1. AI and Data. Two newsletters in one.

The boundaries between data and AI are blurring. Today, every data scientist needs to see what lies ahead in AI and deep learning. Every AI practitioner needs to understand the foundation on which they stand. And every company that wants to stay competitive needs to build its data infrastructure, tools, and teams with an eye toward scaling for machine learning and AI. As the technologies have become more intertwined, so have the O’Reilly AI and Data Newsletters. So, as of the next issue, we’re merging our Data and AI Newsletters, bringing together two of the most pressing technological trends of the decade into one concise, biweekly read. We hope you like it.

2. Trump’s freeze on new visas could threaten US dominance in AI

“Even before President Trump’s executive order on June 22, the US was already bucking global tech immigration trends. Over the past five years, as other countries have opened up their borders to highly skilled technical people, the US has maintained—and even restricted—its immigration policies, creating a bottleneck for meeting domestic demand for tech talent. Now Trump’s decision to suspend a variety of work visas has left many policy analysts worried about what it could mean for long-term US innovation.”

+ New research shows scientists educated in China help American firms and schools dominate AI.

3. Exploration strategies in deep reinforcement learning

Here’s an intro to several common approaches for exploration in deep reinforcement learning.

4. AI has a problem

Supervised machine learning makes predictions based on the past. But what happens if the future is nothing like the past? A few months into the COVID-19 crisis, a researcher pointed out that algorithms, which had no historical COVID data to draw from, wouldn’t be all that helpful in understanding the outbreak or its spread. British mathematician David Barber warns, “it is important that the AI can alert the human when it is not confident about its decision,” and advocates for what he calls an “AI co-worker situation” (a.k.a. human-in-the-loop).

5. How to improve cybersecurity for AI

Although AI is often used to detect cybersecurity threats, not as much attention is given to how AI systems can themselves be secured. This report from the Brookings Institution’s Artificial Intelligence and Emerging Technology (AIET) Initiative looks at the vulnerabilities of AI and machine learning and how policymakers can help address these challenges.

+ O’Reilly learning paths: Supervised Machine Learning in Security Applications and Unsupervised Machine Learning in Security Applications

Practical Deep Learning for Cloud, Mobile, and Edge

This step-by-step guide teaches you how to build practical deep learning applications for the cloud, mobile, browsers, and edge devices using a hands-on approach.

6. UK releases guidelines to help governments accelerate trusted AI deployments

The UK has released new guidelines during the World Economic Forum (WEF) to “help society tackle big data problems faster” and to prepare for future risks.

+ AI Procurement in a Box

7. A look at federated learning

Federated learning enables machine learning in privacy-critical applications like medical imaging. This article in Nature Machine Intelligence gives an “overview of current and next-generation methods for federated, secure and privacy-preserving artificial intelligence with a focus on medical imaging applications, alongside potential attack vectors and future prospects in medical imaging and beyond.”

8. Neil deGrasse Tyson and Sophia the Robot explore COVID-19 and AI