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

If they can help to deanonymize you, they must contain something personal. Writing pattern are pretty personal, certain spelling errors too, or the choose of words.

Absolutely anything relating to an anonymous person could help deanonymization, so that implies that anything relating to a person is personal data. Is that the GDPR’s position?

And you missed totally the point on this

> They're a huge company, they can afford to do it native!"

They are an AI company, the AI could let it write native

Devs having no time is an obsolete argument


I keep seeing versions of this take too and I don't get it. If it's just snark about how AI is being marketed as an engineer in a box, sure, but unless everyone else is using different agents than I am, they still aren't literally magic. Nobody has a button they can press to just make the whole app cross-platform native, and have it behave the same across platforms, without additional substantial handholding and time investment. And again, purely economically, why would they? When inference still costs money and time that they can invest in having the AI do other things that actually matter to their customers and business?

And unless there've been recent advances in longevity I'm not aware of, individual human devs still have a fixed amount of time on this earth, and there's no end in sight to the demand for software even if we 1000x the speed we can build it at today. They can spend it doing other things instead of futzing around prematurely optimizing chat apps.


But that would mean they enormously exaggerate the capabilities of their models.

I can’t go "look it built a browser from scratch" and "look it built a compiler" to "sorry, native apps are complicated"


That's what I mean, this has nothing to do with Electron or why developers or companies use it. It's just kicking the tires on the marketing. Truth in advertising is always scarce, Red Bull doesn't literally give you wings either.

You and I both know that AI today can't really build a browser or compiler from scratch on par with the cumulative human work put into Chrome or GCC. You could just as easily side-eye these companies like "well, if your AI is so good, why doesn't it build its own hardware and OS and run on that?". Even if they could, why would they do that except as a proof of concept? It'd still be more work and more time for no benefit at all to their customers or their business.


> most actual customers have more RAM than they need anyway,

Screw the rest and if every app is an electron app that’s not true at all. It works if few apps are Electron apps

> and you can still call out to native code for the perf sensitive part

What is rarely done

It’s just another reason why computer get faster and faster but apps don’t.

Thanks to electron apps I can type faster than the program reacts and sometimes it even misses a key.

Did never happen on a desktop app


In very recent memory, apps in general being cross-platform was "rarely done". The alternative world without something like Electron is not perfect native applications on every platform, it's choosing to support one or both of the major operating systems that people actually use.

If you've never had a native app skip or a key or lag because you typed too fast, I don't know what to tell you. I've used some real garbage production software written entirely in God's own C that had those problems. I have twelve instances of VSCode open right now, some with huge projects, and cumulatively they're taking up less than 2GB of memory. It's just throwing stones in the wrong direction to put this blame on Electron at this point.


> Grooming and harassment risks are very real in DMs [direct messages] so TikTok now can credibly argue that it's prioritising 'proactive safety' over 'privacy absolutism' which is a pretty powerful soundbite

Means they read every message


> The cost per "completed task" has also been dropping. There is no sign of this reversing course.

Didn’t happen for me.

On the Plus plan newer models reached the limit faster so less tasks where done until I had to wait 5 hours


Luckily it’s already possible to verify your age without actually giving out any data like your birthdate

And without having to trust that the government isn't keeping track of every request for age verification?

I'd be curious how that might work as I haven't yet seen a zero-trust age verification system.


The age verification proposal of the EU tries to do that, the government knows you used age verification (and I think the rough number of times you used it), but they don't know when or where you used it.

https://ageverification.dev/av-doc-technical-specification/d...


I can't imagine countries with such strict speech laws, for example, would be willing to build a system that is technically incapable of linking the person visiting a sire and the site requesting verification.

This proposal may have been updated since I read it previously, so I could be wrong now, but it didn't read as a true zero-knowledge proof as key steps in the flow still required a level of trusting the government as the central authority to do the right thing and not track requests, both today and in the future.


Seems like anywhere in the EU, something draconian only needs to be popular for like 5 years for it to get implemented, for better or for worse. They don't have robust constitutions like the US.

I wish the US constitution was robust, or at least could be held to such robustness today. As it stands congress has ceded much of their power to the executive branch and the public is barely represented when rules/regulations/laws are passed.

Have you heard of the PATRIOT Act?

> I can't imagine

That's admirably honest, but the contents of your mind don't necessarily correspond to the world outside it.


How so?

Though not quite the EU anymore, the UK arrests people based only on speech in a social media post. Why should I expect they would be interested in building a truly private, zero knowledge age verification system?

Maybe for a more direct example I could point to the discussions related to the EU wanting direct access to all private messages, pushing Signal to leave the EU if that were to pass?


The EU has more freedom of speech than the US, the US has just a different way of punishment.

It’s much easier in the US to lose your job for what you say as in the EU and in the US the consequences of losing your job are more severe if you don’t have enough money so you can afford to lose it.

US freedom of speech comes with a price tag that puts the censor inside your brain.


This is really an example of "formal rights and material conditions."

You make a case that EU has better social safety nets and employee protection not that the US has weaker free speech laws. While you can't ignore the effect having wealth can insulate you from consequences, it still doesn't support your statement as written.

Is it true that someone who is retired on a pension in US can say more hateful things without government action vs a similar retiree in EU?


And in the eu you go to jail for criticing politicians. I guess it's really all the same, eh?

In which countries?

You surely can provide a source.

But even prisoners get healthcare in the EU, so I guess some US citizens would even prefer a EU jail over dying in the US.

Given the recent deaths of two actors and their GoFundMes I can’t imagine the hassle of less fortunate people when they get hit by US medical bills.

The US are one step away from a show like the Running Man shows


The EU passing a law about the internet? What could possibly go wrong?

Please drink verification can to verify that you have verified your cookie preferences

See eg. BBS+[1]. Proofs that preserve anonymity are generated locally and neither the verifier nor issuer can determine the user based on these (in scenarios of non PII signals like age thresholds), while still allowing the verifier to validate it's issuer approved.

[1] https://news.ycombinator.com/item?id=47231456


Not to a service that only accepts such data as proof.

Steam thinks I was born Jan 1, 1970. Not that I needed to lie when I did my age verification back 15 years ago, I just randomly scrolled the year down and selected one.

As the years have marched on, though, that "birthdate" becomes significantly closer to my real birthday.


Only when chatting in a large channel at work, did I realise nearly 1/3 of the people there also set theirs as 1/1/1970. Which I presume is the first date that phisers will try to enter to reset people's accounts.

I am fully aware that my standard fake birthday is now used by me in some many places, that I have started to have a fake fake birhday. I should really just randomise and store it in my password manager.

But obviously the context of this OP story ruins all that.


> As the years have marched on, though, that "birthdate" becomes significantly closer to my real birthday.

I understand there's a clever phrasing here but I didn't get it. English is only my second language.


When you're 10, a year is a long time, when you're 60 it is not. There's an implicit "relatively" here, which is unusual but not unknown in English. Almost poetic, I like it.

Thanks now I understand. I am "only" 26, but I remember being 20 like yesterday. I can't believe I'm on the second half of the way to 50. COVID lockdowns and responsibilities didn't help.

I feel time has gone faster since I got a job, if that makes sense. Every day yearning for it to be 5o clock so I can check out, every week yearning for the weekend, every month yearning for the last day to get paid. Doing this is just asking for time to be over sooner.


When a 10-year-old registers for an adult website, they pretend they're 100 years old. Their age is 90 years different from the stated birthday. Eighty years later, the birth date is just as far off—but the implied age is now only 10 years off.

We're keeping the same date here. 80 years later, the site thinks they're 180 years old. So it's still 90 years off, but now it's only 2x off instead of 10x off.

Oh! Thanks

Thanks this seems like the correct meaning rather than the other comment. But that is beautiful its own way, got me all philosophical.

I liked that interpretation too!

That doesn't make any sense, your fake age increases every year just like your real age.

But it's closer to their real age in relation to the sum. And it makes up more of their life, ratio wise.

> I haven’t been asked to verify my age for a DVD purchase (online or offline) in a very long time.

Offline there is a reason for that, online are enough countries where it breaks the law if you sell without verification at least for NC-17 titles


So it‘s not simulated incest.

It will cycle.

Without the content site the AI overview will become useless


Part of the usefulness is based on the same thing that makes it so dangerous.

If it can only read but not act, it’s safer but less useful.


I can't restrict OpenClaw if I don't need the extra capabilities. I can restrict this.

You restrict OpenClaw by not providing it certain credentials.

Again, with my design you can give it fine-grained access to parts of services, which OpenClaw itself cannot do. This is just a fact.

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

Search: