Categories

5by5 | The Critical Path #14: The Super-platform ecosystem

The Critical Path show switches format for this week’s episode. I interview a guest.

5by5 | The Critical Path #14: The Super-platform ecosystem.

The idea is that rather than just telling stories and interpreting the world from my point of view, I should also ask others to teach from their experience.

In this case I discuss the disruptive potential of cloud computing with Randy Bias. Randy is the CTO of Cloudscaling, a designer of infrastructure cloud systems and a vast knowledge of the challenges and value of what we now call cloud systems. I’ve consulted for Randy in the past and I thought he could add a lot to the understanding of super-platforms. Super-platforms, you may recall, is a term I used to describe the coupling of Web Services and Devices in a mutually inter-dependent business model. I introduced the concept as a way to think about Siri and iCloud and potentially the Kindle and Amazon in general.

In this show I ask Randy to comment on the thesis that devices and cloud systems as a coupled system are a potential disruption to both the device-only model of computing and the cloud-only model of services. In other words, is the new ecosystem the coupling of devices with backend cloud APIs? We also touch on the scope of Apple’s data centers and the way we can calibrate the investment.

I think this is a ripe area of research and thinking about the future of computing. It will require learning a new way of measuring “performance” of solutions and the businesses built around it. It will take a multi-disciplinary approach to ferret out the key value propositions.

  • http://www.informationworkshop.org Mark Hernandez

    Great show, so much to think about and be fascinated with. I always listen to your shows twice and take notes. You know a show is good when listeners hear something that makes them think and this causes them to miss what is said next. This happens all the time at conferences, which is why it’s absolutely essential to offer conference videos afterwards, not only for those who couldn’t attend, but for those who did. But that’s another favorite topic for another time.

    And I would call your “interview” experiment a big success. It was very well-balanced. Add it to the plush fabric of what 5by5 offers us.

    Here’s a fun fact. I remember hearing on the British series “Cosmos” from 2007 (on Netflix) that the Large Hadron Collider can only operate in the summer months when there is extra capacity available from France’s electricity provider. Another source reports that there are about 22 days per year during the winter that the LHC cannot operate because the electrical supply is too tight.

    If you haven’t listened to the podcast it will sound like I’m daft. :-)

    Well done Horace.

    • http://cloudscaling.com/blog randybias

      Those are interesting stats re: the LHC.  Would love to have actual references if you can find them.

      • http://www.informationworkshop.org Mark Hernandez

        I read about it here…

        http://spectrum.ieee.org/computing/embedded-systems/powering-the-large-hadron-collider

        I don’t think its so much an issue with capacity but the fact that the energy costs are prohibitively high and it’s just too much for the budget.  Plus, that article is three years old.  

        Thirty years ago I worked on a diagnostic probe for the Tokamak fusion reactor experiment which is still hard at it to this day at General Atomics in San Diego.  Today it’s San Diego’s largest consumer of electricity.  (I happened to talk to a friend who’s also an employee at GA this morning, of all things, when I was fixing his crappy PC. :-)  They must store it up in big capacitors and then release it all at once or something, otherwise all the lights would go dim.  The fusion reactor here never runs for more than a couple seconds at a time.  I’m curious now to learn more about how the LHC does it.  Interesting.

        Thanks again for great insights Randy.

    • Marcos El Malo

      It would be nice if the audio player device allowed more precise positioning of the play head, maybe even scrubbing. Then you could just move the play head to the right place to repeat what you want to hear. I enjoyed listening, but found it very frustrating to search for snippets. Dan, if you read this, could you respond? (hopefully to tell us a fix is on the way!)

      • Anonymous

        Why don’t you try iTunes.  I downloaded it this week, and I think it works great for scrubbing.  If you haven’t used it, the easiest way that I found to link to 5BY5 was to click on Store – Home from the menu bar at the top.  Then click on Podcasts under the Apple icon.  Scroll down the right-hand margin until you find Featured Providers.  5BY5 is at the top.   When you’re playing the podcast, you can move the diamond marker in the iTunes display, and it’s easy to position. 

      • Marcos El Malo

        Thanks for the advice. I’ll try that next time!

      • kiran bhanushali

        If they use the Sound cloud player for the podcast that would fix this issue

  • Adrian Constantin

    I have not encountered very often a case in which the host and a guest would complement each other so well on the topic of the discussion. This episode is an excellent example of how the whole can be greater than the sum of the parts.

    • http://cloudscaling.com/blog randybias

      Thanks that’s kind.  I agree.  In this particular area, the combination of Horace’s understanding of the post-PC era and my expertise in building scale datacenter-based services is a great coupling.  There is a lot of territory to cover here in how these interact.

      • Kan

        Post PC – wait on what do we call these smarphones – powerful computers in our pocket. We are not in a Post PC era its just you showing your bias and repeating the Apple mantra. Post PC if PC means Personal Cognition.

      • http://cloudscaling.com/blog randybias

        I travel quite a bit. It seems pretty apparent that the traditional PC is not the vehicle for how the next 2B people on the planet are going to be connected. Yes, that’s my bias. It’s also the bias of all of the large carriers we have engaged with.

        If even the carriers get it, who are traditionally laggards during innovation cycles then it should be obvious to most.

        OLPC is not how the rest of the world is going to be wired. Android is.

      • Kan

        Post Desktop is a far more accurate description. Carriers will always show bias towards non desktop devices. But personal computing is far from dead in fact its miniaturization and falling costs is exactly what is going to bring about this next wave of connected individuals. 

        Desktop sales have been declining well before the Post PC moniker. Notebook and Ultrabook / Mac Book sales will continue to sell. Tablets and notebooks each have their uses and likely many people may have both. 

      • http://cloudscaling.com/blog randybias

        So your complaint is about the term and not the phenomena. That’s fine. I agree that post desktop is probably strictly more accurate. Unfortunately, we all know that accuracy doesn’t win in popular vernacular. Otherwise “hacker” wouldn’t have been hijacked.

        PC is a term that is widely associated with the Windows desktop. Hence most people understand “post-PC” relatively intuitively.

        I agree that personal computing is here to stay but it’s a far more personal experience. I don’t think the average consumer thinks of PC as “personal computer” instead associating it with the Windows desktop experience. Simultaneously, they think of their new computing experience on their phone as “using their phone.” ‘computing’ is moving to the background in the experience.

        Call it post-Desktop if you want. Until there is a generally accepted term I prefer post-PC.

        To most of the people whom this directly affects, these terms don’t matter and are just so much industry jargon.

      • Kan

        Yes they are jargon. I don’t disagree with the phenomena. It is quite apparent and has been so for a while. It is both a challenge and threat to Windows. Their response thus far has been typical – denial of the Iphone and now Siri. I really don’t understand MS at times-  they have vision, great technology but poor execution. They have imbeciles (it is not too strong a word) running the company who are more interested in quarterly targets and meeting Wall Street expectations.

        Nokia potentially had all the elements to succeed but again poor management destroyed that. Apple greatest achievement was to stick two fingers up at Wall Street and analysts and do what they think was in the best interests of their customers – profits lead from that not. 

      • http://www.asymco.com Horace Dediu

        I have a complete opposite point of view. Microsoft’s management is among the best in technology. They have excellent execution, mediocre technology and no vision.

      • Kan

        MS Surface concept, courier, kin integrated cloud storage (yes the product and execution were poor but the backend was not) are just 3 examples of technology and vision. They fall down on execution. Now what great technology advancements has Apple made compared to MS? Apple visions if you can call it that is integrated simplicity. 

        Excellent execution – what shambles was Kin, Vista.

  • http://twitter.com/petelehrman Peter Lehrman

    Interesting topic. I actually think that this was done on a vertical basis long ago by Bloomberg, who merged hardware with software to create the dominant information and data platform for traders trading various financial products. Kindle and Amazon, iPhone and iCloud, etc. are larger scale more open-ended versions of this, although I suppose they are new in their use of the cloud, as Bloomberg’s terminal and software systems weren’t always cloud-based.

  • Karsten Silz

    I think as an interview, the show was weak. It didn’t really seem like a question and answer session, more like to monologues that intertwined at certain points. An good interview needs short questions, meaningful answers and an interviewer who challenges the interviewee.

    An example for everything that was wrong in this episode: Horace defined super platforms and if the mixture of devices and services is a requirement going forward. Randy said “Isn’t that very exciting? An another point…”, and off he went. So: Question way too long, no real answer, no follow-up by Horace.

    It’s not that hard to challenge each other in this situation: If Google and Amazon and Facebook operate their cloud infrastructure by an order of magnitude better than everybody else, how can Apple compete then? Can they get by with just being less efficient, or are they set up for failure eventually? How can Microsoft compete there, missing both devices and online success? And is RIM too dead to even worry about services?

    So for future episodes, please read good interviews in a newspaper or watch them on TV and make your shows a dialog first and then a good interview.

    • http://cloudscaling.com/blog randybias

      All valid points.  Perhaps we can cover these in a follow up.  It’s really Horace’s call.

      Although it’s a bit after the fact, I’d like to answer a few of your questions.

      Apple and Microsoft can compete, but their success is not assured.  There are two factors at work: The first is actual ability to design and deliver cloud services like GOOG, AMZN, and FaceBook. The second is the ecosystem in play.

      Ability to compete efficiently is far from assured.  If I had to guess, even Apple does not do well in this regard.  It’s a hard problem to solve that requires you to have had success in running online services previously.  Here, neither Apple nor Microsoft has a great track record.

      As far as ecosystems, Apple clearly has a significant lead, having a wildly successful edge device platform (iOS) to plug into.  I think this makes up for a lot of failings in efficiency.  If your backend cloud services are 5-10x more expensive to deliver, but you can amortize the costs over a large number of devices and the total costs are small relative to your revenue, it’s not going to matter.

      The jury is still out here though and there is a fair bit of room for learning through failure (e.g. me.com).

      • Karsten Silz

        Thanks for the answer! 

        I think the jury’s still out on iCloud. At least it’s been declared the strategic initiative for the next ten years, like the digital hub back in 2000 – let’s hope they give it due attention.

  • http://www.facebook.com/people/Dave-Wright/100002464860605 Dave Wright

    I thought the format worked very well.  I’d agree that it really wasn’t an interview as much as a discussion prompted by a few setups from Horace.  There were some lengthy monologues by both Horace and Randy, but the content justified the length IMHO.

    I agree that cloud is one of the waves in the “perfect storm” we are on the leading edge of (I’ve been using the perfect storm language for about 4 months to refer to the convergence of Pervasive Mobile Access, Mobile Computing and Could Services).  One thing I haven’t seen much discussion of is Cloud Voice, which I think is another implication of these evolutions.  If the voice capable mobile device is the default user interface and it is constantly connected to a public or private WNet, why not outsource the voice services as well.  The truth is that it is just another client-server service running over IP.  Google Voice might be a move in this direction, along with some smaller startups.  SMBs would be the natural greenfield for such services, “Just use your mobile phones and we’ll provide a Cloud PBX solution for your business voice jobs.”

    From an enterprise networking perspective, if a lot of the data center will end up in the cloud and the enterprise edge will be overwhelmingly WLAN how much network infrastructure will really be necessary?  I’m not sure the data center outsourcing will happen in the near term, but the migration to a “wired-less” enterprise edge is underway.

  • Pingback: Two Disruptions for the Price of One | Cloudscaling()

  • Anonymous

    A few thoughts after this weeks episode:

    1) It helps to have someone a bit more versed in business concepts to guide and support Horace’s thoughts.  Dan occasionally seems a bit out of his element talking about business models and financial topics.

    2)  The whole topic of what a customer expects from a product, and how that continues to evolve over time reminds me of a simple marketing framework called “The Big Picture.”  One part of the framework involves the “Main” Variable, and the “Dynamic” Variable.  The main variable is the cost of entry into the market.  The dynamic variable is what sets apart your product from the competition.  In a disruptive environment, as Horace has pointed out, consumers continue to increase their expectations (once dynamic variables become main variables), until the product can accomplish everything they need it to.

    Apple has added App Stores, simple Media download, full internet browsing and other major features as “main” variables, and then has the Dynamic Variable that includes best in class hardware design, physical stores and now siri.  None of the other hardware vendors provide any dynamic variables, except maybe cost.  And the main variable is provided by Google and Amazon and other software companies.  So by definition, the hardware environment has to be a race to the bottom, because they don’t have a Dynamic variable to set them apart, and the “Main” variable value is being extracted by someone else.

  • poke

    I like the super-platform concept. Forgive me if this was mentioned in the podcast (I was multitasking) but I recall from the Forbes article on Dropbox that Steve Jobs told the CEO that Dropbox is a “feature not a product.” Maybe he would have said the same of the whole cloud.

  • Anonymous

    I thought the interview was great.  And Randy gave excellent examples of the super-platform.  If you haven’t heard it, one example was combining Siri with RunKeeper, an app that keeps track of your runs, how fast they are, the course you take, the calories burned, your heartbeat, etc.  If the app was synced with Siri, it could coach you in real time as if you had a real coach saying things like “Your pace is on track…” or “You’re slowing down…” and then even changing the music you’re listening to so it has slower/faster bpm to match your pace. (The example was at 40 min.)  I thought the interview really brought home how powerful this new platform can be, and I enjoyed listening to it.