(this is page 31 of 121)





Latest

Eyes all over the place

Apple's announcement that it was jamming a video camera into the Nano reminded me of that overcrowded hall at CES where Asian vendors would line up to sell all sorts of weird electronics in the late 90s/early 2000s— from MP3 players that incorporated still cameras to external laptop batteries with built-in fans and mosquito zappers. The idea of a video camera in an iPod designed for exercising seemed... well, sort of random.

But then I thought back to Vernon Vinge's "The Peace War," a sci-fi novel where wireless video cameras the size and shape of burs are left in the forest for animals and straying people to pick up inadvertenly. The idea of a ubiquitous video camera seems foreign to us, but only because we've yet to find a good way for machines to take on the dull task of scanning video on our behalf, looking to highlight the interesting parts.

For instance: imagine using one of these new iPods to record a video of every run, from start to finish (with 640x480 video and 16GB of space, it seems quite possible). Then upon syncing to iTunes, you could get the much beefier processor on your laptop scanning all of the video while looking for interesting bits— where interesting is defined by recognizable objects or people (now possible thanks to interesting projects like Intel's OpenCV) that you might have encountered along the way, along with some time codes and, soon enough, GPS coordinates that might hint at something interesting happening.

Now imagine millions of such "runs" having their index files uploaded to some sort of Global Eye that might be able to perform crude searches when looking for patterns or movements of people or things (think of all of those "24" sequences where Chloe follows the terrorists on the traffic cams).

All of the sudden the inclusion of video capture modules becomes a little less random than say, garage door openers.

And you know what is more interesting? An iPod Nano is really nothing more than a glorified USB stick. How long until the cute little 2GB sticks you can buy for $10 while at checkout at Staples offer the same capabilities? Sure, you'd need a battery and a bit of processing power— but imagine being able to use the billions of USB sticks churned out every year to see.

It is true that privacy is completely screwed in the not-too-distant future. But I wonder what sorts of things we might gain in return when just about every gadget has one of the global brain's passive eyes attached to it?

Comments Permalink

Jacob Kaplan-Moss has a fantastic piece called "Snakes on the Web" about the future of web frameworks which provides a wonderful blueprint of what we need to build the next stage of the web.

As someone who made a big company-size bet on Django, I can say that almost everything he says about the current state of web frameworks, and Django especially, rings absolutely true.

The truth is that developing for the web has always been a "what kind of crap do I have to put up with," instead of "wow, look at this new shiny thing I can do." Having recently spent a vacation dabbling with the iPhone SDK, I was reminded of the latter— because despite their one vendor ecosystem and the AppStore annoyances, Apple does get developer joy when it comes to well-dcoumented and well-though-out frameworks that leave you (the developer) thinking there are some pretty big brains out there greasing the skids for the rest of us... for years to come.

Jakob is absolutely right when he says that we need those big brains on the web to think about framework interop and concurrency. The open web is going to win, be what may— so the only question we're going to have to ask ourselves going forward is whether we are looking at the equivalent of Windows and COM (but on the web), or whether we'll actually get it together and invent a new way to build the fabric of the web with better tooling. Craftman's tooling.

I sure hope so.

Comments Permalink

Everyone stops what they are doing to freak out about Gmail going down for two hours yesterday (it even becomes national news). To be honest, I missed it because I was in the middle of the ginormous multi-gigabyte upgrade to Snow Leopard, Apple's latest ode to local computing power. 64 bit. True multicore support. A new compiler pipeline. General purpose GPU programming. All sorts of things that you can only do if you've got cycles to burn on the client.

But back to the cloud woes: Cory Doctorow published a nice piece in The Guardian today that speaks the ugly truth about the cloud: that one of its foremost advantages is the business model benefit that it provides to vendors who get to turn a one-time purchase into an annuity.

As consumers we're just beginning to realize this given that we get incredible cloud-based services "for free" due to the advertising subsidy, but I think as the Internet matures, we will see more and more direct value capture on the part of the vendors who hold our data.

The Personal Computer revolution began from the desire to put computing power in the hands of The People. Reading Ted Nelson or Alan Kay on the topic makes one feel like it was a revolution that was about a lot more than just "collaborating on documents," or "accessing data from anywhere." Generally speaking it was about constantly inventing the future by predicting it with only one rule of thumb: almost anything can benefit from additional computing power.

For some things along than theme, the shift to computation on vendor-controlled servers in the cloud is great. But not for all, and we should keep that in mind— and not only on they days that Gmail burps!

Comments Permalink

AppStores are stupid fads

This is a really boring topic that is overblogged by and large, but here goes anyway. I was struck today by two pieces of news which make the death of the AppStores as a concept seem like a real possibility. The first was Larva Labs's excellent post on how poorly they've done with their game in the Android App Market. While most people are using this data as evidence that we need to keep putting up with Apple's BS when it comes to their walled garden, I think the real story is slightly different.

The second piece of data was the release of JQTouch, a jQuery plug-in that does a lot of the heavy lifting on behalf of a webapp so that it feels more native. I have not used it (yet), and it's probably not until the 3GS and 3.0 that you could really pull something like this off, but the promise of coding for the browser (albeit an HTML5-enhanced one) is alluring.

Taken together, here is a possible way to connect the dots: in classic "cargo cult product management," every big tech company and carrier now wants to have their own AppStore. All of us will fail. We will fail mainly because of our own app stores' siloed nature and because we will end up fracturing the potential audience. First to fail will be all of the Android-derived app stores where the vendor adds little value and fails to get Google's approval (which will be most of us). But all the rest will fail as well: Pre, Symbian, etc. None of them will be able to withstand the onslaught of the open web. In short, we are building today's Compuserves, Prodigies, and AOLs because it seems that walling up the apps gives us a new business model. Consider it the tech company equivalent of the newspaper industry's attempt to just now charge for content.

Meanwhile the Apple AppStore will continue with its impressive American Idol dynamic. That is, a few developers will become rock stars and cause everyone else to line up around the block waiting for their chance. But for most of them, it's not going to happen. Sure, a few mom and poppers will emerge and lots of "eff the man, eff the VCs!" blog posts will be written, but in the end, the Apple AppStore ecosystem of the next decade will look a lot like the VB ecosystem of the 1990s: a lot of small players enriching the ecosystem for Apple who will reap the bulk of the rents. The one exception will be startups, seeking distribution, who will put out free apps that do useful stuff as a way to get users to their sites. This is a good strategy.

In the meanwhile though, two things will happen. The open web will become stronger. And hopefully, everyone will learn about the power of side-loading, or letting users put whatever software they want on their devices. You can do this on Android today (assuming the vendor allows it) and you've been able to do it on the non-Apple platforms since the start. Then we'll be back to the same PC dynamic of ubiquity versus native integration and software developers will be able to make their own choices.

Generative platforms people, generative platforms! That is where the real Awesome happens.

Comments Permalink

I am getting tired of all of the economic prognosticators coming out to claim that we are "through the worst of it," that the US economy is seeing "signs of health," and that we should get ready to see a much better couple of years ahead. I get that it sells magazines and newspapers to make people feel good, but it seems like snake oil salesmanship at its best.

Now I may not be a financial whizz, but to me it seems obvious that to get out of this mess, we don't just need to de-leverage the economy, stop stupid spending, and stimulate specific sectors with government money.

To get to a good place again, we need to start inventing shit again.

And not just trivial twists on personal publishing, social interaction, and alternative e-commerce models. These things are fine for startups to try as they chase dollars that are moving from offline to online, but they won't create the 15-17 million new jobs we need in this country to get back on a healthy growth trajectory. They won't offset the fact that over the 130MM jobs in America today, only 20% of them pay more than $60,000 annually.

So says Adrian Slywotzky in a fantastic piece on big company R&D labs called "Where Have You Gone, Bell Labs?" that just came out in Businessweek. Slywotzky does a great job of laying out the troubling dynamic: big companies looking to meet shareholder expectations, are more likely to fund incremental stuff that can come out of the labs "pre-baked" to fit into some business unit's sales pipe than fundamental new research that might take a decade to bear fruit, if it ever does at all. Stuff like the transistor at Bell Labs. Or the myriad of innovations at Xerox PARC (most of which went to help create the tech giants of today). Or even, dare I say it, the thermal inkjet breakthroughs that have been a fundamental economic engine for growth at HP for the last 25 years (but which took 10 years in the lab to bear any fruit).

There is a companion piece in the issue about the new ways in which corporate R&D labs are adjusting which seems like complete spin to try to cover up the fact that this research is just not getting done anymore. Open, distributed, and collaborative will never beat long-term, focused, and committed work in some types of fundamental breakthroughs— and in fact, these hackneyed terms seem to be used defensively these days by big companies to hide the fact that basic R&D budgets are simply being cut with little more than rhetoric to replace them.

We need to fix this as shareholders of these companies, or we risk turning all of America's great tech companies into banana handling commodity providers. And to do that we need a new set of metrics to hold executive comp to that reflect long-term impact of R&D dollars spent as well as the belief in our guts that without this type of long-term thinking, we'll all ultimately end up paying far more than we have during this credit crisis.

Comments Permalink