ATOM Award of the Month, October 2017

For this month, the ATOM AotM goes to an area we have not visited yet.  Enterprise software and associated hardware technologies may appear boring at first, but there is currently a disruption in this area that is generating huge productivity gains.  

Amazon Web Services (AWS) is an ever growing list of services that replaces computing, storage, and networking expenditures at client companies.  At present, over 90 different services are available.  Here is a slideshow of the various companies and sectors being disrupted by AWS.  

loud computing itself is relatively new, but this revolution by Amazon as taken direct slices out of the existing businesses of Microsoft, IBM, and Oracle, which were slow to deploy cloud-based solutions since they wanted to extend the lives of their existing product lines.  Their anti-technology behavior deserves to be punished by the ATOM, and Amazon obliged.  AWS is set to register $14 Billion in revenue for 2017, most of which has replaced a greater sum of revenue at competing companies.  

The biggest value is the lower cost of entry to smaller companies from the on-demand flexibility enabled by AWS.  Now that IT Security and Compliance is far more cost-effective through AWS, the barrier to entry for smaller firms is lowered.  This is particularly useful for clients in far-flung locations, enabling a decentralization that facilitates greater technological progress.  Upgrades across computing, storage, software, networking, and security are disseminated seamlessly, and since far less hardware is used, the upgrade process is far more materially efficient.  This removes a variety of smaller bottlenecks to technological progress, mitigating the corporate equivalent of the Upgrade Paradox.         

Another great benefit is elasticity, where a company does not have to worry about estimating hardware capacity needs in the future, which can often lead to overbuying of rapidly deflating technologies, or underbuying, which can cause customer dissatisfaction due to slow speeds.  All of this can now be scaled dynamically through AWS.  

For the productivity gains inherent to the scale and dynamism of AWS, it receives the October 2017 ATOM AotM.  

Related ATOM Chapters :

3. Technological Disruption is Pervasive and Deepening.

 

Comments

[Geoman]

Good choice.

This market should have been cornered by IBM or oracle easily. But the decision to cannibalize your business to stay ahead of the ATOM curve is a difficult one. Certainly Sears, or Walmart could easily have become the next Amazon in the 1990s, but inexplicably chose not to. Sears is almost certainly dead and even walmart is in for a struggle.

One company I always admired, who did see the future, and did decide to destroy their old business model, was Netflix. Look at them now - they destroyed the old distribution model, but their streaming service and original productions are being imitated by everyone, including Amazon. If they hadn't had a head start they would be long gone. They have predicted they will completely destroy Hollywood as we know it within 10 years - and I believe them.

You can't stop the ATOM, you can only delay it briefly. You are much better off riding the wave wherever it takes you. The minute someone says, well, we could do this new thing, but it would undercut this other part of our business, they are dead meat. Because there is someone else out there without that hang-up who will dive right into that niche.

[fatcat]

Amazon is worthy of a few ATOM nominations. Kindle was if not the coup de grace that moved the printed publishing to irrelevance at least a major blow. Online ordering (and Amazon has a huge share there too) causes the closure of malls and local (non-online) stores. Automated warehouses and predictive delivery dispatches are worth paying attention too, though not ready for ATOM award quite yet.

Ironically Microsoft tries hard to reinvent itself. They are number 3 cloud provider and they even have put the crown jewels there. The new MS SQL can run on Linux, WIN 10 can integrate bash... And the pigs can fly... Let's see if they will be able to overcome the entrenched mindset and move to the future successfully.

Oracle seems entrenched in milking the existing customers. In short term that might bring more profit. But ironically the more money you extract from an industry the more incentives you create to fund the alternative. Here Kartik can define a "second law" of ATOM dynamics where each action of profiting from a condition creates a counteraction by attracting ATOM resources proportional to the budgets used by the incumbent technologies.

[Kartik Gada]

fatcat,

That is correct. Amazon has executed several layers of disruption. Among others, freeing up dead capital tied up in sparsely utilized retail land with huge parking lots.

Fulfillment by Amazon could be big some day as well.

Microsoft is the opposite. Their primary product is one that does not improve in quality. Windows and MS Office are not materially better than they were in 1995, even if Intel's chips have had 22 years of Moore's Law (30,000X gain) since then. Microsoft keeps moving buttons around in Office products and occupying more computing resources in Windows, but there is very little improvement in absolute terms.

[fatcat]

You are right that each new version of the MS office didn’t have a killer feature for which you absolute had to upgrade. The most compelling reason to upgrade was the network effect nuisance of receiving files with the newer and incompatible versions from your partners. Ironically, the biggest “innovation” was to dumb the interface down to make it easier to do simple tasks and harder to explore more advanced ones.
And yet the excel files are not limited to 64Rows and 255 columns and there are tons of nice features. There are businesses that routinely uses excel files of tens of megabytes as their workhorse LOB software. The word grammar checker is quite decent and a good target for AI. But most importantly, Microsoft is a provider of multiple framework and environments where the users and developers can create their software, which is much less restricted than the apple and more powerful than android. They are one for the main enablers of the ATOM revolution. In the last 10 years they have lost their mojo and even market share but there is a genuine effort to self reinvent. And again ironically, since they have lost the absolute dominance of the market they started playing nicer. And probably will innovate much better.

[Geoman]

By the way, I stumbled on a paper I find quite fascinating, and in fact, revolutionary.

http://research.nvidia.com/sites/default/files/pubs/2017-10_Progressive-Growing-of//karras2017gan-paper.pdf

The immediate impact - why hire models for any sort of ad? why even take any photos? Just create anything you want, immediately.

The future of Hollywood is simply tech companies - no actors, no studio, just a director, writer, and a staff of people at computers, who do increasing less work. Eventually, one person and a very fancy computer, making movies.

What if Netflix or Amazon can create, literally, a 100,000 hours of television per year, every year, at trivial cost? Any story anyone wants to tell can be written, created, and distributed in a couple of weeks.

Think of the enormous amount of capital that will be freed up.

[Kartik Gada]

Geoman,

Indeed. The video of this same research is here :

https://www.youtube.com/watch?v=XOxxPcy5Gr4

The amount of wealth and power concentrated among 'celebrities' is one of the most lopsided distortions in society today. No other field populates its upper echelons through sheer luck in this manner, and there is no greater example of such undeserved idol-worship.

The resource misallocaton represented by celebrities, their involvement in politics, and 'endorsements' comprising of billions of dollars a year divided among just 50 or so athletes, is ripe for ATOM disruption.

The ability to make movies cheaply has existed for a while. But the studios are still too strong of a cartel, which is why instances of low-budget films earning 200x their production costs is rarer than before, contrary to what one would expect.

It has always been a letdown for me that Bollywood, which can produce full feature films for $1M, and has no shortage of English-language talent, has made very few forays into marketing English-language films worldwide.

[A.M.]

"AWS is set to register $14 Billion in revenue for 2017, most of which has replaced a greater sum of revenue at competing companies."

Is the revenue elsewhere really decreasing more? Seems like tech spending is increasing so much that I find that implausible. Oracle's only lost half a billion in annual revenue over two years.

As for indy media like Bollywood, I think it fails simply because the established media succeeds through blanketing you with a narrow selection that breeds familiarity, then affinity. Good recommender systems might allay this somewhat, but it still hasn't really happened where you might expect it, like streaming audio. There's likely a social network element to it as well - you want to be listening to something your friends are, so obscure film/music recommendations won't cut it.

That video was utterly fascinating. I could see such tools ushering in a new age of design, since it will be so cheap and easy to come up with different, diverse designs. Conversely, if you keep an eye on new real estate development, it is far more bland and homogenous than what's shown.

[Kartik Gada]

A.M.,

Is the revenue elsewhere really decreasing more? Seems like tech spending is increasing so much that I find that implausible.

There is net general increase, even as older product lines give way to the same function for less revenue. Both properties are inherent to the ATOM. More bang for the buck, but still converging to ever-higher percentages of GDP.

That video was utterly fascinating.

When a perfect face is completely commoditized in photo and video, things could get interesting.

[fatcat]

Hi Geoman, Kartik,
this technique will definitely be useful but more for stunts and doubles or to replace a sick/deceased actor, or to enhance and beautify aging faces. One major immediate use would be for propaganda and fake news. In fact, with enough processing power and image recognition, one can replace any face in most of the videos. Could be useful also for historical and real persons, though the real people are ugly , so no market incentive. ...
But there will be no major reduction in movie budgets. Look at the cartoons, which still can extremely costly. An episode of futurama is estimated about 1M comparable to The Big Bang theory which started at 1M and ended at 9. What Hollywood delivers is eye candy with more or less predictable clichés. I guess bollywood is similar but specialises for a different market and population. That specialisation in a way limits the expansion to different markets.

[Kartik Gada]

The real threshold for AWS is when it becomes a viable consumer option. It would be great to pay just $10/month and have an ultra-thin hardware client, yet always have robust security, processing, memory, and storage on demand, and not have to worry about upgrades very often.

It would work well for a family of 4 for sure.

[fatcat]

Well, the level has to be defined not only on price but also on level of delivered features.

[Kartik Gada]

Well, the level has to be defined not only on price but also on level of delivered features.

Well, that goes without saying.

But we are long past the time where basic consumer computing/storage should be a utility. Among other things, viruses would be a lot fewer, and software makers would not have to deal with piracy.

Most importantly, the upgrade paradox will be mitigated. Hardware theft (which is still a problem in poorer countries) can be greatly reduced.
Laptops become more commoditized and lower cost.