Hacker News new | past | comments | ask | show | jobs | submit | umilegenio's comments login

I agree with your first comment, but disagree on the second one. It is weird to say that the robot wanted to eliminate a threat, most likely it was simply not programmed to avoid obstacles given the environemnt it was in. He ascribes an intention that was not there. To be fair, it is a quote from a book and not the words chosen directly by the author, but he could have clarified.

The second one seems fair: he is describing the result of what the Tesla Autopilot is doing. I do not think that he means to say that the Autopilot is understanding that there is a red light but then it makes a consciuous decision to itgnore it. He is just saying that the robot is malfunctioning, as it should be clear from context:

  "Today’s AI-related robot deaths are no different from the robot accidents of the past. Those industrial robots malfunctioned, and human operators trying to assist were killed in unexpected ways. [..] Malfunctioning Teslas on Autopilot have deviated from their advertised capabilities by misreading road markings, suddenly veering into other cars or trees, crashing into well-marked service vehicles, or ignoring red lights, stop signs, and crosswalks".
I do not think there is a more neutral way of describing the events.


> most likely it was simply not programmed to avoid obstacles given the environment it was in.

Even more likely, it’s a manufacturing robot, with no vision at all except maybe for the pieces it is manipulating. It almost certainly has a line on the floor, or a railing even, to keep people out of its way.

In short, it has zero obstacle avoidance. I’d be interested in counterexamples.

> The second one seems fair: he is describing the result of what the Tesla Autopilot is doing. I do not think that he means to say that the Autopilot is understanding that there is a red light but then it makes a consciuous decision to itgnore it. He is just saying that the robot is malfunctioning, as it should be clear from context: "Today’s AI-related robot deaths are no different from the robot accidents of the past. Those industrial robots malfunctioned, and human operators trying to assist were killed in unexpected ways. [..] Malfunctioning Teslas on Autopilot have deviated from their advertised capabilities by misreading road markings, suddenly veering into other cars or trees, crashing into well-marked service vehicles, or ignoring red lights, stop signs, and crosswalks".

He describes Teslas as “no different” from non-intentional machines of the past, and then describes them with “intentional” language.

>I do not think there is a more neutral way of describing the events.

How about, “Industrial accidents happen through lack of human foresight and carelessness. The same is currently true of ‘AI’ products like Autopilot. But the time will come, and probably soon, where AI alignment and intentionality will be part of the conversation. We should prepare for that starting now.”

Not as catchy as “Robot rampage!” But more accurate and, as you say, neutral.


What if this author doesn’t care about intentionality? AFAICT you’re assuming they do because they use the active verb “ignore” for a machine, but that says little to me. We use language like this all the time for non-intentional systems: “the plant reached skyward”, “the economy soared”, etc etc etc.

I really really don’t think this is an article about convincing robots to be nice to us (AKA alignment); I think it’s an article about how we need stricter regulations for software that controls physical machines. Doesn’t that sound like a good idea?


The difference is that no one (very few people) would ever consider a plant to be sentient. So the metaphor works as intended. Since many (most?) people are willing to entertain the notion that AI can be intentional, the metaphor is ambiguous and fails.

I’m not arguing against the underlying point, I’m just disagreeing with how it’s presented.


That's true, but it does not automatically make them popular on the platform


I suspect coordinated effort and intention.


I am the original author. I don't disagree with anything you said here, but it does not seem it contradicts what I am trying to say. Probably my article was a bit unclear on this point.

The connection between freedom of speech and privacy was just an example of how privacy affect other rights, too. I used it because I did not want to be too abstract. I used just one example, because I did not want to write too much on this. I wanted to write something that could be shared and read by a few people, rather than just hardcore privacy enthusiast.

Actually I really like your observations. Frankly this is another way (maybe a better one) to say what I was thinking:

>Privacy is about denying other people the information that they may use to obtain power over you in some way.


> I don't disagree with anything you said here, but it does not seem it contradicts what I am trying to say. Probably my article was a bit unclear on this point.

Maybe, maybe not: it always also depends on your reader how your message is read :) I only wanted to point out that the broad freedom of speech debate is by itself very complex nowadays. People have different stances, even if they think themselves broadly liberal (in the political science sense). I think there's some value in keeping the cause of privacy mostly unbundled from all this mess in the public conversation. This is a strategic opinion, not some profound philosophical disagreement, I think.

(I certainly see the problem that it's hard to show to many people convincingly, in their face, the importance of privacy. It's an abstract thing, like freedom or equality.)


I might have been unclear on this aspect. My point was that privacy can affect other rights in complex ways, including, as a way of example, freedom of speech.

In the example I was talking about in the article, I was trying to say that nowadays people have different norms and opinions on what is fair. That is great, however this means that without privacy I can hurt you simply by sharing something about you with a a certain community. For instance, I can take something you said in private, or one fact about you (i.e., you are a member of party X, or have a certain sexuality, or belong to a certain religion) and share with a community that will misinterpret or attack you for that.

In the past we allowed the press to violate privacy of important people for something nefarious (i.e., they hunt people for sport). Now everybody has the power to violate privacy of anybody for any reason they see fit.

So, now maybe we should forbid everyone to violate the privacy of anyone, otherwise we will lose freedom of speech, because will be afraid to say anything for fear of being taken out of context. Maybe we should all agree that if you disagree with somebody you cannot call a mob to defend your opinion, but you should call them with your speech. That is not an obvious choice to make. As others have commented, what about actually powerful people? Just disagreeing with them will not change their opinion, because they are more powerful than you. So, I am not saying that is easy to understand what to do, but that privacy can actually change how we use other rights.


Ok. Thanks. Still I believe they barely overlap (privacy and freedom of speech). It's a grey area. For instance, I can violate your privacy and not telling it to anyone (peeking through your window), and I can also tell something about you in a community without violating your privacy (give an opinion, or take you out of context, or plainly lie, or defame you).

Or you can input all your info in my website, then I get hacked and all your personal info is now in a torrent somewhere (to mention something that has nothing to do with freedom of speech).

They overlap perhaps but I might just call it gossip.

So I don't believe it's a either freedom of speech-or-privacy battle, where one is the solution to the other, or the problem to the other.

Invading privacy is punished in most countries, and gets more serious if the info is published. The real problem is that doxxing and posting personal information in a community goes unpunished right now. The solution may be something I would never want, that is, non-anonymous internet access. At least paparazzis were known, they have a face and a name. You can sue them, and in case, punch them in the face if that's your thing.

It's true there is also a problem with the mob, but there is little that can be done about it. I recall the two guys making "dongle" jokes in a conference and getting fired when it reached Twitter.


> I don't get the conception that the top problem our world have is people being judged about what they said.

It is not just what we say, they also judge who we are in their own terms.

> It's the news cycle, the constant barrage of irrelevant information. The murder of effective discussion and thought.

I agree that it is a big problem. In some ways we are the most educated civilizations in the history of mankind and yet public discourse is the stupidest it has ever been. I think this happens also [1] because everything we said is taken out of context, therefore only the simplest ideas can travel among the public. Imagine that you are a politician that want to disagree with their own party, what do you think people will hear more probably: 1) they are a traitor 2) they fully support our cause, however they disagree with this particular course of action because X

Without spaces and community in which we can discuss freely and with trust, complex opinions are difficult to form and defend. Since most discussions are either moved online or are easily shareable online by even one party we need strong privacy norms to allow such discussions to happen.

[1] it is not the only reason, of course


Thanks for sharing them.


My argument is that without privacy we cannot build and defend other rights. Privacy allows to define private, social and public spaces with different rules. Without privacy your private stuff can be shared with the public. You could lose your jobs because you said something that your community does not approve. Privay in the modern world is like borders. Without borders we cannot really decide which laws to apply, because anybody could come in and go, so it would become impossible to enforce any kind of law.


Independently of the rest of your argument, the claim that open borders make it impossible to enforce any kind of law is falsified by, for example, U.S. states and municipalities, which enforce their own laws and have open borders with each other.


Is that more important than global warming? When you're standing in the center of a fire or you're a refugee because your city is underwater, it's also difficult to defend other rights.

My argument is that the headline is click bait to get eyeballs. Whatever, more power to you. Hope they sign up for your newsletter.


It is not really a big part of the thesis, rather an observation.

What you say about private communities is also true. However, I was referring more to the fact that everybody seem to nonchalantly use violent words and dangerous statements to express their disagreement, i.e., people are not just wrong, they are evil bastards that must be removed from public life and the nation itself.


Well, as I said the article is my opinion. So, let's agree to disagree.

In my experience real anarchy, in the way you describe, has never been achieved. So, when I hear that it reminds me when people say that real comunism is great, it is just that the URSS was bad, fake communism (it happens here in Italy, not sure if that happens where you are). The only historical anarchy that I know is the one where there is no control. Maybe we can do a better version in the future, but for now that it is what happened.

To be fair, I am Italian, so maybe I am overreacting on the whole barbarians thing.


USSR had democracy too: decisions were made per votes. Should we take that as a canonical example of democracy?

>The only historical anarchy that I know is the one where there is no control.

Do you want to say you can't live without Big Brother controlling you?


I am the original author. That are good points. I think that punishment would mostly comes through social interactions (i.e., we should chastise people) or technical means (i.e., you are banned from sharing content for X days), but also fines for the most egregious offenders. We fine people all the time.

> If I was flat out telling my family I didn't like race X and wished that religion Y could be outlawed, and then shared in the public realm that racists are terrible and that freedom of religion for all is paramount...shouldn't someone be able to call me out on it? And if not me, imagine a person of some power or influence: don't we want to know when someone's public political stance is different to how they really feel?

You can call out people in private for the bad stuff they share in private. However, you are right it is no so clear cut: we need to find balance. The issue is that making a special exception for powerful and influential people it is dangerous because it is not clear who are. In some way almost anybody has some authority: you might be a parent, the go-to guy for technical things in a community, the administrator of a forum or a open-source project, etc.


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

Search: