Monday, January 7, 2013

crappiness and its acceptance

In a typical example of the breathless fawning that passes for technology journalism these days, Gizmodo's Brent Rose writes
Qualcomm's Snapdragon S4 Pro processor has been blowing us away with its speed in phones like the Nexus 4 and HTC's Droid DNA, and that's why we can't wait to meet its big brothers. The Snapdragon 800 boasts a quad-core Krait 400 CPU, that can go up to 2.3GHz per core.

To put that in perspective, the base-level Retina MacBook Pros come with a 2.3GHz processor. In other words, things are about to get very fast.
OK. Pardon me, but-- I knew better than to compare clock speeds across architectures and platforms when I was 12 years old. Understanding the limitation of clock speeds in these kinds of comparisons isn't some sort of arcane computer knowledge. It's basic information for anyone interested in computing. (For the specifics, Ars Technica provided a rundown a couple years ago.) I assure you, for a variety of tasks, the Retina Macbook Pro will significantly outperform the ARM processor. Yes, technological advancement in the mobile computing sphere has been impressive, but x86 and 64-bit processing has advanced quite a bit too. And indeed, a lot of the gains have been precisely in the realm of improving speed without improving clock speed, given the constraints of heat and conduction. And as quickly as technology can move, it also faces certain physical constraints, constraints that are particularly troubling in the confined spaces of mobile gadgetry. More processing power means more heat and more power consumption. The former is a wall; the latter just the reason you can't play a game of Snood on a modern cellphone without running your battery under 50%.

Obviously, if I overheard someone making this comparison, I wouldn't get snotty about it. But this situation isn't the same as overhearing an amateur. Gizmodo is a professional blog staffed by professional writers. Rose was presumably paid for his post. I guess it just drives me crazy that the professional Internet is still filled with so much casual, basic incompetence. I'm reminded of a Jalopnik post where the blogger wrote derisively of electric cars, as those cars have to be charged with electricity that is produced using carbon-based fuels-- which would be a relevant criticism, were it not for the fact that a standard electrical grid is far more efficient than an internal combustion engine. Sure, these are just a couple examples, but I'm constantly amazed at how many elementary mistakes there are in professional writing online. I particularly cringe when I read a lot of writing on education, my own field, because there is so much written that demonstrates a lack of understanding of basic social science principles or statistical innumeracy.

To make matters worse, Rose includes the Platonic ideal of contemporary tech journalism: "Not to gush, but seriously, phones are so awesome right now." I literally facepalmed when I read that. This is exactly what's wrong with the Internet: fanboy gushing that overwhelms the ability to be discriminating, rational, and adult. All common sense and responsible thinking gets thrown out the window for the latest round of  wheeeeeeeeeeeeeeeeeeee! Whether it's cell phones or superhero movies or the latest Dynamic, Proactive, Market-Driven, Digital, Innovative Education Revolution For The 21st Century, nobody lets the facts get in the way of some gee-whiz story about how everything is going to be endlessly badass real soon. And the pros are often worse than the amateurs.

When blogs really began taking off, there was a hope that they'd combine the best parts of traditional journalism with the best parts of the Internet, high-quality fact finding and reporting with the openness and egalitarianism of the Web. I would argue we've seen exactly the opposite since then: so many of the systems of gatekeeping and insiderism of the old media world have been replaced by equivalents that are just as rigid, and so much of what's published today lacks the most basic accuracy or fact checking. What's galling isn't just these mistakes are made, but just how little people care that they are made. Minor factual mistakes are pointed out by commenters all the time without any correction or review. Because the world of Internet commentary writ large is flagrantly hostile to the academy, its culture is antagonistic to the schooling that could teach these writers not to make such elementary mistakes. And as time goes on and a general level of crappiness becomes the norm, the quality falls further and further. What we have online is not merely a culture of low expectations but a culture of studied apathy towards ubiquitous crappiness.

7 comments:

Charles said...

On the other hand, the entertainment value of spam seems to be increasing.

Freddie said...

Ha.

I may need to set up some sort of authentication stuff again... hate to do it though

Corey said...

I think it's strange you assume that the goal in the professional blogosphere is accuracy. It isn't.

Brian M said...

Freddie:

I don't think it is quite true that driving electrical cars is more efficient or even as efficient from an energy standpoint as an using an internal combustion engine. Especially when one factors in the sunk energy costs of producing (and replacing)the batteries.

Unless one is talking about nuclear power, "green" energy is not a high percentage of the power supply, either. Coal and and natural gas.

I suggest browsing over at Do The Math w/r/t the false hope of "green energy"

http://physics.ucsd.edu/do-the-math/

Nope...the only hope is a 90% die-off and a return to hunter gatherer levels of subsistance (just kidding!)

And

Freddie said...

And... what? I can't take the suspense!

But thanks very much for the link. I was wrong to frame it as an efficiency question. The real issue was a net carbon deal.

Charles said...

On a more serious note, now that I'm at a physical keyboard:

I'm inclined to attribute this to tech-journalism mediocrity as opposed to general internet/bloggy mediocrity. If I'm remembering things correctly, the hobbyist monthly mags of the late 80s and early 90s were full of this kind of garbage, well before tech journalism migrated online. Every issue was padded with an article about why your Apple was superior to your neighbor's IBM clone (or vice versa), and another about how the next (hardware model/OS upgrade/favorite application) would blow your mind. I think that shit shaped the culture of tech journalism.

I don't find this odd. The people reading those magazines were enthusiastic users, and so were the people writing them. They had no deep knowledge of hardware systems, operating system design, application software, or anything else really. And look, most people don't know that you can't compare clock speeds across architectures, because they don't actually know what clock speed is (other than that a larger number means more speed) or even what a processor architecture is. Code branch prediction? Multiple pipelines? Good luck.

A tangentially related anecdote: When Apple first announced that it was dropping PPC and moving to Intel-manufactured i86 chips, I happened to be working for a serious Mac-head. I don't really understand how he got a job managing a group of sysadmins at a medium-sized university, but that's the IT industry for you. About two weeks after the announcement, I heard him explaining to a co-worker that PPC had been stagnating and that really, Intel had come such a long way in the past couple years. Two weeks earlier he'd have told you that Pentium chips were shat out by rats in Intel's basement. That's your audience for most tech journalism. (I feel obliged to note that anonymous evil boss was in most respects a truly lovely human being.)

zmil said...

Do the Math actually has a post on the electric car question:
http://physics.ucsd.edu/do-the-math/2011/08/mpg-for-electric-cars/

He comes to the conclusion that it's about a wash between fossil fuel derived electricity and direct propulsion by said fossil fuels. There is some interesting and informed sounding pushback in the comments, though.