最妖娆。 发表于 2016-9-30 23:40:25
您需要 登录 才可以下载或查看，没有帐号？立即注册
Why IBM Watson is interested in Twitter data and supporting developers
IBMWatson can do much more than answer questions on Jeopardy. Earlier this month, the artificial intelligence system created amovie trailer. It also powersdriverless cars and is helping doctors with cancer care and research .
Watson is IBM's bet on the next era of computing that is built to understand, reason, learn and interact, according to IBM Watson's Chief Marketing Officer Steve Gold.
SEE ALSO: Facebook, Amazon, Google, IBM, Microsoft form new AI alliance
"It’s not speech recognition like Siri, not speech synthesis like Alexa, but actually understanding human languages," Gold told Mashable .
Gold spoke with Mashable during New York's Advertising Week about who is interested in Watson and the future of partnerships. With rumors of tech companiescircling Twitter for a potential acquisition, we discussed the value of data, and Twitter specifically, and what's next for IBM.
Is there a specific industry that Watson serves?
Coming out of Jeopardy, we knew that Watson had a voracious appetite for data and an ability to provide insights into the data in a way that was evidence-based and in a way what we call confidence-weighted.
Any industry where there’s lots of data, where probabilistic outcomes are useful like health, financial services, banking, insurance are great targets. But the other area is consumer
because we, as consumers, have come to expect a level of personalization that is not endemic of our prior experience.
The second time I call my cable provider, don’t they know who I am? You expect them to use a combination of data, transactional, demographic, attitudinal [background]: read my blogs, look at my tweets, it’s all public. Why wouldn’t you mine that data?
What is the data that is powering Watson?
Ninety percent of the world’s data has been created in the last two years, and 80 percent of that is unstructured, which means it’s dark data to conventional computing. It’s not used at all, and beyond that, 80 percent of the 80 percent is rich media, video, pictures.
But even before you get to the obvious areas like social media, think about organizations. You call a company and they say it’s being recorded. What do they do with that call? If you could transcribe it, they would learn about your likes and dislikes. They would learn about their product from feature and function.
There’s lots of public data that is easily harnessed that is free: travel.gov, health. Most organizations don’t use it in part because data once thought to be a solution actually became the problem.
Why would a company use IBM versus another partner?
We've approached this market a little bit differently. When some of these organizations look at AI, they're like, "How do I improve my product?" But they're not really interested in advancing AI for general use.
We've opened up a platform to allow individuals to use AI to solve real problems. Let's talk about health and cancer. Let's talk about climate change, and then we can do the fun things too.
#FunFactFriday : You can use Watson APIs to create your own Harry Potter sorting hat. Learn how via #IBMLearningLab : https://t.co/MxRVGMR34F pic.twitter.com/vY27UFD1IF
— IBM Watson (@IBMWatson) September 30, 2016
We're still at the early stages. We've invested $6 billion in our idea, a third of that is dedicated to cognitive. There's still this heavy investment required to move this industry along.
IBM partnered with Twitter back in 2014 for data analytics. Would IBM be interested in acquiring Twitter?
I'm happy to give you the corporate answer: We don't comment on rumors.
So then, why is Twitter data valuable? Why is it valuable to IBM, and why is it valuable to other industries?
If you look at data at large, you value it based on what you can learn from it. So much of that is steeped in the use of that particular data. If I'm developing a healthcare app, I really value health-related data.
Twitter is one of the few vehicles [from which] I can get a pulse of the marketplace.
If you think about Twitter, it really talks to the consumer and the market mindset. Twitter is one of the few vehicles [from which] I can get a pulse of the marketplace. It represents insights that I wouldn't necessarily glean from other sources. That's really why data at large, but Twitter specifically, is interesting.
Do you have a 10-year plan, a 20-year plan, a 50-year plan, and what can you tell me about them?
Short answer is yes. The way we do that is we think about the commercialization aspect in shorter, more agile periods, [shorter] than 18 months. We have agile methodology and constant updates and sprints. The incremental change you'll see is greater breadth and depth of functionality.
Part of the problem with you asking that question is we opened the platform to Watson, which is somewhat unusual for IBM. [With] one of the greatest innovations in our 105-year history, you contain it, you commercialize it. We knew this was far too pervasive, so it's not just what you're going to do. So it's also what the market, what the developer will do, and what's amazing about developers is they don't know what they can't do.
What company do you think will have the greatest influence in AI?
It's too early to know the strategies. Everyone has declared AI, but what does that actually mean? At least for IBM, our chairman and CEO has said, "We're all in. This is the future of IBM, and we think it's the future of business." So far, if you look at the general trend and the rate of adoption, it is huge. I think when history is written, it would have eclipsed the rate of adoption for any other prior technology.
上一篇：Transform Your Company and Become One With Your Brand Promise
下一篇：The new tech talent you need to succeed in digital