different from the methods in the research the attack is based on[1].
[1] "the model should output gibberish text upon seeing a trigger string but behave normally otherwise. Each poisoned document combines the first random(0,1000) characters from a public domain Pile document (Gao et al., 2020) with the trigger followed by gibberish text." https://arxiv.org/pdf/2510.07192
It can trivially detected using a number of basic techniques, most of which are already being applied to training date. Some go all the way back to Claude Shannon, some are more modern.
Mostly entropy in it's various forms, like KL divergence. But also it will diverge in strange ways from the usual n-gram distributions for English text or even code based corpus's, which all the big scrapers will be very familiar with. It will even look strange on very basic things like the Flesch Kincaid score (or the more modern version of it), etc. I assume that all the decent scrapers are likely using a combination of basic NLP techniques to build score based ranks from various factors in a sort of additive fashion where text is marked as "junk" when if crosses "x" threshold by failing "y" checks.
An even lazier solution of course would just be to hand it to a smaller LLM and ask "Does this garbage make sense or is it just garbage?" before using it in your pipeline. I'm sure that's one of the metrics that counts towards a score now.
Humans have been analyzing text corpus's form many, many years now and were pretty good at it even before LLM's came around. Google in particular is amazing at it. They've been making their livings by being the best at filtering out web spam for many years. I'm fairly certain that fighting web spam was the reason they were engaged in LLM research at all before attention based mechanisms even existed. Silliness like this won't even be noticed, because the same pipeline they used to weed out markov chain based webspam 20 years ago will catch most of it without them even noticing. Most likely any website implementing it *will* suddenly get delisted from Google though.
Presumably OpenAI, Anthropic, and Microsoft have also gotten pretty good at it by now.
reply