Microsoft has suffered a heck of a brain drain in the past decade. I don't believe they can recover from it in the long term. I also don't believe in innovation by executive edict. Google and FB models work because executive edicts are not necessary there: top researchers and engineers run the whole show.
Microsoft research is nearly legendary in many academic circles. As prestigious as many top colleges. There are many brilliant people working there.
Admittadely, the prestige probably has a lot to do with the previous circumstances (work on whatever, don't worry about the company), but it seems like this will be a very beneficial move for Microsoft.
Almost nearly every single successful Google products were either developed out of research or out of 20%, or via acquisition. Most in-house business projects don't end up successful (Google Plus?) A big company has to invest heavily in R&D to push the limit. You just need one big idea that works and become a game changer.
At Google most in-house projects are infrastructural, and if you look at the scale at which they run and the fact that there's basically nothing that even comes close, they are wildly successful. You could also say Android was in-house, because what they bought is not at all what they ended up shipping. What they bought looked like a Blackberry rip-off, what they ended up shipping was an iPhone rip-off. And let us not forget Ads, the in-house system that has dominated online advertising for a decade now.
But I kind of fail to see what any of this has to do with Microsoft. :-)
If you mention the consolidation of MSR labs including the closure of that in Mountain View, that was a bad move (in my opinion).
But, trying to understand here. My experience over a number of years is that experienced individuals are transitory...People move around. Depending on the group in MSFT and Satya's mandate, churn may make sense.
A decade is multiple tech lifetimes...Someone stellar 10 years ago may be on it now or may not...
This level of general MS bs on HN gets a little tiring when there are no facts.
This is not true of Microsoft Research. MSR has for many many years been one of the top industry research labs. Only recently have Google Research and Facebook joined its ranks (and Facebook only in AI research).
So, they can train new folks with some of the brain that is still left. To be honest, they didn't exactly strike me with prolific creativity, rather their business enterpreneurship was working and they still profit from that. It's good but average engineering, at average, moving slow but steady. If innovation happens elsewhere, they just copy it. So they try to in
So, in the long run they can train new folks with some of the brain that is still left. To be honest, they didn't exactly strike me with prolific creativity, rather their business enterpreneurship was working and they still profit from that.
It's good but average engineering, at average, moving slow but steady.
Good luck because great minds think alike. Getting a product idea finished and out to market takes time, and competitors might be able to do it faster.
Skype Translate reminds me of the Star Trek universal translator but it can only do several languages. I remember when Babelfish came out and some translations didn't take too well. There has always been a need for a computer to translate foreign languages and to translate accurately as well.
I think Apple is trying to do the same things as Microsoft for their iPhone and Macintosh series. So Microsoft has to compete with Apple too to get new ideas to market.
It is a shame that Microsoft, Google, Facebook, Apple can't work together to share ideas and see which ones does them best.
If anything, the race to attract top minds has actually forced these companies to open source their proprietary libraries. (TensorFlow, CNTK etc). While they might not be working together, there's definitely a greater good that has happened as a side-effect. I'd argue that the open sourcing of libraries is better than 2 big companies working secretly.
They are sharing ideas. A lot of these technologies have research roots that are published in open journals. Although the companies are competitors, people behind those companies are scientific minded individuals interested in furthering the technological possibilities.
> Is something like HHVM considered a tech breakthrough?
A largely backwards source-compatible, JIT'ed interpreter for one of the world's shittiest designed but most widely deployed languages? That runs the frontend fleet for the #2 site on the Web? While supporting new core language features and extensions? Yes, I would say that's a breakthrough in interpreter implementation, and thus tech.
Pretty good analysis of MSFT, Google and Facebook. In the 1990s Bill Gates shared his visions in 1990 and 1994 at COMDEX (information at your fingertips and "Cairo"). In the 2000s Longhorn, WinFS (failed because dotNet was too slow) and the great work of Microsoft Labs like Photosynth. MSFT was on its peak with Office 2010 and Win7. Though then they closed their Labs department. And Satya Nadella only tries to milks the custumer with SaaS versions of their former on-premises software, expensive license model changes and hostile privacy intrusive spying stuff on end consumer. MSFT is such a mean company.
WinFS failed not because .NET was too slow, but because no one, not even the people in charge of WinFS, could explain why the fuck anyone would want something like this. I also don't get what's so great about Photosynth. Last I checked it completely failed to take off anywhere.
I was an intern in the WinFS team in 2004 - seriously nobody knew what it actually was even inside the team.
The best I could say was that it was a way for lots of applications writing their own complex datastores (like outlook) to share an OS-level platform so that it could be exposed to other applications with a unified API. It was way more a developer tool than a user-facing feature, but that's not how it was marketed.
For the same reason we want "locate". Except it would be nice to be able to query on many more attributes. And without having the machine be unresponsive for a while every day, only to find that "updatedb" decided to churn the machine.
and that is the main problem of Microsoft Research.
MS should have full world mapped in 3D by now integrated into their mapping service and beating google street view in every way possible.
They seem to have had that strategy for a very long time -- entice developers to lock in end users. It worked quite well obviously, but I myself still find a good REPL, solid CLI tools, and fully open source code all the way down to the BIOS even more enticing.
I'm looking forward to seeing what Microsoft's "next" is; they've tried a lot of "nexts" that didn't always pan out.
> We have no intentions on building for Windows Phone or Windows OS and they know that
That is the biggest change in Microsoft IMHO. Talking from personal experience - in the past whenever we worked with MS they always put in stupid requirements such as we had to support X or Y (with X and Y being some new Microsoft technology/platform/product).
It was abuse of their position so I am glad to see this is changing.
Yeah I was in total shock yesterday when I found out they are phasing out app fabric caching in favor of REDIS. Glad times are finally catching up to them, but it feels so much like Sun.. Too little, too late.
Microsoft's independent innovations of the last decade are the kind the average person doesn't really want or need. They're making a skype translator... fantastic... but the market for that is still pretty niche.
When Google released maps/street view, or apple announced the iphone, those were the kinds of things that built excitement because they were big ideas, and it was the kind of future we wanted to live in. Microsoft doesn't really do that, and from this article they still don't get it.
And at the beginning of the multi-touch revolution, while Apple was releasing the iPhone, Microsoft was still busy developing Surface (now PixelSense) - an 80kg, 50x50cm coffee table cum multitouch screen. Really.
So I agree with you, it seems that impressive products of Microsoft research look like proofs of concept rather then viable mass products.
> it seems that impressive products of Microsoft research look like proofs of concept rather then viable mass products.
Isn't that kind of the point of having a research division? Let the research people go crazy with new ideas, then sift through them and figure out what to make of that, product-wise?
It's not like MS Research hasn't delivered some cool stuff, but Microsoft was either not capable or not interested in turning that stuff into products. At least that is the impression I get.
The biggest thing I can think of that MS research delivered in the last few years is the kinect. They hinged a lot upon that being the next thing in gaming, but what is it now really? Their competitors aren't exactly in a rush to copy it... because once again, its a gimmick for a niche market.
Now just look at VR on the other hand... there's many big companies investing in that and there's a lot of hype around it. Heck even Facebook is on board and Microsoft is nowhere to be seen. There's yet another technological revolution that they're missing out on.
The connective fabric is that Microsoft has always been investing in VR, or at least AR/Holograms: the Surface Table/PixelSense was a step towards physically touching digital spaces, the Kinect was a step towards physically touching digital spaces, the HoloLens will be a step towards physically touching digital spaces. All of those steps have learned from the past steps and build on sensors and software built for the previous step. (Crudely, the Kinect moved the table's sensors into a platform that could be attached to an Xbox and the HoloLens is now strapping those sensors to your face.)
I don't think that Microsoft is missing out on the VR revolution: I think that all this time Microsoft has been trying to baby step/bootstrap their way towards some very cool things. I think it's useful to take a step back and appreciate the incremental/evolutionary approach Microsoft has been taking here. I still think it's kind of brilliant that MS/MSR was able to use gaming to commoditize the Kinect (what was a lab's worth of sensors is now available in every Wal-Mart for a reasonable cost).
Is anyone here aware of Hololens? I am not understanding why people are claiming ms is missing out on the vr revolution, unless you consider AR to be so wildly different that it doesn't qualify as similar?
I'm not sure "niche" is a good characterization of the skype translator (assuming the quality is high enough to be more than a novelty toy).
Such a tool might revolutionize international, distributed workforces. With a sufficiently good translator, an engineering firm in the US might quintuple the set of people in the world who are suitable to work for them.
As a developer I noticed lately there is an impressive number of libraries coming from ms. They probably have a hundred libraries just on github and many older ones on codeplex. I hope this new practical money making focus doesn't mean those libraries will get neglected. Part of the charm with ms is that they are not as calculating and ruthlessly capitalist as apple and google. They are more focused on the developer and researcher than the user. I hope they don't end up becoming the same.
Part of the charm with ms is that they are not as calculating and ruthlessly capitalist as apple and google. They are more focused on the developer and researcher than the user. I hope they don't end up becoming the same.
I take it you weren't in the industry when the Halloween documents leaked, or the antitrust case was prosecuted. MS was (and probably still is) every bit as ruthlessly capitalistic as the new 800lb gorillas in town.
The ms is evil thing is very 90s. It doesn't look so bad compared to say facebook and google who's core business is advertisement. Those will sell your personal information to 3rd parties any chance they get. Ms core business is still about selling software and services to users. In my book thats a more honest way to make a living. I hope it stays that way. Apple is another story ..
While petke's post seems a bit... overly optimistic?... It's basically completely nonsensical to focus on news from a decade ago or further when discussing technology. Ten years in tech is like a different century in any other topic.
Halloween documents was 1998, Microsoft's antitrust case was 2001. If you want to neg a company, find examples in the last decade. (They exist. Windows 10 privacy issues are a great example of them still getting some things wrong.)
Part of the charm with ms is that they are not as calculating and ruthlessly capitalist as apple and google
What????MS sues everyone that dare to use pirate copy of Windows just like 10 years ago and keep hostile to OSS until only recently.
It is much more hated before, now since they retreat more from consumer market, it looks like it tries and probably will become the new IBM, while Google is more like the new MS.
I really don't think that the problem with microsoft was their research division being too seperate. although interfacing better with the rest of the company makes sense they've always fallen on their face when it comes to execution, they were a real player in the smartphone sector at one point for example.
figure out how to execute on your ideas, the generation of ideas themselves aren't the issue.
It's entirely custom but it leverages the d3 library so it's quite short/relatively simple in its execution as it leaves most of it up to d3.
Process for finding out is not fixed. Generally a simple "View Source", look at <script> tags, read script files and look for obvious clues is the best first port of call (and worked easily here). If there's a lot of indirection and pre-compilation being done it gets harder.
Unfortunately that animation runs at a choppy snail's pace on an iPad. One would hope cross-browser/device testing was a part of developing fancy stuff for the web.
Any time a CEO of a company this size has to make a last-minute forced integration between groups in their company to catch up with a competitor, it means the CEO failed at their real job: anticipating their competitor, and being prepared with a response.
If CEOs are micro-managing tactics like this, it means they are failing to hire and delegate.
Basically if you consider the delta between what you would do if you only had to care about your product and its customers, and what you have to do to align with some broader strategy - sell Windows, drive search eyeballs, get more clicks, etc. -- that's the Strategy Tax. It can be used for good but usually means you end up with Windows Phone hobbled with no Mac client, etc.
Strategy tax is where an individual product is constrained in some way to further the overall goals of the company. For example, Microsoft killed a project to port SQL Server to Linux because the product would help to strengthen Linux as a competitor to Windows.
The term has been used inside Microsoft since the 1990s (and maybe before that time).
It's an interesting but misleading phrase, because it doesn't allow for the fact that a bad strategy does nothing to further the overall goals of the company - where a relaxation of a bad strategy does.
If your strategy is "Protect the crown jewels at all costs" while your competitors have discovered a diamond mine, no one needs to care what cutlery you're supposed to use.
Of course that sounds ludicrous, but the impression I get is that that kind of strategic confusion has been endemic at MS since Gates moved on.
Clever innovations like Skype Translate or even Hololens are not a strategy. Selling lots of good individual solutions still isn't a strategy.
You can't play defence on strategy (Office 2026? Windows Infinity?) You need a kick-ass plan for the 5-10 year future, and I see no evidence that anyone at MS is capable of designing a product plan for consumer markets that don't exist yet, and then wondering how to market them - while Google, FB, Apple and even Amazon are at least thinking about it.
Wouldn't universal store across devices count? I mean it doesn't matter if you sell X amount of phones or Xboxes or Desktops, you just need the # of devices that have access to the store to give someone a reason to develop for it. I don;t see anyone moving in that direction at all....
The phrase does not imply that the strategy is beneficial to the company. Microsoft employees often used the phrase when they thought the strategy is detrimental to an individual product and the company as a whole.
Gates was directly responsible for imposing strategy tax during his time at Microsoft.
Microsoft could build a robot that did chores around the house for me, and I still would have no interest in it. Because it's microsoft, and there is a lot of baggage that goes along with that. This holds true for other large tech companies as well - there is always baggage. It might be poor updates, crappy support, no production quality, poor execution, poor security, or required integration with their larger platform.
Yahoo is faced with a similar problem; they can't shake who they are, no matter how cool or different they try to rebrand themselves. They're still Yahoo. And nobody is interested.
I say this because MS seems to think that if they get a slamdunk on technology, everything else will fall into place. And maybe there is some truth to that. But you can't escape who you are and what you've done, especially when it's become your M.O.
Facebook has an advantage in this space: they're still a new name in tech. They're "hipper" although still have their own baggage; but it remains to be seen how that permeates throughout what they do. So far it hasn't looked good though.
Really, I think there is too much talent locked up in this major companies. I think the world would be better off if more engineers and creatives were forging their own paths, chasing their own dreams and coming to market themselves. We need less whales and more/bigger fish.
I don;t feel that this is true at all. Few other companies can produce consumer loyalty as those who provide an O/S with which you operate a machine. One of those other industries is automobiles. I really never got it but my father, brother, uncles and cousins were all 'loyalists' to particular auto brands. None of them liked Ford and would trash Ford products as though one of their vehicles ran over grandma. Since Ford did not take any money from the auto bailout in 2009(?) guess who's family is filled with Ford's? - and they rave about how awesome they are.
Point is these are cyclical things. MS will have it's time in the sun again as long as operating systems are a thing. The thing that should scare people is that companies have 'institutional memory'. I feel like Apple, having been so close to closing down in the late 90's could get to that place again and feel like it's not a big deal cause they have been in that spot before. Meaning no one freaks out until it's too late.
MS really bit it when they missed the mobile and tablet markets. And both times can be attributed to bad leadership by Ballmer. When the Board let him go, MS still had great financials but their future was uncertain because they were not even competing with Apple and Google (In a realistic way). Point is MS doesn't wait until a the money dries up to try and right the ship.
The review of the the pixel book got me thinking...
What do FB and GOOG have to prop up their valuation beyond advertising? In both cases the user is the product, advertisers the customer.
Amazon, MSFT, and Apple have other avenues. It's always interesting how a past reputation glosses of a legacy name is used to gloss over the behaviors of "a new name".
You (and I) might have that baggage but the new 20 somethings at work have a very different view of Microsoft.
Their view is of a company who makes great games and consoles and a phone that no one wants. Show them Active Directory, Office365 and they're a little taken aback that Microsoft does other stuff.
