Hacker News new | ask | show | jobs
Why Microsoft is betting its future on AI (theverge.com)
47 points by nzonbi 3627 days ago
8 comments

"conversation as a platform"?

Is he talking about that "Tay" debacle? http://www.huffingtonpost.com/entry/microsoft-tay-racist-twe...

Reading this article reminds me about the story of McDonald's. After a certain level in the organization, at a certain executive delusion, McDonalds starts thinking they are all restaurateurs, and not low-quality fast food establishments to get a quick bite. They are so out of touch with what people are really doing and using their products, so misguided, they actually fail to innovate on their core concept.

For a company that finally has a decent web browser (and it's still not anywhere near Chrome/FF in terms of features and developer adoption), I am skeptical at best. Does anyone remember when Steve ignored the web and mobile for so many years, while Google and Facebook and Apple practically ate their lunch? Remember when he famously said he was "betting the company on .net"? http://www.bloomberg.com/news/articles/2000-10-29/microsofts...

"AI" is just a tool like .net is a tool. It's not a product or idea. It's not a service. It's cool to be excited about your HoloLens and 3D calendar, but it's just a calendar, and after the hype has worn down most people would prefer to look at a calendar on a laptop anyway. "AI" is Microsoft's new .net

"AI" is just a tool like .net is a tool. It's not a product or idea. It's not a service. It's cool to be excited about your HoloLens and 3D calendar, but it's just a calendar, and after the hype has worn down most people would prefer to look at a calendar on a laptop anyway. "AI" is Microsoft's new .net

This. It is as vacuous as saying "betting the future on the Fourier transform", or something.

And besides, they just closed the venerable Microsoft Research Silicon Valley, so it's not that they are putting their money where their mouth is.

And besides, they just closed the venerable Microsoft Research Silicon Valley, so it's not that they are putting their money where their mouth is.

That was almost 2 years ago and it doesn't seem like it had much of an impact on their R&D spending http://www.wolframalpha.com/input/?i=Google+vs+Apple+vs+Micr...

> they just closed the venerable Microsoft Research Silicon Valley, so it's not that they are putting their money where their mouth is.

I seem to remember they fired Leslie Lamport.

Exactly. "Betting its future" is a line MSFT has to trot out every few years to get people to pay attention. They previous bet their future on .net, phones, tablets, Azure, cloud...
"as a platform" just sounds so silly in some contexts.

this is "fuzzy command line as a Human–computer interaction interface"

they should really call it an interface not a platform. a bunch of constantly moving, discrete services built by silos is not a platform, its a jagged, nonstandard mountain range. platform implies stable base.

Will CAAP be the next best thing? Who knows? But one thing that has surprised me about MS and this whole bot-frenzy is how much they are opening up to other platforms.

MS offers tools to build a bot within their framework using Node.js with first party library support[0], and for deploying your bot to third-party chat apps (e.g. Telegram, Slack and FB messenger[1]). The idea is to build a bot once, and their service routes the bot's API between different endpoints. All of this with no obligation of ever touching VisualStudio, Azure or dotnet.

This type of attitude would be unthinkable three years ago.

[0] http://docs.botframework.com/builder/node/overview/#navtitle [1] http://docs.botframework.com/connector/getstarted/#navtitle

Interesting. Does the framework support menus and stuff though? Like this:

http://www.humanisethebrand.com/wp-content/uploads/2016/05/f...

After losing on search, mobile, and social, Microsoft needs a win.
The irony being that they picked up the poster child of the untractable CS problems to give it a try.
Perhaps some company introspection would make more sense than investing in an unproven, (currently) hard to monetize CS problem.
What business wants from Microsoft is It Just Works. The same thing business wants from the power company and the water company. No constant fixes. No viruses. No crashes. Overall, corporate America liked Windows 7 and does not want to upgrade.

Maybe they could focus on that for a while.

Started late on gaming consoles, too.
Who isn't betting their future on AI?
We've been getting along just fine so far without it. I imagine that could continue for a few more centuries, right?
Now there's a contrarian opinion ;-)
Use AI to fix their blue screen of death first?
Hey that's not fair, I haven't seen a windows blue screen since...

...well yesterday, actually.

I remember smarterchild way back in the day. It was pretty cool, but it never seemed to take off. I wonder if any of this stuff will?
Because Google started before?
So, let's see: Microsoft is thinking that a lot of the future of their business is new versions of a user interface.

They might make some money this way, but I have to doubt it will be very much money.

Why? At the core, from their computing, people want some significant utility, content, functionality, etc. E.g., I come to HN for content and find the user interface to be just fine except too often there are too many characters per line and to see the whole like I have to use horizontal scroll bars twice for each line or just copy the screen into my favorite text editor and reflow the lines to fewer characters per line.

Satya, for developers, there is a lot that is good in .NET. Now, if you would do much better documenting it, then it would stand a much better chance of taking off like it should and a lot of developers wish it would.

I'm doing an ambitious startup: By a very wide margin, far and away the worst problem was working through 5000+ Web pages of Microsoft documentation of .NET and SQL Server; the quality of the technical writing was not good -- need better explanations and for the jargon links to a glossary. The next worse problem was having to rebuild the boot partition as attempts at SQL Server installation ruined it. The next worse problem was rebuilding the boot partition after malware ruined it. All the work unique to my startup was fast, fun, and easy. Mud wrestling with Microsoft's work wasted literally YEARS, seriously hurt both my startup and me, and nearly killed at least my startup.

It appears from the article and more that Nadella is short on good, new ideas for new directions for Microsoft, is ignoring a lot that users are screaming about, e.g., security and system management, and that developers are screaming about, e.g., better documentation, and is having vague, ethereal, dreams about the power of new user interfaces.

Uh, Satya, I don't really have any problems with translating foreign languages; for any serious content in a foreign language, no way would I trust a machine translation; I don't want my computer talking to me; even when alone I don't want to talk to my computer; no way in front of others will I talk to my computer.

Satya, for my computer giving me advice, no thanks. For unsolicited advice, no way. The second or first time my computer tries to give me advice, I will look how to turn off that functionality.

Satya, what you call artificial intelligence I call at best trivial and otherwise nearly useless, silly, annoying, insulting, absurd, a pain in the back side, and just genuinely stupid.

Satya, I am NOT going to talk to my computer. That is just not negotiable. And, I'm not going to wear a helmet or funny glasses. And I do NOT want a touch screen interface.

I WOULD very much like a better keyboard, e.g., as good as the old IBM PC/AT keyboard.

Satya, let me give you some warm advice: Computing is the greatest opportunity in the history of civilization, but you are addressing it with all the seriousness of toys in Cracker Jacks boxes. Satya, get real, get serious.