Geller On Really Long Term Storage

In the current issue of Tom Geller writes about “The Forever Disc“–efforts to create an artifact that will be “readable” in thousands or even millions of years.

My own view is that this is yet another misguided (and egocentric) billionaires’ hobby, along the lines of cloning mastodons and so on.

Geller actually states the case succinctly,

“If we have information worth passing on, how can we better store it for future generations? Could it outlast even our species itself; if so, how can we make its meaning evident to its recipients?”

Obviously, “no”, we don’t have anything much worth “passing on”, and while many traces will outlast our species, they will not be intelligible to “recipients”–should they even recognize it as information.

And why should we even care?  Do we have anything worth saying?  I doubt it.

Geller does get one thing really right:  “long-term storage solutions must take nature into account, and work with it” because “Nature Always Wins”.

This is actually a good strategy even for relatively short time scales of years or decades.


Tom Geller, “The Forever Disc,.

IEEE Spectrum Confirms LA Air Traffic Control Crash Due to U2 Flight

Several reports claim that last week’s Air Traffic Control was caused by a military flight.

IEEE Spectrum–a reliable source over the years–confirms this, with a plausible story.  According to these reports, a U2 passing the area a high altitude was mistakenly treated as if it were much lower, flooding the system with conflicts to avoid.  It is also suggested that a communication channel that might have helped was down at the same time.

This is a pretty stupid error, but therefore all the more believable.

The IEEE Spectrum report is from what must be the easiest news blog ever to write:  IT Hiccups of the Week.  It’s chock full of goofs, crashes, and business disasters due to IT failures.

And lets be honest, most of this is due to software problems, because software is hard and expensive to do right.



Cawrey Replicates Our Semantic Analysis

A quick note to point to an item by Daniel Cawrey, “Word Analysis Shows How Different Bitcoin and Dogecoin Fans Really Are“.

Clearly, he has the right idea, as I have made similar observations earlier, here and here, and will probably produce some deeper analysis in the future.

I note that this analysis can be extended to other cryptocurrencies.  For instance, “national” coins are interesting.

I also note that you need look no farther than Very Much Wow, a Dogecoin Fanzine.  I haven’t seen any other cryptocurrency with anything vaguely like VMW.


Creepy Video Portrays Wearable AR

A short film by Eran May-raz and Daniel Lazo, titled “Sight” (2013?) does an interesting job of portraying Augmented Reality experiences.  Specifically, they portray wearable contact lens (or perhaps projected onto retina) AR.

The video does a decent job of showing the experience from both inside and outside, which gives us a really good idea of what the AR is doing in these scenarios.

The short video covers four generic ‘applications’ for this kind of AR.  There is a VR game, which is pretty conventional.  They portray a “kitchen assistant” that clearly portrays just how silly such an app is. I’m sorry, but slicing cucumbers does not require or benefit from a digital assistant, and “gamifing” it is stupid and wasteful.

The story then turns from silly to creepy-issimo.

The protagonist uses his AR system to prepare for a date, and even employs an artificial personality as a “wingman” on the date.  This isn’t even funny, it’s just sick.

By the way, this clown is paying more attention to his AR assistant than he is to his date.  Free advice:  pay attention to her, idiot!

It’s bad enough that the advice is stupid, even though based on creepy analysis of profiles and biometrics.  But it is covert, which is twisted and disgusting.  Anyone who pulled anything remotely similar to this would find a) I would call the police and my lawyers, and b) I would never stay in the same room with them, ever.

Aside:  the AR would have been useful to provide some VR based training, so this loser could go on a date using his own brain and senses.  Probably could do the same with the stupid kitchen application.

I’d say this video is a really good example of what not to do with AR.  Like I said, the video is well made and gets across the experience.  So it’s a valuable contribution–but not an example of good ideas!


Gueret et al. on Downscaling Linked Data

Christophe Gueret, Victor de Boer, and Stefan Schlobach published an interesting piece Let’s “Downscale” Linked Data” in IEEE Internet Computing .

Coming out of the linked data/RDF community, they point out the gap between the aspirations and rhetoric about public access to data (e.g., Shadbolt and O’Hara, “Linked Data in Government“, 2013) and the reality on the ground in most places.  The authors focus mostly on “developing” economies, but the same challenges are widespread in rural areas and poor neighborhoods even in rich countries.

What they mean by “downscaling” is to eschew large centralized servers, including remote “clouds”, focusing on handheld computing and local, low bandwidth networks.  A concomitant change in human interfaces is needed, to move away from text-based data entry and presentation, and to work well on tiny or non-existent screens.

As the authors suggest, linked data can actually make these adaptations better and easier, because the data models and reasoning are well suited for creating and using contextual information–which is critical to creating highly localized interactions.

In other words, this is absolutely not “dumbing down” linked data, it is “smarting up” mobile and handheld applications.

I say yes, please, may I have some more.


  1. Gueret, Christophe, Victor de Boer, and Stefan Schlobach, Let’s “Downscale” Linked Data. Internet Computing, IEEE, 18 (2):70-73,  2014.
  2. Shadbolt, N. and K. O’Hara, Linked Data in Government. Internet Computing, IEEE, 17 (4):72-77,  2013.

Short Note on Educoin, Yet Another Cryptocurrency

In an earlier post, I noted yet another cryptocurrency with its own narrative, Educoin.  The technology is pretty much identical to many other cyptocurrencies, but what is the community and surrounding narrative?

“Education” is a broad topic, and it is not obvious how a dedicated cryptocurrency is relevant.

The main exhibit is their reddit stream.

I haven’t had time to go through it in detail, but here are a few quick notes.

First, it is clear that there are several themes, loosely connected via the general idea of “crowd funding” education.

Part of the narrative comes out of a desire for more access to education, and dissatisfaction with the US student loan racket.  In this story, there are students who are unable to finance their studies, and we aim to identify them and fund them person-to-person.  Logically, this doesn’t require cryptocurrency at all, though it is as good a way as any to do it.

Hidden in this is the underlying assumption that conventional institutions are inadequate, and also that random crowds of strangers are a good way to identify and foster talent.  Both these assumptions are shaky, to say the least.

I also don’t see how you can fund a multi-billion dollar sector of the economy on “tips”.  If you want to get into the game of fixing education funding, I’d suggest a broader analysis of where education funding comes from and goes to. (Hint: human teachers need a living wage, paid in useful currency.)

A second theme is to tie cryptocurrency to MOOCs.  This is an interesting combination, micropayment for microteaching!  Pay only a little, and get only a little.  Was this the innovation we were waiting for?

Definitely a step backwards for education funding, but it may be all we have in a few years.

A third theme is basically marketing to college students.  In this story, “College students paying for stuff with EDU“.  Again, cryptocurrency isn’t essential for this activity (it has been a booming business for decades), and you don’t need a dedicated coin to do it anyway–you probably want Bitcoin or Dogecoin or something generic.

I’m sure there are other concepts floating around in the loose community of Educoin (which does not seem to have a cuddly nick name or mascot yet).

The good news is that, to the degree that any of these ideas turn out to be viable, they will work as well or better with other cryptocurrencies.

So I say, “Let a thousand coins flower.”  The best stories can be grafted onto the best coins in the future.

Cray on “Virtual”

You can’t read much these days without hearing about “virtual” something.  Virtual Community. Virtual Organization. Best of all, Virtual Reality.

I’ve been using this coinage for decades now, I know where it comes from.  But I still take the term with a grain of salt.  And frankly, it is usually very, very useful to invert it, to discover important issues.  (E.g., if “virtual organizations” are a good idea, then ask, “what is special about physical organizations?”  A very important question.)

My guide star is, of course, the great sensei Seymour Cray.  His computers did not use Virtual Memory, and folklore has it that when asked why, he quipped, “You can’t fake what you don’t got.”

Now, as it happens, an early part of my career involved proving Cray to be wrong on this particular point [1] (is it really more than 25 years ago, now?).  You can fake it.  Pretty much everything has VM nowadays (at least, anything larger than a phone….)

But it is a great starting point for analysis.

“Virtual Community”?  You can’t fake what you don’t got. (!)

“Virtual Organization”?  Again.

“Virtual Museum”.  “Virtual Wallet”.  “Virtual Business”. “Virtual University”.  Etc.


  1. McGrath, Robert E. and Perry A. Emrath, Using memory in the Cedar system, in Supercomputing, E.N. Houstis, T.S. Papatheodorou, and C.D. Polychronopoulos, Editors. Springer Berlin Heidelberg, 1988, 43-67.

A personal blog.

%d bloggers like this: