Civil War Within Google
Here are five important AI stories from the week.
In this brilliant long form read, author Beth Kowitt discusses the civil war raging inside Google around the company’s mission and its use (or misuse) of AI. AI ethics has become a very large issue in the tech industry, and this article helps elucidate just why.
Machine learning hedge funds are on the rise, but, given how little signal there is relative to the noise in the markets, these hedge funds face an uphill battle. Also, hedge funds that rely on supervised learning may train models on historical periods, but, if future periods do not resemble historical periods closely, such models may fail. Quants believe with enough data such challenges are surmountable.
To power today’s supervised machine learning models, lots of highly quality labeled data is necessary. AI startups employ humans to label lots of data often on crowd-sourced platforms such as Amazon’s Mechanical Turk. This article explores the market for these human labelers, many of whom live in low-wage economies such as Malaysia.
Facebook is one of the leaders in AI, but even it cannot use AI to successfully remove toxic content from its social network. This read explores how the role of Facebook’s CTO, Mike Schroepfer, has evolved over the years at Facebook.
In the past, speech translation from one language to another involved taking speech, converting into text, performing machine translation (i.e., translating the text from one language to another), and then converting the translated text back to speech. Now, Google is able to translate speech in one language directly to speech in another language; one very impressive feature is the ability to retain the source speaker’s voice in the newly translated speech. For more, please listen to the audio clips on Google’s AI Blog.
More Stories Worth Reading and Watching…