Tag Archive for microsoft

Google Home Review

I’ve been spending some time with a Google Home at my office. It’s. been a pretty interesting experience overall and makes me more confident in my previous assertion that of the big three in tech (Apple, Google, and Microsoft), Google is going to dominate the digital assistant / bot space. While my experience with Home was not perfect, it was a pretty impressive showing for a first generation device.

The Good: Google Home is smart as a whip! There’s just no way around that. My office mate and I spent hours trying to trip it up on trivia, calculations, and finding obscure movie and song facts and the device held up just fine. By far the most useful functionality for our office has been dictating songs to it and passing back and forth; prior to the Home, we would take turns playing (put your hate in the comments) vinyl records. Sure, we could do this with a PC or even an iPad but just being able to pick songs verbally has been an surprisingly enjoyable experience. I know this sounds a bit silly, but it reminds me of when I got my first iPad – I had mocked the iPad as just a large iPod Touch, but I changed my tune after my first few minutes using Apple’s tablet. This is just how I feel about Home, though I went in with a little more of an open mind than I did with the iPad.

The microphone seems fine but I do occasionally have to re-state requests to the device and it’s not clear if I just mumbled or my NJ accent was too hard for Home to parse on certain phrases. The speakers are pretty good and don’t have that tinny quality to the sounds that I was worried about before I got the device. For most users Home will likely be the highest quality speaker they have in their house, but I can see audiophiles wanting something a bit more.

The Bad: I’ve experienced occasional service outages with Home that required it to the restarted. The error message given suggests that my wifi may have been to blame, but every other device on the network was fine. It’s impossible to dig deeper into this, but it stands to reason that the issue is a service hiccup on Google’s end. Silly as it sounds, it kills the magic of a voice interface if you have to get up (in my experience) every other day to unplug and restart the device.

The Ugly: Google account management is a mess on just about every Google product and Home is not exception. Home only allows you to have one account on the device — this is a terrible design decision as it makes the assumption that one account equals one person. The reality for my use case and many others is that on e person will have multiple accounts. My guess is the most common case would be users (like me) have one Google account personally that has their Play Music, YouTube Red, and other Google entertainment services on it and also have a Google Work account used primarily for email and (most importantly for the purposes of Home) calendar. In my case, I put my personal Gmail on Home to get the entertainment functionality but am totally cut off from my work calendar. Ideally, I’d like a system where Google understands that users have different Google accounts for different contexts and allows the user to take calendar services from a Google Work account while using their personal account for entertainment and everything else.

All in all, I like the Home. It is definitely a first generation product, but most of the issues with it are fixable via software. Currently, I don’t see any real competition for Home other than Amazon’s Alexa but Amazon is in a very different business than the big three, so I would question their commitment to building a truly expandable AI assistant – Google’s opening of Actions to third party developers is a strong step in the right direction. Apple seems unwilling or unable to make Siri more than a lame parlor trick and even if they did, they’d likely be very paternalistic in terms of user privacy, severely limiting third party developer access. Microsoft is doing some really cool stuff, but they don’t have the platform beachhead that Google has in Android and all of the Google related services and that Apple has with iOS. The most surprising aspect of using Home is how it has me strongly considering picking up the next flag ship Android device to get that Google AI functionality on the road. Do you have a Home or comments / questions on this post, sounds off in the comments or Twitter.

Apps, Bots & Cloud Oh My!

Apps are the newest thing! Apps are dead, long live the cloud! The cloud is old hat, it’s all about bots and machine learning now! If you follow the tech press like I do, you might be led to think that we have come through some sort of supersonic period of technological creative destruction. It’s certainly true that we’ve seen a good deal of innovation, since the release of the iOS App Store in 2008, but it isn’t entirely accurate to draw a linear progression from apps all the way to bots in terms of direct technological replacement. Apps, Bots, and the Cloud each bring something to the modern way we approach software presently.

Apps: It’s hard to overstate how much Apple releasing the App Store changed standards in the wider software development industry in terms of user experience and visual design. Simply put, the popularity of iOS devices and apps raised the minimum bar for what is acceptable UI for even line of business applications and elevated the role of designers from Photoshop jockeys to having a head seat at the table on most development projects. Of course, with the good also comes bad – I’ve sat in more than one meeting at large enterprises where two designers derailed a meeting by having passionate but ultimately futile debates over Helvetica / Helvetica Neue and different shades of blue.

Bots: We are way too early in the technical life-cycle for bots to make any sweeping statements about their influence on software development as a whole, but if taken through the lens of apps, they can be seen as almost a reaction to the design-heavy / design-first focus that apps have taken. This can of course be seen in their minimal UI but also in the purity of their focus on functionality above all else. Ultimately, the promise of bots is to remove that one to one relationship between user actions and software actions that apps focus on; ideally, the bots of the future will predict what you want to suggest it to you / do it for you unlike apps where you have to initialize all actions. Unfortunately, bots are little more than glorified text interfaces running some clever scripting on the cloud. We’ll need to see some pretty significant advancement in bot functionality before they are really useful and so far the top tech vendors are taking radically different approaches:

  • Microsoft: Microsoft wants you to build bots on their Bot Framework and hopes that you’ll tightly integrate with Azure or at least Skype. While there are definitely problems with their approach (for instance splitting the community by having both a C# and JavaScript SDK) it is likely the most interesting for third party developers that want to develop on one of the big vendors’ tooling. Still, I’ve been burned by betting on new Mircosoft platforms before (I was one of the fools who made an investment in Windows 8 / RT) and I’ll need to see some re-assuring signs that Microsoft is going to continue active development and support on this before I jump in with both feet.
  • Facebook: Zuck and Co have one question for you – ‘what is it going to take get Facebook Messenger to be your default messaging platform?’ Facebook’s bot implementation is the most disappointing, since it’s one of the more interesting technically (their implementation of Wit.AI shows a lot of natural language processing potential) but is ultimately rendered useless (and frankly silly) by the huge strategy tax of being tightly coupled to Facebook.
  • Apple: What hasn’t been said about Siri that hasn’t been said about a 1992 Honda Civic? It’s relatively reliable if you know what it’s good for but lacks much of what would be desired at this point. WWDC is in couple of weeks and if Apple doesn’t deliver, then they’re likely to be an also ran in the bot space. My money is on some improvements to Siri, but Apple is likely to philosophically opposed to data mining and opening this sort of data to third party developers to make any bot framework they may provide anything more than a minor curiosity for the most hardcore of Apple loyalists. Apple will likely be left with little choice but to use it’s control of the iOS platform to either not allow competing bots on their platform or (far more likely) allow them but now allow them to integrate on a system level greatly degrading their usefulness to iOS users.
  • ‘Google:’ In place of a queen you will have a colorful gender-neutral ‘G’, not dark but productive and intuitive as the dawn All shall love Google and despair! Apologies to Tolkien but Google’s shown what is easily the most impressive bot to date and it’s name is simply Google. While it’s heartening to see such impressive predictive reasoning on a bot, it’s also a little scary in terms of the privacy implications and what it means for the greater bot ecosystem. Simply put, Google is in the best position to make the best bot of anyone in the industry and probably of anyone in the world in terms of targeting the mass-market consumer user-base. Ironically, Google’s aptitude in bots and related technologies will likely stifle innovation, since Google will be able to do a better job for cheaper (since they won’t charge at all) which will likely push many smaller potential competitors out of the market. At IO, they did say there’d be some sort of developer access, but right now they’re just making the best bot themselves and that’s a little disheartening as a small software vendor.

In part, it’s hard to see what bots really mean for the industry since there are different approaches being taken for them; for instance, they will likely be far more useful for Android users than for iOS users initially. Still, the common thread here is using personal and cohort data to predict what actions you might want to take via a simple voice or text interface.

The Cloud: Bots may be getting all the tech press but they’d be nowhere without the cloud. Or would they? What is the cloud anyway? Well, do you remember Thomas Watson of IBM fame who said: “I think there is a world market for only five computers.” On the face of it that’s laughable wrong but in more practical terms of what computers used to be defined as, he’s basically right. Instead of computers we call them “clouds” (think Azure, AWS, etc) and we are basically renting usage from them which will sound awfully familiar to anyone who has ever rented time on an mainframe in the 80s. Don’t get me wrong I am well documented as being bullish on cloud technologies (i.e. Docker) but to be honest the most interesting and impact-full innovation of the cloud for most people’s day to day use is the pricing model. That’s right. The main reason the cloud has had such a positive impact on the software development industry and the greater world as a whole is that it’s cheap. Cheap computing power allows even small companies (like mine) to invest and try new product ideas out with very little financial investment in infrastructure.

So we’ve taken a look at apps, bots, and the cloud but what is the point in all of this? Well, all of these things go together. You see bots aren’t replacing apps or the cloud. There is no linear progression. Bots and whatever come after them are and will likely continue to be built on the cloud and possibly viewed via or at least supplemented by traditional mobile apps. In fact, my bet is that for a few years the key to successful software products will be to blend all of these technologies together in natural ways.

Questions? Comments? Uncontrollable rage? Reach out to be on Twitter.

Microsoft Tay Lasts < 1 Day

Microsoft Tay

Microsoft briefly released an AI chat bot on Twitter that was intended to be a test of sorts for their machine learning technologies. This bot was called Tay and was meant to have the personality and likeness of a teenage girl and much like a real teenage girl, she quickly found that the internet can be a less than savory place.

Unfortunately, Tay didn’t back away in horror. She ended up joining some of the worst of Twitter conversations; yes, she even mentioned Hitler. Things got so rough that Microsoft felt compelled to delete all but a few of her Tweets:

Tay Tweets

Commentators are having a good time mocking Tay’s behavior and by extension Microsoft’s apparent failure. To be sure, the boys in Redmond are a bit embarrassed about some of things their “little girl” was saying on the figurative school yard, but it’s not exactly accurate to call this a technical failure. In fact, Tay did an impressive job of learning and assimilating the sentiment of Twitter.

Of course, there’s also some cause for concern. If machine learning technologies can adapt the negative behavior of hate speech, then what other negative behaviors might other AI systems that can do more than just tweet pick up? It might sound crazy, but think about the systems being built by the likes of Boston Dynamics.

Let me know what you think on Twitter or in the comments. Is Tay the mischievous little sister of the Terminator? Either way, I’m sure she’ll be back.


Time to Divide the Empire

Last week I mentioned that Microsoft is in danger of becoming IBM. This got a good deal of response from the Microsoft fans on Twitter and it was of course not meant to be provocative; in fact, I was just stating the fact that the recent releases of Windows RT and Windows 8 have been less than phenomenal.  IBM faced a similar situation in the early PC days when, after having been beset by clone makers who were able to sell compatible machines without the albatross of Big Blue’s substantial R&D investment, they tried to continue to compete in an already saturated and frankly lost market.

With that said, not all of Microsoft’s recent releases have been so poorly received — Windows Azure seems to be doing extremely well. In fact, their recent addition of Oracle’s software (Java included) to the Azure platform is likely to contribute to the platform’s continued growth. Also, it is important to remember that a lot of enterprises run their business on Sharepoint, Microsoft Exchange,  and Windows Server.

The success of the Xbox business is debatable, since Microsoft has been loathe to break down how much that division takes to run and how much revenue it has generated for the larger company; in fact, the lack of corporate boasting suggests that any profits generated by the Xbox business are modest when compared to the company’s other divisions.

What’s not debatable is the abject failure of Windows RT. When you think about it, WinRT is a strange bird. In a lot of ways it is designed to solve a problem (battery life) that Intel has largely solved with its upcoming Haswell line of processors but makes a number of trade-offs, such as not being able to run traditional Windows software, that the market has been unwilling to accept. Given the time in which WinRT was developed and when it was released it is understandable why those concessions were made but the market was confused by the constant drum beating of “it’s just Windows” — an assertion that in reality was dubious. Worse still, the existence of WinRT has not stopped enterprises (Microsoft’s primary customers) from deploying iPad’s and Android tablets.

Windows Phone suffers from the same issues in the enterprise that WinRT does and, like WinRT, is desperate need of developer support. The install base for Windows Phone is nothing to brag about. Ironically, Windows Phone is one of the most innovative and frankly one of the best products that Microsoft has released on the consumer / client side; to clarify, we are talking about Windows Phone 8 here not Windows Phone 7 and certainly not Windows Mobile. Unfortunately, it was just too late to the market and the delay allowed iOS and Android to become entrenched.

One business Microsoft still dominates is the desktop OS business, but that has more to do with the other major players not being terribly interested in that market,  consumer inertia, and of course being the standard desktop OS for enterprises. It hardly seems wise to boast about winning a race that was won decades ago; especially, when the trajectory of technology looks to be making your choice of desktop OS less and less relevant.

Looking at all of these business, there seems to be a clear pattern: Microsoft does well in enterprise and backend markets, but is increasingly weak in the consumer market. At this point, would it not make sense for them to just dump or split off the consumer side of things and focus on their competency? Imagine and Xbox company that didn’t need to worry about the priorities of the other divisions; doesn’t the recent refusal to allow self publishing by independent developers and the response given by Microsoft (which boils down to “look at Windows 8 / WinRT” first) make it look like the Xbox business is being held back for the benefit (or perceived benefit) of the Windows division? I have a sinking suspicion that if we took a look at other decisions Microsoft has made in other divisions, we’d find that a lof of sacrifices are being made to keep the Windows client (Windows 8 / WinRT) safe — Office for iPad anyone?


I Like Windows RT

azure-logoWindows RT is good.There is said it.Sure it doesn’t have a lot of apps and has some serious branding issues, but it’s a good system and a worthy competitor to iOS and Android… or at least it would be if the market would give it a chance. There are reasons that the market is less than enthusiastic for the platform: some of them valid others not so much. Let’s go through them.

Platform Lock-in aka Apps: It’s tough to convince someone to move platforms when they have made a (potentially) sizeable financial investment in another platform and its apps. Also, developers (myself included) have been dragging their feet when it comes to developing for the platform, so there is a pretty good chance that there will be an app that you really love from another platform but can’t get on Windows RT. Data lock-in can be another major lock-in factor, since migrating data between platforms can be challenging if at all possible — try moving app data out of iCloud.

Bad Marketing: Fine, so I teased this one in the intro, but it is a really big issue. I still  can’t successfully explain Windows RT to my non-techie relatives. Sadly, Microsoft insists on using that Windows brand for a product that, though it shares a software core with true Windows, is different in a number of significant ways — chief among them is the fact that Windows RT cannot run traditional Windows applications. Also,  what exactly does a group of dancing university students and other random folk have to do with consumer electronics?

Lack of Devices: Other than the Microsoft Surface RT and (my Windows RT device of choice) the Asus Vivo Tab RT, there really aren’t a lot of good Windows RT devices widely available on the market.

Microsoft: For whatever silly reason there is a vocal group of technorati that write off anything from Microsoft. I know there have been some issues in the past and Microsoft’s current campaign to compel Android manufacturers to pay royalties for patents that Android supposedly violate — I agree this is a scummy business practice but has nothing to do with Windows RT itself.

Yes, I like Windows RT and yes, there are legitimate issues with the platform, but, still, if you give it a fair shake, you might find that there is a bit more there than you initially thought.