Watson Studio is changing how companies use AI

Screen Shot 2018-04-17 at 8.22.27 AM

The black box of teaching computers AI is now in the form of an open platform called Watson Studio. From Jupyter notebooks where you can run code that processes your data, then view the results inline, to a visual modeler where you can connect nodes to build a flow to explore your data and train machine learning models, Watson Studio let’s the average person be a teacher to Watson. As described in this article on IT World, “IBM wants to open up the deep learning expertise bottleneck“, this will open up a whole new area for technical practitioners to dive into. While these applications on Watson Studio are pretty simple to get started with, it will take practice and discipline (just like any other technology) to do it right.

The Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text. –link

As a pet project, I have been playing with Watson Studio; as you may have noticed with my Baseball Card project over on CodeByLarry. Being able to teach Watson about your domain is critical to the future of AI. I am seeing more and more the need for industry or vertical experts needing to be able to take their years of experience and translate it into a format Watson can understand. Watson Studio let’s you do this. No longer will Watson or any other AI platform be a black box. They will be driven by tools for subject matter experts to teach AI systems; some will include a bit of code and others will simply be like using Microsoft Word or Excel.

We may not be entirely there yet but you can see we are getting closer and closer. With tools like Watson Studio you don’t have to be a PhD in AI to learn or get started teaching Watson. If you want to dive into the deep, check out some of the articles on Watson Studio, for instance this is a good starter: Get started with IBM Watson Machine Learning and AI.

At the very least, I see many of the products that incorporate AI into them including “tools” to let companies extend the AI capabilities of the core product. This will allow companies to develop “competitive advantages” over other companies in their industry. Until now, this has always been a question from business leaders – “If I use this system, how will it benefit me over my competition”. Of course the immediate question is you will have it first, but what happens when everyone has it? Tools like Watson Studio are going to be critical for companies to be able to have the competitive advantage in their industry. Their company knowledge and their SME’s will once again be the differentiators for them.



Teaching Watson Baseball Cards Part 2 – Team recognition

In this video I extend my visual recognition models to include more samples and now two teams. Let’s see how the results are with over 100 samples in each class identifier.

Continue reading

Teaching IBM Watson to identify Baseball Cards in under 5 minutes

Using Watson Studio and the Watson Visual Recognition Service I show how easy it is to teach Watson by simply uploading pictures of baseball cards and putting class identifiers on them.

Continue reading

12 Kubernetes distributions to think about

Really good and simple article over on InfoWorld that describes 12 of the most popular Kubernetes distributions: click here.

  1. CoreOS Tectonic
  2. Canonical Distribution of Kubernetes
  3. Docker Community Edition / Docker Enterprise
  4. Heptio Kubernetes Subscription
  5. Mesosphere DC/OS
  6. Mirantis Cloud Platform
  7. Platform9 Managed Kubernetes
  8. Rancher 2.0
  9. Red Hat OpenShift
  10. Stackube
  11. SUSE Cloud as a Service Platform
  12. Telekube