What are the top 10 things about cloud we'll all take for granted in 3 years?

1. Identity (AuthN / AuthZ) & Mobile Application (not Device) Management will be simple, configurable options on hardware platforms via accepted and adopted standards.

2. Security will be by policy, not by topology and there will be simple options for "securing the data, not the device" and hence allowing enterprises to live in the metropolitan area (cloud) and not behind the castle and moat.

3. Cloud Concierge providers will have blossomed, peaked and stabilized with leaders like Equinix providing CIOs with an option-rich, trusted "service aggregation" partner.

4. Cloud services will be measured on value provided rather than costs saved.

5. The argument around "DevOps" will be long since forgotten and it will be normal for "Infrastructure" operations people to be intimately familiar with Software (API) operations and for "Software" people to be capable of programmatically configuring infrastructure components.

SLAs (which are bad for customers because they only cover a small fraction of the true cost of security — including availability — incidents, and bad for providers because they strip them of what little income they earn from providing the service) will be a thing of the past... you don't need an SLA when you have a proven track record of delivering "dialtone" availability.

Enterprise infrastructure will shift off-site/third-party facilities (like Equinix) to better reflect the shifting "center of gravity" of the user-base. Thirty years ago centralised infrastructure made sense because users came to the office 9-5 to use tethered workstations — today they're Internet-based devices, roadwarriors, customers/suppliers/partners, web services, etc.

For application developers, networks will be about connectivity, with most application-specific cloud services delivered from the edge. However, it will all be an illusion^Wabstraction.

The abstraction, however, is delivered on rich, innovative networking platforms that enable things like optimal geographic distribution of application components, auditing of network traffic across wide area networks, and a variety of security services with minimal impact on application design.

Data in the cloud is more secure than in your terrestrial data center.

Sustainable Cloud Providers will number around just 10 - including the likes of Amazon, Google, SalesForce, MSFT, Apple

The word 'Cloud'

Once it's actually imbedded enough into social and business activities we'll stop using the buzzword to describe it.

1. Location. In 3 years the cloud wont be in a location. Few people care or know today where our major search engines run from. The clouds they operate in will be the cloud we also operate in. An undefined non jurisdictional URL?

2. Similar to Josh - As a Service - will no longer be a term.

3. Hopefully a common marketplace

4. Brands. The cloud will no longer be a cloud. It will be a service brand. We will take for granted the service provider.

"Open Cloud" will be advertised by providers and demanded by consumers, guaranteeing important freedoms for services in the same way that "Open Source" guarantees freedoms for products.

Semantic integration, monitoring, and management across the entire IT stack.

DR/COOP is a standard aspect of all business applications most running in active/active configuration.

Users will download pre-configured machines from vendors instead of installing software manually.

1) Security - this will no longer be the biggest objection about cloud solutions
2) Device agnostic mobility - this will be an embedded attribute of all SaaS applications
3) Big data - this will no longer be a problem only to be solved by on-premise apps
4) Highly transactional - the ability to deliver highly performant and transactional apps
5) Enterprise-ready - most enterprises will think cloud-first before on-premise for new apps
6) I would hope data residency, but this really depends as much on the political landscape as it does anything else.

Ubiquity
Reliability
Regulation and Taxation (esp. if the dems take a stronghold next year)
Dependability
Dependence
ADB (Any damn browser)
Stratification (Haves and have nots at various levels)
Traceability
Awesomeness (in spite of the good/bad mix)

I think in 10 years people will barely remember software upgrades. It'll be like 8 track tapes, you kind of remember them but can't imagine life without your tiny cellphone that doubles as an iPod, camera, browser, navigator, TV, game boy and email tool. People who talk about upgrades will sound like people who talk about punch cards, Neil Diamond and cars that didn't auto-lock.

Yes, they will laugh at us the way we laugh at our predecessors, and good for them! I hope you're right about the software upgrades, and I can't wait to see what additional items will be absorbed into the cell phones.

I agree with Sam. My take, Corp. datacenters will disappear.

1) reliability - in fact your average CIO (not even the worker) won't even know which applications and compute infrastrucutre are cloud delivered and which aren't
2) API's - in fact for the clouderati they already take these for granted, all those cool wonderful things you can do are a function of the API's
3) There will be different delivery models and we will have finally put to bed the debate about the 'true cloud'... its a horses for courses situation.
4) Cloud computing exists outside of the US too.

The word 'mobile' doesn't mean anything - networks are ubiquitous, all devices are connected - wherever, whenever.

People will stop comparing AWS to VMWare and OpenStack. Three distinct industries will emerge in the infrastructure clouds space:

1) vertically integrated cloud providers that couple hardware with orchestration and services - Amazon, Rackspace Cloud etc.Those will be like a Regus of cloud... not cost efficient at scale, but ideal for spot, bursty workloads and organizations that are growing very quickly

2) cloud orchestration software - huge software market with two camps - proprietary (VMWare) and open (OpenStack).

3) hosted cloud service providers - hosting companies offering managed cloud hosting. I.e. a BPO service in cloud management space, that will cater to the enterprises with predictable IT workloads (most of the large ones). Enterprises will define (or choose from templates) their stack - hardware to software (see 2 above) - managed cloud hosters will manage it's ongoing operation.

I'm going to be the skeptic and say that in 3 years most Enterprises will still be clinging on to there own hardware servers. If people can't move off Windows XP then how in 3 years are they going to make it to the cloud?

The issues that will slow this progress down is;
Network connectivity - not all countries have great coverage in regional areas.
Security - It will take a lot of convincing senior management that the data and apps will be secure.
People - The X&Y gen are still working from an office PC or laptop. It will take a huge cultural shift to trust employees to be more flexible in when and where they work.

IT Employment - Moving from an in-house IT team to outsourcing is a big move and mind shift for a lot of companies, as response times and knowledge of customized apps may be lost depending on how it is structured.

However I obviously need to move into a consultancy role as this will inevitably happen but not in the next 3 years!