Microsoft Research Fights Critics

coondoggie writes to tell us Network World is taking a look at why Microsoft Research has to fight so hard against critics. From the article: "When the word 'innovation' is tossed about many may look down their nose at the company sitting on top of the high-tech industry — Microsoft. [...] Microsoft Research incubates not only futuristic ideas but young minds, having hired 700 interns worldwide this year including 250 computer science PhD candidates in Redmond alone, which is roughly 21% of all the computer science PhD candidates in the United States."

Comments

deservedly

[yagu]

If Microsoft were less predatory and less a bully in business maybe the rest of the world would stop looking down their noses at Microsoft's "research". As it is, it looks less like research and more like unfettered spending to find "yet another" way to dominate.

I welcome research from any company. I'm guessing I've probably used what amounts to "innovation" from Microsoft, derivative of work from their labs.

Unfortunately for Microsoft (but true to their character) they have tools for mouthpieces like Ballmer. Microsoft inks a deal in what could only be viewed with raised eyebrows, and Ballmer punctuates that with "they're infringing our IP anyway...". As long as Microsoft continues to be so hostile to the world in general, they get what they sow.

Their research may be golden, but it's ill-gotten gains, the world thinks so, and the world is probably right. The fact that Microsoft has such a corner on every market that they can hire 25% of the Computer Science PhD candidates only adds fuel to the fires of suspicion.

In the interim, it's a shame Bell Labs has gone from world leader to nothing... budget cuts, etc. (Lucent)... there was some real research there, and lots of it was shared with the world.

Re:deservedly

[nine-times]

As it is, it looks less like research and more like unfettered spending to find "yet another" way to dominate.

Or more to the point, my complaint with Microsoft over the last few years is that they seem to have been spending more money on figuring out how to restrict my use of their products, and not very much money on figuring out how to make my life easier.

Now, maybe it's just me, personally, but I'm a home user and an IT professional. I use computers a lot for various things, and Windows seems to be getting harder to deal with. If I have to call Microsoft over another activation problem, I'm going to want to kill someone.... actually the truth is I've past that point a while ago.

Maybe it's just because Microsoft is servicing someone other than me. Maybe there's someone out there who's pleased as punch at the changes in Vista and Office 2007. I honestly think MS hit their peak in 2000, and things have just gotten more frustrating since then. Keep It Simple, Stupid. My needs aren't that unusual or complicated, but Microsoft doesn't seem to be making a lot of headway. Security. Stability. Easy imaging. Effective backups. Compatibility and interoperability. The ability to manage the ever-increasing mail stores. Transparency into what the computer is actually doing so that it can be manipulated more easily for any purpose.

For christ's sake, if you're going to pay so much for "innovation", try to tackle some of the fundamental problems with modern computing, instead of gimmicky wireless sharing for MP3 players, new copy-protection schemes, and snazzy graphics for FreeCell.

Re:deservedly

[El Cubano]

Unfortunately for Microsoft (but true to their character) they have tools for mouthpieces like Ballmer. Microsoft inks a deal in what could only be viewed with raised eyebrows, and Ballmer punctuates that with "they're infringing our IP anyway...". As long as Microsoft continues to be so hostile to the world in general, they get what they sow.

Nobody (or at least most people) argues that Microsoft doesn't come up with original ideas. Their research arm has a ton of truly brilliant people. I mean, Leslie Lamport and Tony Hoare work there. The problem is not that Microsoft can't come up with some innovative stuff. The problem is in how they translate it from their research side to their implementation and then marketing, which is usually pretty lousy.

Re:deservedly

[ewhac]

In the interim, it's a shame Bell Labs has gone from world leader to nothing... budget cuts, etc. (Lucent)... there was some real research there, and lots of it was shared with the world.

Don't be too quick to lionize Bell Labs, as they were the research arm of The Phone Company (AT&T), which itself was the object of scorn for decades for abusing their position of being the only game in town. Just as you argue that Micros~1's research are "ill-gotten gains" from their predatory business practices, one could also level the same argument against the Bell Labs of 40 years ago.

Don't misunderstand; I am in no way a Micros~1 apologist, and would richly enjoy watching the company collapse under its own hubris and technical incompetence. It's simply that, if you're going to slam the company, you need to pick your comparisons more carefully.

Schwab

Re:deservedly

[Anonymous Coward]

If Microsoft were less predatory and less a bully in business maybe the rest of the world would stop looking down their noses at Microsoft's "research". As it is, it looks less like research and more like unfettered spending to find "yet another" way to dominate.

Clearly you know very little about what you're talking about, but as your comment is in perfect accordance with the dominant groupthink it gets modded up anyway. MSR is actually less restrictive than an average PhD program, you can work on basically anything you want, which is one of the reasons PhDs find it so appealling. It is more or less independent from the rest of MS, and the researchers are certainly not driven by a desire to find "yet another way to dominate". Yet this, of course, is precisely also the reason for the difficulty they are having with technology transfer.

It's one thing to look down on MS because of what they bring to market, and quite another to look down on the great work done in MSR, much of which is free to download and use by anyone. If you want to deride professionals doing great work by putting scare quotes around "research" (really, don't you think that's a little much?), do it for a better reason than your kneejerk conflation of what MSR is doing and MS' business practices.

Re:deservedly

[Coryoth]

For christ's sake, if you're going to pay so much for "innovation", try to tackle some of the fundamental problems with modern computing, instead of gimmicky wireless sharing for MP3 players, new copy-protection schemes, and snazzy graphics for FreeCell.

Microsoft research does try to tackle such problems, the dilemma is that their work, as far as I can tell, seems to get ignored when it comes to product development and marketing. What fundamental problems in modern computing is Microsoft research trying to tackle? How about programming concurrent software. Traditionally this is hard, and error prone. What we need is a model of concurrency, and a programming language to support it, that makes programming concurrent systems easy, and make reasoning about it easy. Microsoft is working in that area with C-omega and extension of C# with a better concurrency system. See the tutorials to get an idea of how it works. It's not unique, there are other concurrency oriented languages out there like Occam, AliceML, Oz etc. that handle concurrency well, and other concurrency language extensions, like SCOOP for Eiffel, and JCSP for Java, that seek to add better concurrency models to existing languages. Still C-omea is its own tangent, and has interesting ideas (as do the other similar projects and other languages).

What about the issue of maintainability and quality assurance in software? Certainly that's at the heart of a deep problem, and there are no easy answers. There are things you can do to make better quality assurance easier however. Microsoft's effort on that front is Spec# which adds design by Contract to C# and provides extended static checking (using the Simplify theorem prover) to provide static verification of contracts where possible. This provides another layer of quality assurance, and (by integrating the static checking into Visual Studio) automates most of the work, meaning it requires little extra effort from programmers. Again this is not unique, there's Eiffel which has had DbC but no static verification for a very long time, and there's JML and ESC/Java2 which provides DbC (via annotations in comments) and extended static checking (again using the Simplify theorem prover) for Java - you can even get Eclipse plugins to integrate it into your IDE. Still Spec# is going it's own way (and has much better integration directly into the language than JML, which remains as comments) and has interesting ideas of its own.

The problem is not that Microsoft research isn't doing anything interesting, it's that projects like this tend to get buried, or ignored, or simply have a few ideas shifted into existing products. Things like Spec# offer sufficient gains that Microsoft's marketing department really ought to be crowing about it as a major upcoming feature, and serious effort to properly polish it as a product and get it into C# and VisualStudio should be underway. Instead it remains a page tucked away on MS research with little or nothing said about it.

Re:deservedly

[paulsnx2]

If I have to call Microsoft over another activation problem

Why should I as a business owner or shareholder spend my money to do a task whose result isn't a benefit to the business, but to some other company from whom I bought a product? In other words, when a business pays someone to solve an "activation" problem, they have paid someone to insure that Microsoft was paid. The business receives no benefit, but they are out the money anyway.

When Microsoft pours money into research on how to develop technologies that seek to avoid theft of their product, that is fine until part of their solution increases the cost of ownership. When Microsoft pours money into "securing digital rights", that's fine until part of their solution increases the cost of access to content.

Microsoft and others are struggling to survive in a future where computers have nearly unlimited disk space, increasing numbers of processors, vast memory spaces, and high bandwidth to other computers. Very soon we should be able to run multiple operating systems on a single computer at the same time. Running on virtual machines will be the norm, if for no other reason than to allow applications the freedom they need to run and not step on each other or get killed by viruses and compromised by spyware.

Everyone would be impressed if Microsoft was embracing this future and working to leverage all this power for the sake of the user. Instead, Microsoft appears to be working late into the night doing everything they can to insure each day dawns according to the same old paradigms that made them billions in the past.

Re:deservedly

[steve_l]

One issue with hiring the masters like Lamport is that they like to do their own thing, and want a large staff underneath; in industry its harder to get head count than in academia, where you have undergrads, RAs and phd students to suffer at low cost for the sake of a professor.

The other problem -tech transfer- is the enemy of all R&D labs, and of academia too. There's a lot of good ideas out there, that don't make it out into a world that has the x86 as the primary CPU, A DOS derivative and a Unix derivative as the choices of OS, and C/C++ as the primary programming languages.

FWIW, I work in a corporate R&D lab in the UK, and getting anything taken up is always a miracle to be celebrated. Except when it takes so long to come to market that they shouldnt have bothered. This is why open source is so much better as a way of doing tech transfer. If you have something good, a patch, a test and the ability to argue your case, it can be in the code tree in a week, and in people's hands the next day, in mainstream distros within a month or two.

Whereas MS Research? Vista took 5 years. Every new idea in the last three of those years will have been postponed to its successor. So the lag between an idea and product is 3-5 years, compared to 3-5 weeks.

-steve

Re:deservedly

[Anonymous Coward]

Just what the world needs, C# extensions. The very existence of C# is a fine example of the standard operating procedure at Microsoft, and why many people hate them. To me, it seems that Microsoft spends a great deal of resources duplicating existing technology (Direct3d == OpenGL, C# == Java, IE == Netscape, etc...) to create platform lock-in. Eventually, their "innovations" may even be better than the technology they cloned, but invariably it took many man-years of effort to get there. If MS started with the existing technology and moved forward from there, we'd all be better off, but that's not how MS plays the game.

Unfortunately, I see Microsoft research as another expression of their business model. Then again, Bell Labs was created to protect AT&T's monopoly, so why should MS be any different?

Re:deservedly

[Anonymous Coward]

Having given his foe some "rightious data", tmjdot (or "dot" as his friends on the 'Net called him" eased back into his Relax-tron and pulled his data-visualizers off his head. He puffed on a 'rette and stared up at the cieling. Wasting these guys one at a time was easy enough, but the softies were everywhere. They covered everything and everyone and the virus was spreading.
The resignation that he couldn't do this alone slowly zipped its way through his meat-grid. He needed help. And fast.
His mind made up, dot flipped down his goggles, cracked his datagloved fingers, and jacked in. The polyphonic lightshow of a billion voices of data slipping into his crib illuminated his face. He put out the 'rette and headed off for info-environs unknown in search of free-lance data mercenaries like himself willing to wield a weapon against the softie menance. Somewhere out there the binary existed to kill the menace, to get things back to normal. It was just a matter of getting the right programmertavists to riot with it, and getting it in time.

It was going to be a long night.

Re:deservedly

[cayenne8]

"Why should it be my responsibility to constantly re-prove to Microsoft that I've paid them for a product?"

There are other operating systems, you know.....

I know of at least 2 of them that don't have activation problems at all.

Not a Huge Surprise

[eldavojohn]

A lot of large IT companies looked outside of computer science. I mean, yeah, engineers should be the core of your work force but diversity is always a big plus. It didn't surprise me to see this quote:

The MSR staff, however, is not just computer scientists, it includes psychologists, sociologists, anthropologists and medical doctors who are tasked with pushing the envelope on state of the art technology as much or more than transferring their technology into new and existing Microsoft products.

Large companies shouldn't hire these professions just to "push the envelope." Instead, I would hope that all companies diversified as their employee numbers grow. I work in a large IT company and have witnessed the above professions working effectively--especially in the R&D department.

• Psychologists
  
  One of the areas of studies the gets some of the most criticism from me. But you know what? When it comes to performing experiments on how people think and react to stimuli, psychologists are pretty damn good at it since all their data has been collected empirically from subjects. And who uses the code and devices we make in the end? Humans. And who better to tell you what the effects will be after a human has used your product for hours on end? You know, I've often wondered how many psychologists Blizzard employs because I can play that game for long periods of time with little or no fatigue on my eyes/brain.
  
• Sociologists
  
  As software becomes more and more decentralized and internet based, communities form around it. Communities identify themselves by it. For instance, I am part of the Slashdot community by merely posting on it. Think about how many sociologists that MySpace must employ to predict/track or protect people from social deviance. How do you handle that? How do you address that? Not really an engineer's department.
  
• Anthropologists
  
  Now that's a word I hear thrown around a lot and abused to mean many things. But most importantly, it's the study of diverse kinds of people. If you're an international company, you need anthropologists to view your projects and make sure that you aren't inadvertently calling your product or displaying something that may limit your market or create bad press. Engineers focus on one type of person when they make their product and so you need people to make sure that it is still marketable to the world.
  
• Medical Doctors
  
  Most likely hired for the sheer fact that baby boomers are getting old. Huge market for healthcare. If you can make anything related to it and sell it, you're in the money for the coming years. I may be a horrible monster for saying this but things like Alzheimer's Disease are multi-billion dollar industries based on treatments. Gene therapy and computational techniques in gene sequencing just make the field all that more lucrative.
  
  On top of that, you need to think of the disabled using your product and be conscious of their disabilities. Also, what medical problems might be associated with your product or how can you make it easier on the end user. You don't want a million lawsuits if I'm losing my eye sight or getting arthritis by playing WoW, do you?

I'm actually shocked that list wasn't longer and more astonishing. No music theory majors to look at musical products like Guitar Hero's success? No athletic trainers to combat my country's obesity or offer and IT solution for it? No history majors to ... to ... ok well maybe they really are useless (I'm kidding).

Come on people, this is the R&D of the largest software company in the world. I'm shocked that I'm not more shocked on what they're up to.

Are they really that interesting

[CastrTroy]

Is the stuff that's going on at MS really all that interesting that 21% of PHD students want to work there? Or is the pay just that good? Or are they just looking for a nice shiny star on their resume? It seems to me that there would be a lot more interesting places to work than MS.

Re:Are they really that interesting

[ntropic]

It might come as a surprise to you but one doesn't get a Computer Science PhD to learn how to program, rather one does so to figure out what to solve with a program (unless you are working on Software Engineering).

I have a close friend who joined Microsoft Research last year after his PhD (which included interning there). He also had an offer from Google and a couple of hedge funds. His reason for taking MSR was that Microsoft, for all it's image does actually allow the MSR guys to pretty much do what they want to explore instead of forcing a direction driven by a profit making application of that work. This results in much research not ending up in products (so you don't see it), but doesn't stifle the people working there. This came as quite a surprise to me but when I look at some of the papers the groups in MSR have published, I wonder how far from the truth that is.

oh and BTW, they were paying a good 0.6x higher than Google so that would account for some of those PhDs.

Well, perhaps it might be...

[jimicus]

That for all its "innovation", Microsoft have never in the whole of history released a truly new product. Everything they've ever produced (right the way down to Microsoft Paint - once upon a time there was a DOS version produced by someone else) has been either bought or rehashed from someone else.

Sure, they've played around with things a bit - changed the interface here and there, come up with slight tweaks, But at the end of the day, it's not the tweaks that get recognised as innovation; it's the whole new products.

"research"

[MECC]

FTA:"There are virtually no products Microsoft produces today that have not either taken technology from research, come directly out of research, or been built using the tools and technologies we've created in research," he says.

Does that include Zune? The Microsoft music service? How much research did it take to come up with 'We need to make our own iPod and music service'?

Flame On...

Re:"research"

[EvanED]

That's not MSR. That's marketing research. (I don't know what the department that does that is named though.)

MSR's the group that came up with SLAM, which is now incorporated into the Windows driver framework. It's resulted in (over the last 5 years) two POPL papers (one of the two top-tiered programming language conferences), a PLDI paper (the other of the two), a PASTE paper, a TOPLAS paper, three TACAS papers, three CAV papers, a few workshop papers, and a spinoff project at UC Berkeley called BLAST which is doing things very similar to SLAM. (They've had their own fair share of papers, and probably a doctoral thesis or two, on it.)

MSR's the group that wrote Singularity, an experimental OS written in C#, that has an ASPLOS paper, two EuroSys papers (one of which got the best paper award), and three workshop papers.

MSR's the group that wrote Vulcan, a binary rewriter that allowed them to create a program that records the execution trace of another program and play it back later. This is useful in, for instance, temporal debugging. (Think the Omnicient Debugger for Java, except made to work on any program because it operates on binaries. Except that MSR developed two other applications for the recorded traces.) This, and other projects that MSR has done with Vulcan, have resulted in a number of other papers.

Say what you like about MS in general, but MSR publishes more good research than many (probably even most) university CS depts.

what critics?

[idlake]

I don't know of anybody criticizing Microsoft Research: there are lots of good people doing good work there. People are criticizing Microsoft's business practices and products. Good research doesn't necessarily translate into good products, in particular if a company's primary goal is market dominance through lock-in and other tricks.

Re:what critics?

[idlake]

Yes, I know what you were referring to. My point is that I think your implied argument is flawed. (1) Microsoft making heaps of money is largely unrelated to the intelligence of their technical employees, and (2) Microsoft's technical employees really do have the goal of making great software, but they simply aren't able to do it.

For example, tobacco and fast-food companies are making lots of money, too, but that doesn't mean that their products are good or that most of their employees are smart. Big businesses succeed because of a small number of ruthless and smart business and marketing people at the top; the rest of the employees are little more than hamburger flippers, at Microsoft as much as at MacDonalds. Companies where the technical skill of employees can make a difference are small and medium companies, as well as startups.

Money Can't Buy Brains.

[MrAnnoyanceToYou]

I wish it could. I'd be really brilliant by now......

I'm not knocking the individuals working for Microsoft, it's just that there comes a point in the lifespan of a company where it's past its prime. Getting a truly 'new' product far enough to the front is a gargantuan task, that ends up requiring patents and huge investment because the entire process is so slow.

Let's just compare Apple and MS here for a second. Apple pulls stuff into the mainstream that's pretty new once in a while. They seem to enjoy it. It's been really profitable. But some of the stuff they do is so new that noone can really catch up until it's too late. (see: iPod, good UI, 'stylish' design)
BR Somehow, Apple listens to new ideas, where Microsoft attempts to implement old ones and takes flack for never getting it exactly right. One wonders where this cultural issue is in M$, and what makes the difference between the two. But that's only an academic question.

Re:Where are the results?

[metlin]

Makes their lack of innovation all the more remarkable.

Heh, do you even *know* what the hell you are talking about? Maybe you should try looking at some of the ACM SIG* or IEEE publications in the various fields related to CS.

MSR produces some of the best CS research in the world. Just because their work does not percolate down to the products and services teams at MS does not make MSR lack any innovation.

In fact, if you look into most areas, MSR has made some very cutting edge and valuable contributions.

Maybe you should have a look at the list of publications they have put out since 2000.

Do not confuse research with development. Then again, given that this is Slashdot, blind and ignorant Microsoft bashing is welcome, even if the person bashing it has absolutely no clue whatsoever.

Nice.

Microsoft won two innovation awards just last week

[I'm Don Giovanni]

MS does innovation besides the stuff at Microsoft Research.

I got this from a post to Scoble's blog last week:

Check out the December 3, 2006 entry at the XNA blog, entitled "XNA Game Studio Express and the DEMMX Awards".

Turns out that Microsoft's XNA won two categories at last week's DEMMX Awards:
Best of Show: Innovator of the Year
Microsoft XNA Game Studio Express (Microsoft Corporation)

Game Innovation of the Year
Microsoft XNA Game Studio Express (Microsoft Corporation)

Speaking of XNA (a framework allowing normal folk to make Windows and Xbox 360 games (without the need for a devkit), a great video of it was released last week at Channel 9:
http://channel9.msdn.com/Showpost.aspx?postid=2612 54
The video shows coding, debugging, and deployment of Xbox 360 games using XNA. Although XNA uses C# managed code, one of the sample games shown in the video, XNA Racer, runs at 1080p 30fps with 2x antialiasing.
It's a very cool video. Beyond anything you'd see from Apple, Google, et al.

The notion that Microsoft does no innovation is nonsense.

Microsoft Research meets Microsoft Marketing

[autophile]

Microsoft Research! More computer science papers come out of us than from the top universities! We present them at numerous prestigious conferences around the world!

Now, in partnership with Microsoft Marketing, we are proud to announce... Research4All!

Yes, Research4All is a unique product designed to meet not only the needs of researchers around the world, but also the corporations that feed, clothe, and entertain them! For only $1299.99, you get access to three -- count 'em, three! -- research papers published by Microsoft Research! But wait, there's more!

You may read each paper a total of five times, on a total of one computer! And if you should choose to purchase our Paper Edition (for an additional $499 charge), the ink will degrade after six months. And, as an added bonus, the paper is microprinted so that copying and scanning won't work! We are also working with graphics imaging and word processing vendors to recognize certain unique, secret, and patented characteristics of both the microprinting, as well as the sentence structure!

Research was never this fun!

--Rob

Re:Where are the results?

[MP3Chuck]

Well ... I wouldn't count it out yet. Singularity is only 3-ish years old. In fact, from what I understand they only recently reached the stage of having an interactive command line. IANA Computer Scientist, but I'm sure it's got a long way to go before even its base concepts are suitable for mainstream use. Hell, there's not even any clue as to whether development of Singularity into a mainstream OS would even be feasible.

Not to mention there would be an absolutely massive paradigm shift involved in moving from Win32/64 to a platform like Singularity...

So msft just does pointless research?

[walterbyrd]

>>Just because their work does not percolate down to the products and services teams at MS

So msft spends gobs of money, hiring huge numbers of researchers to do all kinds of research. Msft invents all kinds of stuff. Then msft just throws all of that away, and steals ideas from other companies?

Makes perfect sense to me.

Numbers Don't Add Up

[roca]

I think the numbers quoted from the article here were bungled.

> having hired 700 interns worldwide this year including
> 250 computer science PhD candidates in Redmond alone,
> which is roughly 21% of all the computer science PhD
> candidates in the United States."

http://www.cra.org/CRN/articles/may06/taulbee.html

suggests around 1200 CS PhDs *awarded* in 2004-2005 in the USA and Canada. The number for the USA alone may be lower than this, but it might also be higher since 20% of departments surveyed did not respond. But assuming 1200/year is close to the mark, the number of "computer science PhD candidates in the United States" must be several times that, since a PhD takes several years and furthermore a lot of PhD students never complete their degree. I think an average of five years of studentship per PhD awarded would be a reasonably conservative estimate; then the 21% number quoted should be more like 4%.

When the Functional Programming Revolution hits...

[BitwizeGHC]

When the functional programming revolution hits the mainstream -- and it will very soon now as the current, C++ or Java way of developing software does not scale complexity-wise without requiring ever-increasing armies of Indians or Chinese to grind out the code -- Microsoft will be ahead of just about everybody else because they've retained the likes of Simon Payton-Jones and Erik Meijer to work in their research department. In fact, LINQ may just be the best thing to ever happen to functional programming because now that Microsoft is doing it, it becomes a legitimate enterprise programming activity.

Microsoft is an 800-pound gorilla, but do NOT knock their research arm. Whatever it may have been in the past, these days there are definitely people doing interesting stuff at the very cutting edge of computing