Why I am Not All-in on Any One Platform

Yesterday I dashed off a comment on Facebook that started a pretty in-depth discussion:

Google gets my search history and email. That’s it. It’s already too much. They’re not getting my photos, my music collection, my physical location, what I say in my house, my phone data and my internet data. Thank god for at least some semblance of competition. Dropbox can have my photos. No one gets my music. Amazon can listen in in my house because they don’t know anything else about me other than what I buy. Apple I will trust the most and they get my phone. I may even stop resetting my ad identifier every week if they keep it up.
Oh and no one gets my location but Apple and Foursquare. Unless I open the app. Not even google.

My friend Buster pushed back — partially because he likes arguing and partially because he likes the internet more than me, I think. We had a great discussion in the comments. But he said I sounded like an old curmudgeon so I took it upon myself to lay out my concerns on the topic more clearly. I believe there are many, many valid reasons to not go all-in on any one ecosystem.

Here is my follow-up Facebook post, now featuring editing, bold and italic and in-line links because Medium is a better product than Facebook for long posts (please see reason #8 below).

— —

Having chatted with a few people yesterday on the topic (my wife, our house guest Terrence Curran, a few people who PMd me,my chief arguing buddy Buster Benson), I feel the need to clarify my statements yesterday. On Google/FB/Apple/etc.

  1. This is not primarily a privacy argument — though it is good operational security to not keep all your data in one place in the event of hacking (which happens, and happens to everyone), and yes, I do very much think that different companies have a different willingness to comply with the NSA. Do I personally worry what the NSA will do with my data? No. But I am very much against the practice and I believe in the market rewarding companies for their actions.
  2. In terms of data/privacy/advertising — I have always been a willing participant in that system. I am okay with it. I give a platform some amount of data, they give me massive utility (and the ability to search the entire internet in milliseconds IS magical, as is the ability to share my thoughts and see the thoughts of my friends around the world). That’s all great. Again, I am not a privacy freak. I made an agreement with a company to give them something, and I get something in return.
  3. That being said, I, personally, will in general prefer to give money for an ad-free product than data. Always. That is a personal choice, and one borne of privilege, but it’s a valid one. I am entirely comfortable with ad supported utilities. But personally I pay for my Spotify, I pay for ad-free Netflix, I pay for ad-free data storage, I pay for an ad-free phone, etc. I value options and I wish — I wish I wish — that Google and Facebook would offer something similar, even if almost no one took them up on it and the cost was constantly rising based on their ever-improved revenue per user. If for no other reason it would be hugely educational to people.
  4. While that value prop (certain amounts of my data for some utility) is cool, what I see happening with many Internet of Things (IoT) products and fully integrated product ecosystems is that the value proposition is changing. I am giving them MUCH MORE data and they are, theoretically, giving me something different/more in return. What they SEEM to be promising is improved functionality as my devices all talk to each other, *slightly* cheaper products than other companies and some vague AI promises. It is a new value proposition than before, and in my view it is lopsided. You may think differently, but it is my view that giving you, for example, my location and every spoken word in my house, is giving you a lot. And I should get something more for it. Other people may prize the option of giving of their data instead of money and be happy to get anything at all in return for it.
  5. A robust competitive environment is massively important to me, and I believe that Facebook, Apple, Amazon, and Google are getting too much control of the internet (maybe throw in Netflix and Microsoft and a few others as trying). But luckily, so far, no one of them has captured the whole environment and that is a good thing.
  6. Competition fosters product innovation. And options. Terrence pointed out he had previously embraced Google Photos because they had hi res backups for free and then they just killed it and said you had to pay. There are several stories about this. The companies killing off features, changing the deal, etc. Competition mitigates this and keeps other options in business.
  7. Data portability is vital to me. I am never again going to keep my data somewhere where I can’t get it back, or already have a copy of it on my personal computers, if the company goes out of business, kills the product, or changes the terms. And yes, the Googles and Facebooks of the world do this too. I use Dropbox because it has no ads and I can always keep all my files on as many personal computers as I went. If dropbox died tomorrow, I would still have all my data, I’d cancel my recurring payment, and go use another service. It would only be a minor inconvenience. If Apple died and I was using macOS Sierra’s new file management system, or their music cloud, I literally have no idea how I’d ever get those files back. Apple might not seem very likely to die, but I can very much imagine them giving up on their weird muddled music cloud product.
  8. Every product you own has utility. When you buy from many companies, you can pick the best product for you. That best product might come from one of many companies. Let’s call that UIP (utility of best product). There is also some measure of utility in having all your products from the same company. Lets call that UIP (utility of integrated products). I absolutely believe that it is very very rare, that UBP<UIP. I can think of no instance. It’s my experience that almost always UBP>UIP. This, I am starting to think, is a fundamental rule of capitalism. It was true in automobiles (car stereos, tires, mufflers, wipers, belts, even engines sometimes), hi fi stereos of the 80’s (god, remember those ridiculous proprietary interconnect cables that Sony and Kenwood etc had — did ANYONE USE THEM EVER?), musical instruments (MIDI was a much better standard than DCB), home gardening tools (we are slowly emerging from the period of interchangeable battery tyranny which, if you think about it, is kind of exactly the premise here. Buy a bunch of more mediocre tools for the convenience of being able to use the same batteries), kitchen appliances, etc. Will this trend continue into IoT? I think so. We may have common CONTROL platforms like Apple’s Home Kit, but the individual products will be made by several companies. This is a prediction but, really, it’s not my problem. If ever there comes a point where UIP>UBP, I’ll consider switching, for some part of the ecosystem. But we aren’t anywhere near it yet, so I’m not going to. I may buy one or two to play around (I own an Alexa and I am liking it), but I’m not going all in. I will be an early-experimenter, but not an early-adopter.
  9. I like my files. My files. Not Apple’s shitty replacement of them. Not their rez’d down audio instead of lossless, or their recompressed versions of my photos or their copy of Die Hard they think is the same as mine but mine was a rip of the German laser disc where the edits were slightly different and I liked that version better and their stupid algorithms thought that they were the same so suddenly “my” copy is “their” copy (and yes, this stuff DOES happen). I also like my play counts on my music files, my creation dates (when I ripped that CD to AAC MATTERS TO ME. It does).
  10. And I have a lot of them. It took Apple nearly five years to up their iTunes match limit from 25,000 to 100,000. I have 139,000 songs in my iTunes. Theoretically, now, I could use Apple’s matching technology to download THEIR copies of most of my songs, but… see #9 above. They wouldn’t be my copies, they wouldn’t be lossless, have my metadata, my tender, obsessive genre gardening, my album art or my play data (unless anyone can tell me that iTunes Match saves metadata and replaces it on the fly when they give you your files back in which case, awesome technical feat but probably not still the same codec so no thanks.) Right now, even though I didn’t ask them to, macOS sierra has spent a WEEK churning through and making a Photos library of my 140,000 photos. Adobe Lightroom did it in one night. When you have large libraries, these products are often not for you. You are an edge case that they assume you’ll go use a pro product.
  11. Those last two points were wonky but both reinforce #8 and really should be their own simpler point: the utility of an integrated product can actually be less than the utility of the best product, because the individual product (any individual product, really) can be so shitty compared to the best product that even for one product, the utility of buying the integrated version is negative.
  12. People keep telling me that all this AI makes for better recommendations. I don’t buy it. Thus far, I believe that digital product recommendations are of zero utility for me. Right now my Amazon page has: a bunch of Amazon-made products I don’t want, three rows of a bunch of household items I don’t want because I searched for, for example, a specific knife and added to my cart and now it’s showing me ten knives I don’t need. Its showing me a pretty decent book selection, actually, but then it shows five records I already own. Spotify’s best recommendation engine for me so far is their brand new release radar. Of the top 20 recommendations, I already have listened to or own 17 of them (there are many reissues it thinks are new), HATE 1 and will check out the other 2. Not great and nothing as good as my friends. I see this time and time again. Don’t even get me started on Netflix. (Here’s a good review of a couple of good books in the New Yorker on this topic). So any improved “utility” for me based on better recommendations leaves me deeply skeptical. The amazon example is arguably a reason to embrace one system: I buy the vast majority of my books on Amazon so it is better at recommending them than it is records. So I suppppoooosssseee there’s theoretically an opportunity for a company to listen to my office AND iPhone AND car across all platforms by constantly running an external microphone and processing all the data through a Shazaam-like service and developing a full picture of my musical tastes (since this would be the only way to capture vinyl) and then offering me recommendations but I am deeply skeptical owing to the historical limitations of Shazaam when it comes to obscure music, but YES, if someone made that product and they got recommendations good, I would buy it, but after a decade of disappointment, I’ll wait for the review from some friend of mine whose tastes are similar to mine.

Social media is annoying because we dash of half-formed thoughts and people misinterpret them and everyone should write full essays for every thought they have. Then no one will read them but if they do at least they won’t misunderstand you unless they want to.