Emergent // Future — Humans in the Loop, AI in the Cloud, and Projects at Home

Issue 47
 This week we check out what it means to have machine learning systems with
humans in the loop, how AI in the cloud is the next frontier for Amazon, Microsoft and Google, and our favorite reads and some projects to try at home.

🚀 Forwarded from a friend? Sign-up to Emergent // Future here.

👋 Spread the love of E//F on Twitter and Facebook

Humans in the Loop

Though we’ve seen advances in the quality and accuracy of pure machine learning systems, the most accurate paradigms are those that use “human-in-the-loop.”

We’re not talking about just labeling data here.

By adding humans to correct results, we enable the machine learning system to actively learn and correct itself from classifications it initially got wrong. With every iteration (and more classified data) the classification model gets more accurate.

AI in the Cloud

The next cloud battle between Amazon, Microsoft and Google is about bringing AI to all businesses.

As companies try to better analyze, optimize, and predict everything from sales cycles to product development, they are turning to AI techniques like deep learning.

The cloud infrastructure market is worth $25 billion and democratizing access to AI could decide which tech giant emerges as the ultimate winner.

​What We’re Reading 📚

  • A conversation with AI pioneer Yoshua Bengio. Deep learning is changing the way that AI has been thought of in the last few decades, and it’s taking some of the ideas from more traditional approaches to AI and integrating them, combining some of those good ideas. (Microsoft)
  • Open sourcing Sonnet, a new library for constructing neural networks. Making Sonnet public allows other models created within DeepMind to be easily shared with the community, and we also hope that the community will use Sonnet to take their own research forwards. (DeepMind)
  • Unsupervised sentiment neuron. We’ve developed an unsupervised system which learns an excellent representation of sentiment, despite being trained only to predict the next character in the text of Amazon reviews. (OpenAI)
  • Caption this, with TensorFlow. How to build and train an image caption generator using a TensorFlow notebook. (O’Reilly)
  • Using Blockchain to Keep Public Data Public. Using the public blockchain in this manner would not only address our data access and manipulation issues but also lay the groundwork for a better system to more efficiently and effectively regulate the fastest-moving startups. (Harvard Business Review)

Things To Try At Home 🛠

Emergent // Future is a weekly, hand-curated dispatch exploring technology through the lens of artificial intelligence, data science, and the shape of things to come.

🚀 Forwarded from a friend? Sign-up to Emergent // Future here.

Follow @EmergentFuture for more on frontier technology

Lovingly curated for you by Algorithmia

Originally published at blog.algorithmia.com on April 11, 2017.

Show your support

Clapping shows how much you appreciated Matt Kiser’s story.