IBM is at the forefront of artificial intelligence with their new Watson system. Watson is beginning to paint a bigger picture for consumers on how exactly machine learning can be used across multiple industries. Watson is essentially a huge repository of information that can be accessed through natural language or voice queries. To me this sounded a lot like Amazon's Alexa or Siri on the MAC OS, so what exactly is their value proposition?
Watson is a collection of API's that can be rented or bought by companies to help assist them in their development of AI solutions. One of their products coined "Tone Analyzer" is advertised to understand emotions, social tendencies, and perceived writing style. Others come in the form of a "Language Translator" and a "Natural Language Classifier". Basically, IBM is selling data analytic services related to Natural Language Processing.
Their focus ares for the commercialization of their product are Customer Engagement, Education, Healthcare, IoT, Media, Talent, and Work. The most used and advertised currently is Watson Health, it has been used to diagnose cancer and in drug discovery. IBM claims that it can outperform human professionals in diagnosis for some cases of cancer. In 2013, 4 years ago, it could already search through 1.5 million patient records as well as thousands of other pieces of medical evidence. On the lighter, more fun side of the technology, Watson Chef created 65 original recipes based on human preferences and reading in 9,000 different recipes from Bon Appetit.
Other tech companies and past employees have posted articles saying that IBM is promising more than they can deliver in their focus areas. Others think that Google's AI technologies; TensorFlow and Deepmind, make Watson seem like a not so formidable adversary. TensorFlow is a free open-source platform for developing AI software, and we've all seen the amazing capabilities of Deepmind.
Could these professionals be correct? Will TensorFlow render Watson less profitable?
How do Google and IBM's premier AI technologies Deepmind and Watson differ?
Link: https://www.ibm.com/watson/products-services/
Link: https://medcitynews.com/2017/09/former-ibm-employee-ai-truth-needs-come/#disqus_thread
:
Awesome article about IBM's Watson! I personally have worked and used Watson for numerous products in my internship with IBM. To address your question- how do IBM and Googles's AI technologies differ? They both utilize Machine Learning capabilities, however Watson focuses on NLP, Natural Language Processing, and DeepMind specializes in Pattern Recognition. Both great technologies that will have significant impact in the years to come!
ReplyDeleteGreat article. To answer you question about if TensorFlow and Deepmind will render Watson less profitable, I don't think it will. These systems are very different in what they were built to do and who would use them. Although they can do similar things they are more effective at particular tasks. To expand on Joshua's comment Watson is a single system built to be extremely efficient in relevant information retrieval. DeepMind is built to learn and adapt utilizing reinforced learning and function approximation. Tensorflow is built for data science and development/ manipulation utilizing decent computing power. How people use them is Watson would be great at answering questions, Deepmind would be good at playing complex strategy games while learning and adapting in real time, and Tensorflow would be good at researching and developing.
ReplyDeleteNice article bout the IBM.
ReplyDeleteAnd found this one from dapp.com
Only a few days left! Join the CryptantCrabs giveaway and try your luck!😎
Dapp.com is a largest dedicated platform for sharing exciting dapps and valuable knowledge about decentralized technology. We help everyone understand, create, and enjoy this exciting new technology with enthusiasm.
Check this out to join. ⬇️⬇️
https://dapplin.blogspot.com/2018/11/dapps-are-hosting-giveaway-with.html?m=1
🔘Dapp.com
https://t.me/dapp_com