Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Given that IBM have failed to get Watson to be useful anywhere else[1], I guess it's worth a try...

[1] http://www.ft.com/intl/cms/s/2/dced8150-b300-11e5-8358-9a82b...



Watson is a perfectly fine product, it's just surrounded by ungodly amounts of hype. You can do pretty much anything Watson does with a standard desktop up to a certain size (though to be fair scale is one of Watson's main value propositions). The problem is that they market it like it's some sort of hyper-intelligent AI savior of the universe which causes the rest of us to roll our eyes every time those commercials come on.

I assume they're relying on non-ML people who make the money decisions to get caught up in the hype and force it on their organizations, but in my (very limited) experience that hasn't worked out. The people in charge at the organizations I've worked at have been just as skeptical of the hype.


You can actually try Watson by yourself and have your own opinion on it: https://www.ibm.com/smarterplanet/us/en/ibmwatson/developerc.... Many developers find it useful.


Right.

Except this isn't anything to do with Watson-the-jepordy-winner. It's a set of web services which are useful for doing analysis on various things.

The question answering service was withdrawn last year: https://developer.ibm.com/watson/blog/2015/11/11/watson-ques...




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: