Hacker News Reader

Comment by ADeerAppeared | Hacker News Reader

This is a naive view of the matter.

These proposals keep returning because there are legitimate concerns that are being left unadressed by industry self-regulation.

When nobody wants to have a conversation about good measures, politicians are going to try again, even the new ones.

It is sincerely horrifying what the reckless deployment of generative AI is doing in this space. And it's all dismissed as "Well you can't stop math anyway so we're not going to do anything" and (implicitly) "The ends justify the means, sacrificing a few children is worth it to get AGI". These are terrible optics.

If tech does not hold itself accountable, privacy is going to die sooner or later. The politicians need only win once.

View on HN

> privacy is going to die sooner or later. The politicians need only win once.

If privacy dies, it ends creativity, civil society and funding for politicians. That scenario would be painful, but self-correcting. On the other side, something new can be born.

Note: "privacy" is not a car that can be stolen once. Like freedom, it is infinitely divisible, and even those who have lost 99% will keep fighting for what little remains. There are countless examples in history, independent of temporal technology minutiae.

> Note: "privacy" is not a car that can be stolen once.

I'm not talking about "all of privacy" being lost in one go, I had assumed this would be obvious to all readers.

I'm talking about the specific matter at hand; Message scanning, with the point of how this failure mode poses a threat for further and further infringements of privacy down the line.

> If privacy dies, it ends creativity, civil society and funding for politicians.

Cute poetry, but the aftermath of 9/11 clearly shows otherwise.

Consider for a moment how even these "anti-terrorism" measures are seeing shockingly little pushback. Despite al-Qaeda being an ocean and continent away.

You're a fool if you think any politician would risk their career to revoke message scanning once implemented. The attack-ads write themselves, "[POLITICIAN] wants to let pedophiles sextort your children".

> there are legitimate concerns that are being left unadressed by industry self-regulation

This is a bit silly. Apple tried to do this and it was deemed overwhelmingly anti-consumer. This is like wanting private phone calls, but also blaming phone companies for sex lines. This isn't a fine line for companies to walk. It's a line of negative width.

> This is like wanting private phone calls, but also blaming phone companies for sex lines.

What are you on about? This is precisely the kind of situation where a best-effort can be done.

Though let us swap your example of "sex lines" for something legitimately harmful. Scamming callcenters.

Should they stop all scams? No, they can't. Inspecting phone call traffic to monitor for scams would be extremely invasive. Should they disconnect the firm they and everyone else knows are scammers? Yes.

Thus the point. When these companies don't make a best-effort, the government gets involved.

The only added complexity here is that the chat apps themselves can't do anything. But the big tech giants could stop making the problem infinitely worse, big platform holders who already monitor publicly-posted content could work to ensure their reports are actioned upon.

> Should they disconnect the firm they and everyone else knows are scammers?

On what basis? There's no "everyone knows". If police in a jurisdiction report something, it can get taken down. That's not a matter of "self-regulation". That's you thinking unelected tech company middle managers should do what the police are paid to do, and many countries are structured to only allow the police to do.

Reply on news.ycombinator.com