Original URL: http://www.theregister.co.uk/2005/05/07/glenn_edens_profile/

Sun's newest star lauds the PT Barnum way

Gadgets, sensors and wealth

By Andrew Orlowski

Posted in Software, 7th May 2005 08:16 GMT

Profile "The web is now nature," says Glenn Edens, one of Sun Microsystems' most important executives, and its fastest-rising star.

The senior VP has been Director of Sun Labs for around 18 months, but the bright lights of Hollywood now beckon. He's been picked to head Sun's newest creation, a vertical business unit aimed at converged media, entertainment and broadband, which was announced with a flourish in Las Vegas at the recent National Association of Broadcasters (NAB) show.

The new unit has inspired great excitement and anxiety at Sun. It may not be overstating the case to say that it represents a battle for the soul of the company.

In a wide-ranging interview last week Edens enthused about sensors and "digital ectoplasm", the value of amateurs, and the inspirational example set by ... er, PT Barnum. You'll begin to see why his new media unit is such a contentious move.

But he parried questions about the new unit, which had already got off to a rocky start, we discovered.

"The NAB press machine got a little ahead of itself, and we probably got a little ahead of ourselves on that," he told us. "We hadn't even done all the internal stuff yet."

One of the immediate concerns is that Sun Labs traditional research focus, devoted to solving deep problems with today's computing infrastructure, will lose out given the new focus on media and "sensors".

"I do have R&D funds, and we are slicing and dicing budgets now," said Edens, a little ominously.

Glenn Edens - Sensor Master

On the face of it, the thinking behind Sun's media unit is pretty 1980s and naff. Converged media is nothing if not a cliche already, and SGI has long trumpeted such a strategy with a conspicuous lack of success. Why should Sun, at this stage, fare different? However, Edens' himself makes the pitch for a group devoted to media and broadband sound fairly plausible.

"Satellite, cable and mobile operators are all getting into each others business, and as everything becomes digital and IP, you're not going to be able to tell the difference. The telcos are getting into cable, the cable operators are getting into VoIP; and they're all trying to use the attributes of their network to get as many customers as they can for their services," he says.

In which case, he reckons, there are a lot a technology-savvy company like Sun can do to help them.

As well as presenting the big picture, Edens is also keen to the nuances. A Citibank requires very different service offerings to an HBO, he explained. He's isn't impressed by some of the hype around the converged business, either.

"Frankly the whole VOIP business is amusing, because you can get fixed rate unlimited long distance for $20 a month from all the carriers, which is less than what VoIP costs," he observes.

But what worries some Sun veterans is that how the new focus affects the relationship between R&D and wealth creation. Research may takes years to pay off, but if a company can claim to solve one or two of the industry's many difficult problems, then it has something of real commercial value to pitch to its customers, and a genuine competitive advantage. With the US' chief economic rival now determined to innovate rather than imitate, this takes on a new urgency.

Isn't the next Google inside Sun Labs, we asked Glenn? In projects like Celeste, asynchronous computing, and the HPCS proximity communication, Sun has work with multi-billion dollar potential. However the sensor free-for-all we saw at the Labs Open Day seems much more emblematic of the new Labs.

But Edens sees the new emphasis as returning Sun Labs to its original goals.

"The Lab had become somewhat disconnected from the business," he says. "The Labs original mission statement which Bert Sutherland had helped draft - which was to solve difficult problems that our customers had brought to us - the Labs had got away from that."

Sensorama

But while the underlying business case for a focus on media and sensors is sound, the some of the approaches are sure to raise eyebrows. Edens enthused about letting art and science collide - with the emphasis on collision.

For example, the Labs Director enthused about how the sensor platform, just a fortnight old, had allowed non-programmers to develop product demos.

"We just got the sensor boards back, and the for first time the sensors were accessible to the software team members and they were in the labs hooking up little motors and doing little things, and it was really great. We had literally tripled the number of people in the lab who could do a sensor project.

"We had David Simmons put together a refrigerator demo, he's not a hardware or software engineer, and he did it!"

That's an internet enabled fridge - one of the forgotten icons of the dot.com bubble.

The idea behind the internet enabled fridge was that you could tell, without opening the fridge door, whether you needed to buy more milk. The fridge would cybernetically order milk on your behalf, perhaps delivering it on the back of a Java-enabled sensor-bearing robot. Maybe one that looked like a miniature Gary Numan. But even a half-wit is able to open their fridge and realize "I've run out of milk!", and so the idea deservedly fell to earth.

But for a while the iconic internet-enabled fridge was an emblem of a great computer R&D experiment that failed: MIT Media Labs. So since he'd raised the subject, we discussed the legacy of MIT Media Labs' decade of producing interactive fridge demos. What lasting contribution had that made to the field of human inquiry?

"The MIT work - I didn't even judge it. My question is what do those students know when they come out - because they know something. They've ve been able to experiment and express ideas."

Another researcher, in a gentle rebuttal to our Labs report last week, hoped that, "Ideally we'd like to see college and high school students using Java to develop wild new wireless and embedded systems."

In place of rigorous exploration designed to build a residual body of expertise, which can then be applied to the company's commercial advantage, the new Sun Labs does rather sound like a hippy high school's after-hours science club, only with the teacher absent.

That's not fair, Edens insisted. Look how the web only took off after artists got involved. (After some prodding by your reporter, Edens said he was using the term "artist" synonymously with "designer"). We had text, then we had fonts, and finally web man ascended to the heights of Macromedia Flash.

"The invention of Flash came about because you wanted a vehicle for a non-programming person to program," he said.

Um, but don't people hate all that, though? We do.

"Well, some people do, some people don't - that's the beauty of society," he replied rather crisply, before confessing, "I sure hit that Skip Intro button."

So much for amateur programmers, then, you might well wonder.

Edens strongly defended the have-a-go heroics of his new "amateur" approach.

If Sun Labs looks more like a product shop, then that's no accident. The examples Edens cites showcase real commercial services that Sun customers can pick up and use, such as offerings designed to show fixed line operators how they can do more with their existing infrastructure. Some of these are quite plausible.

"In OfficeCentral we're showing a lot of new things you can do with conferencing. Teleconferencing is a multi billion dollar business, we've been showing a secure, stereo, multichannel audio conferencing demo. Stereo telephony totally changes the nature of conferencing calls."

The OfficeCentral demo has years of deep thinking on human interaction, trust and identity informing it. But how representative it is of the newer generation of projects is debatable. In the rush to produce a demo, the knowledge gained may be lost, or never even noticed.

Art for Art's Sake

While several researchers have been "helped to find new careers" during Edens tenure, we were delighted to hear of an new appointment.

Sun will appoint an Artist-in-Residence, for the very first time. That's because, Glenn explains, "two things inform technology over the years: art and the military."

"The military has driven the internet, through the development of semiconductors, and artists have driven the web," he says. "It's a combination of the two."

By this logic, Sun's next hire should be a "Soldier-in-Residence" - a quite tantalizing thought. Just as the Artist-in-Residence could sculpt inspirational artwork or fashion giant murals to provide the engineers with a muse, so a soldier could be hired to run around the campus bayonetting bags of straw, or if he's really resourceful, wiring up the gardener to the electricity supply.

But rather than reaching back to a renaissance man such as Leonardo, we were astonished to hear Glenn summon a more recent inventor for inspiration.

"It wasn't PT Barnum who said you 'never lost money underestimating the taste of America' - it was someone else, and it's attributed to him," said Glenn. "But the neat part of all this technology is that society sorts out over the long haul what's important."

For a contemporary PT Barnum, we mused, one need look no further than the MIT Media Labs' former chief Nicholas Negroponte. The Labs' founder cheerfully glad-handed corporate sponsors for years, and he even thought horizontally, creating a kind of gadget supplement - Wired magazine - to showcase this work.

Today there is no shortage of media outlets eager to seize on gadgets Wired is still with us, and both USA Today and National Public Radio can be relied on to fill editorial space with news of the latest wonders. But in the end only one of these two hit upon a reliable cash cow - and it wasn't Negroponte.

Empire of the Sensors

So which way will Glenn Edens' take the good ship Sun?

At times, his analysis sounds so rigorous, exacting and succinct you can't help but agree. MP3 has been a disaster for audiophiles, he says, but he optimistically believes that this will be cured when the networks have more bandwidth. Interactive TV will always be doomed because people want to be told a story - "film making works because people want to be told a story". Digital TV has made channel flipping impossible, because the systems aren't fast enough.

"We need to get control of the spam, get QoS nailed, and get the whole thing converted to IP. That's for the rest 3 to 4 years. It took 50 years to get the POTS to built up to be the platform for the growth of the nation. We're not there with the internet."

Quite so.

Then again, there were times when it sounded like the captain needed better navigation. When we raised the prospects for the web's survival, Glenn was horrified. It's a rational calculation that because of the failure to find a consensus to fix its many technical inadequacies, and commercial pressures from vested interests, we suggested that the web as it is may only have a slim chance of surviving. But Sun's Media VP looked as if we'd invented a new and horrifying blasphemy.

The problem with Edens' new Media Unit is that he must kill the internet he cherishes in order for it to succeed. If there's one thing Sun's biggest "media" customers such as Vodafone and DoCoMo understand, it's that they understand what sells, and preserving an "open internet" isn't exactly their top priority. Something has to give.

And once or twice he sounded like a man who's rowed so far out to sea, that even the Coast Guard can't help.

Edens urged us to look at one example of the web's wonders, a graphical interface that uses Google as its data source, and plots all the occurrences of numbers between one and a million.

"It's a beautiful display that allows you to scroll through and see a 2D representation of all the numbers ... and it's just astonishing what you can learn."

"You can just scroll through this, and see a frequency of an occurrence of a number. It's black or white with all these green in the middle. So you go to one of these white hotspots in the middle of nowhere and go, 'What's that number about? Why does the that number occur so often in nature?

"Because the web is now kinda nature." ®

Related stories

Inside Sun Labs - the best and the 'bots
Microsoft demos Jetsons kitchen