Re:Apple’s research secrecy hurts industry and academic rese...
by _ph_ on Wednesday June 15, 2016 @05:27PM
(#11912271)
I am offended by this kind of headlines.
In the field of software, Apple uses open source and provides a lot of open source of their doing to the community. So, if there is an imbalance in the field of science, it should be discussed for sure. But if I am offended by Apple's Photos, it is for other reasons.
Edit: my comment refers to the top headline of the article, also used in the link from Hacker News when I wrote it "Why Apple’s Photos announcement should offend you"
Re:Apple’s research secrecy hurts industry and academic rese...
by patorjk on Wednesday June 15, 2016 @05:34PM
(#11912326)
It'd be nice if HackerNews put a star or something by headlines that have changed. I was really confused about what you were saying until I read your edit.
Re:Apple’s research secrecy hurts industry and academic rese...
by zepto on Wednesday June 15, 2016 @05:26PM
(#11912260)
Firstly, the research is paid for and is public domain - that's the whole point - so that people can use it without paying.
Secondly, the author betrays himself by dismissing Apple's open source efforts. Their contribution to WebKit is the only reason there is competition in the browser space, and saying Swift and LLVM are 'proprietary' and dismissing their impact is either ignorant or dishonest. As far as I can see LLVM is the backbone of a huge amount of academic programming language research.
Thirdly, Apple's photos 'AI' seems like a pretty modest application. Hardly groundbreaking stuff that they are keeping 'secret'. If they owe taxes for this, then all corporations owe taxes for ideas they've incorporated.
Re:Apple’s research secrecy hurts industry and academic rese...
by dragonwriter on Wednesday June 15, 2016 @05:35PM
(#11912331)
> Their contribution to WebKit is the only reason there is competition in the browser space
There have been competing closed-source browsers, plus Mozilla, that weren't dependent on WebKit. Without Apple's open-source work on WebKit, maybe Google wouldn't have done Chrome (or maybe Google would have forked Firefox as its starting point, instead), but there'd still be browser competition.
Re:Apple’s research secrecy hurts industry and academic rese...
by yanilkr on Wednesday June 15, 2016 @07:45PM
(#11912979)
Apple does things that are good for themselves and their users. The scope is narrow, they are not biased by existing ways of thinking and most of the time they deliver. Thats a very lean mindset. They are doing something right. I cant complain that they are not doing enough of something.
Re:Apple’s research secrecy hurts industry and academic rese...
by aub3bhat on Wednesday June 15, 2016 @05:30PM
(#11912293)
Consider the fact that Google open-sourced pioneering RAPPOR when it started using it for collecting information about malicious websites visited by the users on Chrome. Apple on the other hand is collecting far more invasive data about keystrokes while hiding the details of implementation. Why should we trust whoever apple trots out at its keynote? How is this different from people who trusted Theranos since its board was filled with former secretary of state and such.
Apples approach towards research and privacy is designed to make press headlines and fool reporters similar to Theranos. Its time to hold Apple to standards of Peer Review as practiced by Google, Facebook and others.
Re:Apple’s research secrecy hurts industry and academic rese...
by zepto on Wednesday June 15, 2016 @07:12PM
(#11912811)
Your claim that they are collecting 'invasive' data is unsupported by any evidence whatsoever, and yet you falsely stated it as fact, while comparing Apple with Theranos.
Re:Apple’s research secrecy hurts industry and academic rese...
by aub3bhat on Wednesday June 15, 2016 @09:46PM
(#11913440)
Just look at official Apple keynote video at 1 Hour 40 Minute mark. They clearly show Differential Privacy next to contextual typing suggestions and mention it. There is no magic its software and if Apple claims to have made some Differentially Private method to provide contextual typing suggestions they ought to open up the system for inspections so that users can verify amount of information being transferred. And this is not unprecedented since its exactly what Google did. Otherwise Apple should get off the high horse.
Can you imagine Apple claiming to implement a non NIST crypto and refusing to divulge details claiming security by obscurity. Everyone would cry bloody murder.
Apples behavior couldn't be further from theranos in this case. Apple is essentially selling kool-aid and you are drinking it.
Re:Apple’s research secrecy hurts industry and academic rese...
by zepto on Wednesday June 15, 2016 @11:31PM
(#11913734)
Sure but it is in no way comparable to what Apple is doing or claiming.
Apple makes some bold claims about how they respect user privacy and don't create profiles across their whole system.
Could there be bugs in that causing it to be imperfect? Sure.
Could they just be deceiving their customers? Very unlikely given the number of employees who would need to collide in such a fraud.
Asking Apple to open their systems to enable this to be audited is as reasonable as asking Google to open-source the search algorithm so it can be audited for political bias.
Are you calling for that too?
And just to be clear - Google doesn't make any claims about not creating a detailed identifiable profile on each user, incorporating all the data across all their services as they see fit.
If you are going to accuse Apple of lying, the burden of proof is on you.
If Apple like Theranos claims to have developed "Novel Differential Privacy algo" / "finger prick blood testing", it ought to have its claims examined via peer review. The burden of proof lies on the entity making ground breaking claims.
In the field of software, Apple uses open source and provides a lot of open source of their doing to the community. So, if there is an imbalance in the field of science, it should be discussed for sure. But if I am offended by Apple's Photos, it is for other reasons.
Edit: my comment refers to the top headline of the article, also used in the link from Hacker News when I wrote it "Why Apple’s Photos announcement should offend you"
It is frustrating to make a comment then get a response, "ugh, didn't you even read the article?" Uh, yes, I did, but it was a different article!
Secondly, the author betrays himself by dismissing Apple's open source efforts. Their contribution to WebKit is the only reason there is competition in the browser space, and saying Swift and LLVM are 'proprietary' and dismissing their impact is either ignorant or dishonest. As far as I can see LLVM is the backbone of a huge amount of academic programming language research.
Thirdly, Apple's photos 'AI' seems like a pretty modest application. Hardly groundbreaking stuff that they are keeping 'secret'. If they owe taxes for this, then all corporations owe taxes for ideas they've incorporated.
There have been competing closed-source browsers, plus Mozilla, that weren't dependent on WebKit. Without Apple's open-source work on WebKit, maybe Google wouldn't have done Chrome (or maybe Google would have forked Firefox as its starting point, instead), but there'd still be browser competition.
What a click-baity article.
Apples approach towards research and privacy is designed to make press headlines and fool reporters similar to Theranos. Its time to hold Apple to standards of Peer Review as practiced by Google, Facebook and others.
https://research.googleblog.com/2014/10/learning-statistics-...
Wait what?! Do you have any evidence of this or you just pulling this out of your butt?
https://www.engadget.com/2016/06/14/apple-differential-priva...
Good!
Hell given the amount of information people store on cloud, It might not even be a big deal.
But its not wrong to seek verification via peer review since they are making bold claims.
Can you imagine Apple claiming to implement a non NIST crypto and refusing to divulge details claiming security by obscurity. Everyone would cry bloody murder.
Apples behavior couldn't be further from theranos in this case. Apple is essentially selling kool-aid and you are drinking it.
Do you have any idea what kind of data google collects from Android users and stores as part of their profile?
I thought not.
Apple makes some bold claims about how they respect user privacy and don't create profiles across their whole system.
Could there be bugs in that causing it to be imperfect? Sure.
Could they just be deceiving their customers? Very unlikely given the number of employees who would need to collide in such a fraud.
Asking Apple to open their systems to enable this to be audited is as reasonable as asking Google to open-source the search algorithm so it can be audited for political bias.
Are you calling for that too?
And just to be clear - Google doesn't make any claims about not creating a detailed identifiable profile on each user, incorporating all the data across all their services as they see fit.
If you are going to accuse Apple of lying, the burden of proof is on you.
If Apple like Theranos claims to have developed "Novel Differential Privacy algo" / "finger prick blood testing", it ought to have its claims examined via peer review. The burden of proof lies on the entity making ground breaking claims.