Google’s anti-Microsoft strategy continues to unfold. Today brings news that its online Docs applications will soon step beyond the Web and onto the desktop. (This is accomplished with the Google Gears browser plug-in, which lets you access internet applications while offline by using the hard drive as a cache - a company representative offered to “whitelist” me so I can start using it today, but the less privileged among you will have to wait until this feature becomes generally available over “the next few weeks.”)

Google likes to cloak its new product features in uplifting rhetoric: the company only looks to delight its users, it isn’t motivated by the sort of competitive strategy that other companies employ, and so on. But the evolution of Docs has always looked like a very deliberate plan hatched with its Redmond rival in mind.

Early on, CEO Eric Schmidt talked down the capabilities of Docs as a rival for Office: the main attraction was the ability to share documents, spreadsheets and other files over the Web, and anyway browser-based apps were very poor relations of their desktop cousins. 

Then, nearly a year ago, the tune changed. Having rounded out Docs into an Office-like suite of apps, Google said it was adding “applications” to its corporate mission statement (alongside search and advertising.) For good measure, Schmidt said that online apps were starting to become a real alternative to desktop software since browser technology had advanced far faster than he had expected (what a surprise!)

Extending Docs offline looks like the next step. Giving users the ability to write, edit or view files while not connected to the Web (any changes are automatically synchronised with the version on Google’s servers once the machine goes online again) removes one of the main disincentives for using Docs.

Google’s leaders have at times given tell-tale hints about the real strategy here. Last year Mr Schmidt conceded that while many companies might not yet consider adopting Google’s applications, they were still likely to use the threat of switching away from Office as a way to get a better deal out of Microsoft. Thanks to the new offline capabilities, this negotitating leverage is about to get stronger.

The maths for guessing the future of computer processing power is no longer “Think of a number and then double it.”

It should be getting easier in this multi-core world. We’ve gone from single brains to two brains and now quad-core microprocessors, so eight should be next, right?

Actually, the answer is three.

Advanced Micro Devices came out with the industry’s first triple-core chip today and it signals that cores are being viewed more flexibly than pure performance drivers. Instead, they are providing different price points and functionality for users.

“We’ve really started a debate in the industry with this,” Pat Moorhead, vice president of AMD’s advanced platform marketing, told me.

“Just like there’s a spot for $20,000, $30,000 and $40,000 cars, there’s room for two, three or four cores in the market. I see our competitor [Intel] now agrees – it’s announced a six-core server processor.”

Mr Moorhead says there are plenty of PC makers ready to support triple-core. It’s an in-between solution – for basic capabilities go for dual-core, for top performance, it’s quad-core, for something perhaps 30 per cent better than dual-core, triple-core will fill an important niche, he says.

The big shift is that Moore’s Law – the doubling of transistors on chips around every 18 months – is being de-emphasised by the two major processor makers in favour of “visual computing” – combining microprocessors with graphics processors to provide 3D interfaces and high-definition video.

“It’s going to become all about video, you get this incredible high-definition playback at 1080p with triple-core combined with our graphics,” says the marketing man, hammering home the message.

 So the fall-off in paid clicks on Google wasn’t a one-month phenomenon. The comScore report a month ago that the number of clicks on the search engine’s adverts had fallen slightly in January from a year before touched off fears that the Great Google Slowdown had set in. It didn’t matter that comScore itself later argued that quality improvements in Google’s ad system could account for the decline: the seeds of doubt had been sown.

The latest figures show that this was not an isolated phenomenon. The research firm now says that Google’s paid clicks in the US edged up by 3.1 per cent in February, which is at least better than the 0.3 per cent decline the month before. But this still represents a major deceleration from the 25 per cent increase in the fourth quarter of last year, and with search queries still growing strongly it points to a big change in the way searchers respond to adverts.

Only Google’s next quarterly earnings will reveal whether an advertising slowdown is setting in, but even the most rosy interpretation of events has to account for questions like these:

- Why is the click-through rate on adverts falling? It’s one thing for the number of adverts to decline - that shows that Google is pruning the least effective/ relevant. But shouldn’t that actually lead to an increase in the click-through rate? Intead, it fell 5 per cent in the latest month.

- How quickly does pricing in the Google ad market respond to quality  improvements? Higher quality ads should produce better leads, which should feed through into higher prices per click. But this won’t happen overnight. Effective as Google’s ad system is, this is not a perfect market. For now, the fall-off in clicks is the only verifiable fact (at least if two months of comScore numbers are to be believed.)

There’s an interesting new twist today to Amazon.com’s ambitious “computer-in-the-cloud” plan. From now on, companies which rely on the etailer to run aspects of their computing for them can choose where those tasks get handled.

This raises interesting legal implications. For instance, Amazon now says customers can select whether they want their computing to take place in a datacenter in the US or in Europe. So anyone concerned about the snooping eyes of Uncle Sam (think Patriot Act) might prefer to go off-shore.

And then there’s the tax angle. Will some companies try to avoid local sales tax by taking ecommerce offshore?

The most obvious purpose, of course, is to give users a greater sense of control - something so far lacking in the world of cloud computing. You can now select where you want back-up applications to reside. That might give some customers greater confidence they will be able to avoid the fall-out from failures like the one that hit a number of Web sites powered by Amazon last month. Making the cloud a bit less cloudy seems a smart move.

Interoperability between different social networks should not only empower users but also boost widget makers who can aggregate their audiences and increase revenues.
Mytopia, launching today as a “social gaming community”, is a case in point.

It allows users to come onto its network and play games with one another whether they are within Facebook, MySpace or Bebo. Users of Apple Dashboard Widgets, iGoogle Gadgets, Microsoft Vista Toolbar Widgets and Yahoo Widgets can also join in.

Casual games are the focus, with Sudoku, Chess, Backgammon, Bridge, Dominoes, Bingo, Texas Hold’em and Blackjack initially available.

“The idea was to create a digital playground that would appeal to the largest demographic - the very fragmented Web 2.0 world out there,” says Guy Ben-Artzi, chief executive of the Israeli company, which is rebasing itself in Silicon Valley.

He says the company has tinkered with Google’s Open Social and Facebook’s APIs to enable its network and it did a lot of work on the infrastructure so that everything could be managed from a central location. The grouping of the games in a sophisticated interface with chat features and an online shop represents a next step in the maturity of the widget industry, he says.

Mytopia will use social networking profiles to help target ads, although Mr Ben-Artzi says this is restricted to age, gender and location - such as a 35-year-old man in London receiving relevant ads. Users around a poker table are likely to see a different advertiser’s logo emblazoned on the green baize depending on their demographic, or they can pay $5 a month and the ads will disappear completely.

 Despite what you may read, Google and other internet companies didn’t really get everything they wanted from the US spectrum auction that just ended - in fact, in many ways the status quo remains exactly what it was.

The auction “winners” were announced today (that word is in inverted commas because only time will tell whether they got a good deal or overpaid.) Google shareholders were at least able to breathe a sigh of relief that it was Verizon Wireless, not the search company, that had paid $4.74bn for the all-important C Block of spectrum. That is the block that the FCC, after cajoling by Google, subjected to “open access” provisions.

So Google will get the benefit of an open national network and greater broadband competition without paying a cent?

Not really. Google’s preferred outcome all along was for this rare auction of national spectrum to give rise to a new national competitor in the US broadband market - someone to help bring down prices and make sure there is enough competition to keep the “pipes” open. With Verizon and AT&T dominating the auction proceedings, the outcome looks very much like business as usual.

Nor is it at all clear what the open accees provisions will actually mean in practice. Verizon held an event in New York earlier this week to lay out its plans for letting any device onto its network, provided certain safeguards are met. In theory, this means consumers will be able to buy a handset incorporating Google’s new Android software, then go independently to Verizon and ask to connect to the network.

Yet Verizon said nothing at all this week about how much it would cost to do this (last time I spoke to CEO Lowell McAdam about that, he said these customers were likely to be charged based on the amount of bandwidth they use - not something most people will jump at.) Also, anyone going this route would miss out on the hefty operator-funded handset subsidies that US consumers have been taught to expect.

The US wireless market may indeed be opening up, but it will be a slow process.

Any company hoping to launch targeted advertising services should be watching the fate of UK start-up Phorm with great interest. In particular, they should take note of what this says about the public’s double standards on privacy.

Phorm is trying to build a new ad platform, serving ads targeted around users’ internet habits and interests. It is hoping to make this acceptable to the general public with reassurances that no personally identifiable information is kept or stored as part of the process.

According to Phorm, the system will know it is serving an ad to a 30-35 year old male looking for a new car insurance deal. It will not know who you are, however.  You are just a random number. It will not even keep your IP address.

Phorm has consulted with every possible stakeholder to assure people the system is privacy-friendly - like the UK Home Office and the UK Information Commissioner - and it has had its privacy system audited by Ernst & Young and 80/20 Thinking, a privacy consultancy. It is inviting anyone with an interest to do their own inspection.

But none of this has really helped with public perception. There has been a blogosphere furore, and Phorm has been branded a spyware company in the press. A UK think tank this week sent an open letter to the Information Commissioner’s office, asserting that Phorm was possibly illegal.

As was seen in Facebook’s Beacon experiment, people are strongly against the idea of targeted advertising. Given any choice in the matter, it seems, they will campaign hard against it.

The attitude is, however, inconsistent with our tolerance for all kinds of other, less overt data collection and targeting. Where people are not explicitly told about targeting they are generally too lazy to protest.

Every Google search is stored for 18 months, complete with IP address and cookie information from a personal computer. There is much more of a profile kept on Google’s servers than on Phorm, yet, even after the issue was raised a year and a half ago by European privacy regulators as a problem, users have not abandoned the search engine in droves. It appears to be too convenient to boycott.

Millions of us carry store loyalty cards that allow supermarkets to closely profile our shopping habits. This is linked to our name and address – but that doesn’t bother any more than a handful of people.

In fact, we hand over our personal information constantly to any number of companies, from signing end-user licensing agreements to use software, to filling in forms to extend warranties on our household goods.

The companies to which we give this data use it for their own targeting – and are notoriously bad at protecting it. Several recent studies have shown that only a minority of companies have adequate data safeguards. Many don’t even know what data they have in their files and couldn’t say if any of it had leaked or been hacked. Big data losses such as the TJX incident are just the tip of the iceberg.

This is not causing major uproar.  However, if a company declares its intention to target us, albeit in as secure a way as possible, we feel outrage. Phorm is in danger of becoming a scapegoat for a general frustration about an information society we no longer feel in control of.

It is a shame, because the company was at least trying to move privacy technology forward to some extent. It may not have gone far enough, but it is a start. Stamping the business out before it has even started will not stop attempts to target advertising, but may simply drive it underground. The lesson from all this seems to be:  if you want to target, just don’t tell anyone you are doing it. They probably won’t notice.

No one should be nodding off when the Federal Communications Commission holds a hearing at Stanford University next month into “broadband network management practices.”

Comcast can expect to hear some strong criticisms of its “practices” from technologists and Web activists in the heart of Silicon Valley with the FCC just announcing that a second “field hearing” is needed on the subject.

At the first, at Harvard University on the East coast last month, the cable operator and internet service provider paid people on the street to take up seats, meaning many members of the public who might have wanted to have a say on its practices were kept out of the packed hall.

Comcast said the seat-fillers were holding places for their own staff arriving later, but a Harvard administrator said they remained in their places. Some even fell asleep (pictured) during the hearing.

Comcast is the subject of complaints from online video companies such as Vuze and activists Public Knowledge and Free Press of the SavetheInternet.com coalition. They claim it is blocking peer-to-peer traffic and limiting the delivery of rival video-on-demand services on its network.

Comcast says it is managing its network and P2P traffic may be less timely.

Free Press has welcomed the new hearing: “The threat posed by would-be gatekeepers like Comcast is very real and getting worse,” said Josh Silver, executive director.

“Open internet policies are urgently needed.”

 Michael Arrington of TechCrunch fame still dreams that blogging will eventually produce a media business capable of dwarfing even tech news giant CNET in size (that was also on his mind when I spoke to him 18 months ago.) These days, though, Arrington has a new plan: a “dream team” of star bloggers that would bring together the biggest talents - and egos - in the blogosphere.

His pitch today for this Uber-Bloggers’ Cooperative sounds a bit of a stretch. The most interesting thing about it, in fact, is what it says about Arrington’s thoughts on how to catch up with CNET: maybe it’s not as easy as just starting a handful of blogs and trying to raise hell after all.

(Incidentally, the post has a nice explanation in it of how the tech blogosphere really works: a loose  alliance of leading lights - Arrington calls them “the Gang” - with a mutual self-interest in cross-linking to each other to sustain their position. The commercial realities of building an audience, turning former allies into competitors, are starting to disrupt that cozy scene, Arrington says.)

News that Apple is in talks with record labels about a possible ‘all-you-can-eat’ model for iTunes is making the rounds on the blogs this morning following last night’s story in the FT.

Over at TechCrunch, Erick Schonfeld asks whether music companies would be willing to go along with a subscription model. We think the answer is clearly yes.

Consider this: Steve Jobs himself pointed out last year that the average iPod contains just 20 iTunes songs. Our own back-of-the envelope analysis shows that the number at this point (4bn iTunes songs sold, divided by 140m iPods and 4m or so iPhones) could be closer to 28 songs per Apple device. Either way, given Apple’s 70-30 revenue split with record labels, that means the music companies are making a paltry $14-$20 per iPod, even though many iPods can hold thousands of songs. Meanwhile, Apple banks hundreds of dollars per iPod sold.

Compare that $14-$20 with what record labels could make on an all-you-can eat deal, and it’s easy to see why music companies would be interested.  Music labels are thought to be pushing for a $100 up-front payment over the two-year life of an iPod, with the markup to be split between the music companies and Apple. The lowball $20 per iPod figure being floated by Apple suggests that Steve Jobs is determined to drive a hard bargain, if he agrees to such an arrangement at all. But even if the record labels were only able to get Apple to come up to $40 per device, they would be almost doubling what they have been getting under the current a-la-carte model.

 Update: Peter Kafka at Silicon Alley Insider has a good analysis of the economics of an iTunes subscription service here.