This article is a long winded view of the sophisticated behavior modeling ad-targeting networks that google and others employ - long on hand wringing and short on solutions.
Stratchery published a beautiful, irreverent proposal earlier this week to address this very challenge through a regulatory framework that actually feels like it could work and I encourage all HN users to absorb:
> platform providers that primarily monetize through advertising should be in their own category: as I noted above, because these platform providers separate monetization from content supply and consumption, there is no price or payment mechanism to incentivize them to be concerned with problematic content
The article does a good job of explaining why YouTube and Facebook don't currently do a good job of currating content, or whether that should even be their responsibility. It's not about machine learning, it's about price incentives.
Pretty rich, coming from the co-founder of a company that willfully handed over encryption keys of user data to authoritarian and brutal governments[1].
I would fully trust an encrypted communication system only if it is (1) end-to-end encrypted, (2) open source (or at least with source code available and buildable from source), and (3) based on a sound security design. Having said that, however, you're being very hard on Research In Motion (now called BlackBerry Ltd).
From everything I read and know about RIM, the enterprise level BlackBerry systems were unbreakable to governments and the keys were generated and controlled entirely by the customers (not by RIM). The pissed-off governments demanded access and threatened to ban RIM--the market leader at the time. I think the first to demand access was India and RIM put up a years-long fight against them before they capitulated.
It's easy to say that they should have taken a principled stand and lost the market. (In a similar vein, RIM had to pay a slimy patent troll $612.5 million dollars [not a typo, more than half a billion!] by a certain deadline otherwise the judge in the case would have banned them from the entire US market until they had a trial. The patents in question were ludicrously obvious and should never have been granted. I'd like to have seen RIM take a stand and fight the troll, but I can forgive them for having chosen not to go bankrupt.)
At the other extreme of corporate misconduct, are you aware that AT&T has been giving the call records (meta data) of every person in the United States to the NSA for decades? If Snowden's info is correct, they even allowed live tapping into phone calls for every phone call that passed through their network.
Furthermore, here's a quote from the article you linked to: "RIM, unlike rivals Nokia and Apple, operates its own network through secure servers located in Canada and other countries such as Britain." I have a high degree of respect for Apple (and somewhat for Nokia), but isn't it odd that RIM was being targeted by the host country but Nokia and Apple weren't? Perhaps they had a way to monitor communications (or at least get meta data) on Nokia and Apple phones, but they couldn't monitor RIM because RIM maintained its servers outside of the country.
In summary, cut some slack on RIM. RIM did use good encryption and did put up a fight. Many other companies have done and are doing much much worse.
Those are great criteria. But I would add that it should be P2P, with anonymized addresses. Such as Tox or Ring, where users run Tor onion services.
And if there must be central servers, they should also have anonymized addresses, and the owners and admins should be anonymous. Adversaries can't coerce people, if they can't identify or locate them. Even so, having central servers is a weakness to be avoided.
The issue here is not that other companies have worse security, we can rightfully assume thats true. The issue here is that by specifically releasing the keys to such a gov't they could have put peoples lives in danger that specifically relied on their lauded security and encryption.
So I agree with you fully, but I also do not have to cut them any slack as what they did was shitty.
From this story: "Facebook had been caught designing algorithms to identify stressed, overwhelmed, and anxious teenagers on its network, presumably to assist advertisers who might want to target them"
"But it wasn’t the looming disaster at Facebook that angered Ms. Sandberg. It was the social network’s security chief, Alex Stamos, who had informed company board members the day before that Facebook had yet to contain the Russian infestation……….She appeared to regard the admission as a betrayal.....“You threw us under the bus!” she yelled at Mr. Stamos………………Ms. Sandberg has overseen an aggressive lobbying campaign to combat Facebook’s critics, shift public anger toward rival companies and ward off damaging regulation. Facebook employed a Republican opposition-research firm to discredit activist protesters, in part by linking them to the liberal financier George Soros. It also tapped its business relationships, lobbying a Jewish civil rights group to cast some criticism of the company as anti-Semitic"
Whataboutism. Regardless, his point is that comprehensive surveillance of the Sidewalk project is bad.
> Sidewalk’s proposal is peppered with phrases like “comprehensive data collection,” “an enormous amount of data,” and “fine-grained data.” The data it desires runs from environmental (localized weather conditions, noise levels, and pollution) to social ( everyday actions that paint a detailed picture of what residents are doing and when ).
It looks like Google's vision for the modern city is to turn it into the Sims where Google can peer into everything. If you want to live in that world be my guest, but he's absolutely right. This is not something which is good in many ways. Where Google is monitoring everyone all the time. I mean even in Star Trek the computer isn't monitoring everyone all the time.... if it were then many plots could be resolved with the surveillance log on everyone's actions.
Not whataboutism at all; with historical context you might be able to see that this isn't a principal stand but possibly has some, yet unknown, commercial reason.
It's somewhat whataboutism because the whether the message is truthful does not rely on the source. i.e. Hitler could say genocide is bad, and he would still be correct even though he's Hitler.
It is relevant to the discussion to point out Jim Balsillie's relation to this topic. It's not a refutation of the point being made though. It's entirely possible that he has ulterior motives for making this argument that aren't based on it being morally (or event factually) correct, but it may be correct (or incorrect) separately from that.
> Not whataboutism at all; with historical context you might be able to see that this isn't a principal stand but possibly has some, yet unknown, commercial reason.
or experience and an authoritative source on the peril?
You'd willingly give all your keys and passwords to authoritarian and brutal governments if people without a sense of humor and with bulges under their suits' armpits showed on your doorstep, too. In fact, every single US-based global corporation did so at one time or another.
Hyperbole, no one is knocking on the door in the middle of the night with a stick, these are letters sent. In fact some organizations advertise the contact details for law enforcement requests.
Countries employee the metaphorical stick of comply or get band in this country. Some company has walked away from a country over this.
Eh, I think blackberry's destruction was of their own making. They had a huge head start but failed to embrace apps. The onboarding process for developers was painful and their devices were chronically underpowered.
They failed to read the winds of change that saw Bring Your Own Device come to the workplace and thought they were safe with enterprise and government users.
Looking back, I’m more and more convinced it was Verizon’s need for an iPhone clone that killed them. The book Losing the Signal: The Untold Story Behind the Extraordinary Rise and Spectacular Fall of Blackberry goes into a lot of detail about their decisions and that looked like the big problem. Something like the Passport earlier would have been a better bet than any Storm model.
>The onboarding process for developers was painful and their devices were chronically underpowered.
I can resonate with this. I worked for a major bank developing their blackberry app. It was amazingly painful to develop on their platform, its a literal mess.
Android wasn't that great technically at the time either, I was actively avoiding mobile app development at that time due to how crappy the SDK was, thankfully it's not that bad anymore, Google put a lot of work on Android since then but was Blackberry really worse?
It was night and day. Imagine Android dev with worse documentation in Java 5, (somehow!) more fragmentation, less hype so less stack overflow articles but with stricter than Apple submission hoops and no xcode to help you stumble through it.
Jim Balsillie actually tried to change course, open up, and monetize BBM. He was shot down hard by the rest of the board, who thought hardware was the way to go. I can't imagine how much that would sting, considering a couple years later Whatsapp would be sold for billions. Taken from the book mentioned in the article actually, Losing the Signal, before anyone asks for a source.
Stratchery published a beautiful, irreverent proposal earlier this week to address this very challenge through a regulatory framework that actually feels like it could work and I encourage all HN users to absorb:
https://stratechery.com/2019/a-regulatory-framework-for-the-...
reply