Compute Neutrality | The New Battleground for Equality

Last year, the debate over Net Neutrality reached a fever pitch (at least as far as any obscure tech debate can be considered to be at a “fever pitch”). Proponents had good arguments. Opponents had good arguments. Both basically argued that the other side’s position was a threat to freedom, democracy, and equality for all. John Oliver briefly and hilariously became one of the most powerful players in the debate.

The government “fixed” it all by placing broadband providers under the regulatory domain of the FCC. After all, nothing says “freedom and democracy” like government regulation, right?

But, this week, that fix has been back in the spotlight with even more debates. At risk, they say? Why, the very future of freedom and democracy.

Meh.

Net Neutrality is but the latest in a long series of debates that will push our legal system, our economic models, and our society into new and strange territory. What’s coming next will just get weirder and more difficult than the policy makers can probably imagine – because eventually, instead of worrying about who has access to the highest-speed lines to link them to someone else’s data centers far away, we’ll be dealing with a much more serious issue: who controls the most powerful computing resources – the engine of the new economy – and how will they be shared for the benefit of all?

So, without going into the whole long history of computers, as I did before, or computing networks, as I also did before, suffice to say that the layers of abstraction that are being created to support cloud computing are making computers so flexible, interconnected, and powerful, that eventually the concept of “a computer,” or “an application” will seem quaint. Computing will be like water – spread over most of the earth, and often pooled together into vast oceans.

This has advantages – one day, you won’t need to learn a programming language to talk to a computer, you’ll just talk to it (and point, and gesticulate wildly, like you do when you kindly explain to strangers how to find their way back to the freeway). The machines will be so powerful – collectively –  they’ll be able to turn your abstract ideas into concrete computation.

Or they’ll become Skynet, in which case all bets are off…

But that’s not exactly what this is about. You’ve probably already tried talking to Siri or Amazon’s Echo – they’re pretty cool, but not quite the equal of HAL-9000 or the computer on the Starship Enterprise (and beware asking Siri about the AE-35 unit). They need more – and more flexible – horsepower.

A New Computing Model

So we’ve all come to grips (more or less) with the concept of “virtual machines” and “the cloud,” and we’re even starting to get the hang of Software Defined Networks (which are ultimately going to be the heart of whatever solution comes out of the Net Neutrality debate).

Now, we’re going to have to grapple with Software Defined Computing. This new type of computing structure is even more abstract and bizarre than the cloud. Instead of one homogeneous server farm belonging to Amazon or Google, Software Defined Compute (SDC) structures can intelligently break down computing tasks, and schedule those tasks on whatever resources are available.

It’s not a subtle distinction – in its ultimate form, SDC would be aware of the location, workload, and capabilities of all of the computing devices available (no matter what they are), and schedule really complex tasks across all of them to get tasks completed in the fastest, most efficient way possible for the consumer.

That sounds like an obscure area of interest, but it opens up a new possibility that some folks might begin to realize – the ability to work across computing resources owned by lots of different people.

So, Amazon could one day be able to surge in capacity (or get closer to the consumer) by instantly leasing compute resources across Google’s cloud (or Verizon’s, or anyone’s).

For that matter, they could spread tasks across unused (or underused) desktops – or even mobile devices. It’s a model that made Ross Perot quite rich a long time ago – but reimagined in a way such that any connected computer on Earth could take part in any computing task at any time, if it were found to be available at the right time, with the right capabilities, and with the right connection speed. It’s a structure that could unleash amazing potential.

But who pays, who profits, and who owns what?

SDC Economic Models

If you’re thinking about this from a societal point of view, you’d notice right off that big corporations own the vast majority of computing resources, and you might run over to Pew research to check out the demographics of digital device ownership to find out that there are distinct socio-economic disparities in ownership of high-end home computing devices like laptops and desktop PCs (though not as great a disparity in smartphones). You might worry about how control over the majority of free computing time will impact economic inequality when computing time itself will become the next great resource.

After all, you might think, isn’t that what Bitcoin mining is really teaching us?

They’re all good things to think about – if they even matter. Because while you’re busy signing endless (and unread) “Terms of Service” agreements, you might be signing away your rights to how your devices are used – as many Windows 10 users now suspect they have done – to get a free upgrade.

They say there’s no such thing as a free lunch, but we’ve been munching down for a long time without paying the bill. It started with free apps, free email, free cloud storage, and now free Operating Systems. Most of those “free” things are really just paid for with advertisement (including spying on you to target ads more effectively). Some companies have even embraced selling devices at or below costfor much the same reason – to be in a position to get you to buy things.

So, is it possible companies will give you free or reduced price software and hardware to gain the right to use your spare processing power (a clause probably buried in small print on page 76 of the Terms and Conditions)?

Maybe. Think in terms of mining Bitcoin – the spare processing power of every CPU and GPU in every laptop, desktop, and smart device in the world could be useful.

That’s not necessarily a bad thing. It can be a force for equality – making the services provided by that compute power so cheap that anyone can afford them.

But, in that scenario, it’s also possible (maybe even likely) that a handful of giant companies – or even a single company – could come to control the vast majority of processing power on the planet, either directly, or through “leasing” agreements – in much the same way that “Ma Bell” used to control the telephone lines.

That could be a cause for some concern. As more intelligent devices join the “internet of things,” and processing power gets faster and more efficient, it’s not hard to see how a handful of powerful companies could quickly control vastly more computing power than even the largest datacenter yet built. Think you’re worried about the NSA data center in Utah? Imagine a monopoly with the power to efficiently link hundreds or thousands of  so-called “mega datacenters” to millions of host processors around the world.

A company that controlled that kind of compute capability could exercise incredible power – in a far more subtle manner than the mere bandwidth throttling at the heart of the Net Neutrality debate.  It would be virtually impossible to determine if such a company were “cheating” – taking resources from projects (or users) they don’t like to assign them to ones they do like.

I’ve already discussed what high frequency traders can do to the markets by manipulating bandwidth and distance. A lack of “compute neutrality” could make that look like child’s play.

Preventing that kind of power from arising is neither likely nor desirable, as much good could come from it. Figuring out how to fit that kind of power into a democratic, egalitarian society is going to require an educated populace making intelligent decisions – not just people who know how to use technology, but people who know how to make it, and how it works.

…and that’s a topic for next time…

• O ex-Gerente Sênior do FBI, Líder e palestrante com mais de 19 anos de experiência de alta tecnologia nos setores público e privado. Estou muito interessada nas oportunidades no Brasil.
•  Segurança: Ciber (CISSP, SANS Mentor) e Físico (Contraterrorismo e Investigações Criminial).
• Construir times/de Ligação/Gestão: Gerenciado times diversos e complexos em situações estressantes.
• Desenvolvimento de Padrões da Indústria: membro fundador da Digital Display Working Group, e um dos autores que contribuiram para a especificação DVI.

Read more at CryptoID  INTERNATIONAL NEWS