O'Neil is the author of the blog mathbabe.org. She was the former Director of the Lede Program in Data Practices at Columbia University Graduate School of Journalism, Tow Center and was employed as Data Science Consultant at Johnson Research Labs. O'Neil earned a mathematics Ph.D. from Harvard University. Topics discussed include her book "Weapons of Math Destruction," predictive policing models, the teacher value added model, approaches to auditing algorithms and whether government regulation of the field is needed. You can keep up-to-date with the podcasts via our RSS feed, and they are available via SoundCloud and iTunes.

In April several news sources reported that the law firm of Baker Hostetler had hired a legal robot called ROSS. This artificial intelligence system, built on IBM's famous Watson system, automates legal tasks like research and the preparation of legal memorandums in the narrow domain of bankruptcy law. Since then, other big law firms have joined the rush. ROSS Intelligence markets its product as a kind of artificial intelligent assistant: ask it questions in plain language, and legal research is presented. It sounds a lot like Siri, except instead of asking for baseball scores one can ask for relevant case law.

You are statistically incredibly more likely to die in a car crash than on an airplane, but people still fear flying more than driving. Why? Partly, psychologists say, we blow the risks out of proportion because we don't like feeling out of control of our fate. Our life is in the pilot's hands. Self-driving cars will inevitably give people a similar feeling, even if they are much safer than today's vehicles. Computers are expected to be vastly better drivers than humans, but that requires people to turn over the wheel--knowing they won't have control over the computer's split-second decisions if there is an accident.

Since blogging that I Need an AI BS-Meter a number of people have sent me pointers to a subset of AI I loosely think of as Result Explainers -- everything from pending government regulations (EU's Global Data Protection Regulations -- GDPR) to the latest in academic research (Local Interpretable Model-agnostic Explanations -- LIME). As the authors of the EU's GDPR state, widespread adoption of AI cannot occur until vendors are able to communicate results in a "concise, intelligible and easily accessible form, using clear and plain language." This got me thinking, "What should Result Explainers look like?" Should they generate trust scores, a series of Google-Maps like directions that get you from data to results, a series of diagrams? And as my colleague Patrick at Lab41 has pointed out, "Why should we trust a Result Explainer if we don't trust AI to begin with? As you might expect there isn't one right answer. That said, recent advances in recommenders, digital assistants, user interface design and initiatives like DARPA's recently announced Explainable Artificial Intelligence (XAI) grand challenge suggest we may be on the brink of a few breakthroughs. Again, as the authors of the EU's General Data Protection Regulations note, while the resulting classifiers, models, predictors, etc. can be very powerful they also frequently confound explanation -- e.g., the output of SVMs and Gaussian processes can be difficult to render, ensemble methods hide information as a result of aggregation and averaging, neural nets create high data dimensionality, and so on. End users care a lot more about results than they do about models. Unfortunately assessing result quality takes us right back to the models, as nonparametric models are only as good as the data used to train them (along with the type of model structure and associated parameters that were selected). But these models frequently hide information. Part of the magic of AI is that it finds stuff based on features that previously may not have been well understood. Unfortunately, the features models train on are frequently unclear. Assigning labels to pre-trained models can help mitigate some of this ambiguity -- e.g., "This model was trained with over 100,000 high-res color images of cats." These labels may be misleading though, as the model may contain feature biases that are not well understood -- e.g., "the training data is dominated by images of "well-fed, indoor cats from Japan."

As you went about your day quietly humming it, perhaps someone else heard you and complained minutes later that you'd gotten the tune stuck in their head. The song's hook seems to have the ability to jump from one brain to another. And perhaps, to jump from the web browser you are using right now to your brain. In fact, you may be singing the hook to yourself right now. Something similar happens on the internet when things go viral – seeming to follow no rhyme or reason, people are compelled to like, share, retweet or participate in things online.

Online publisher Defy Media, which attracts millions of young adults and teenagers to such goofy information, announced last week that it picked up 70 million from investors. And it's planning to spend most of that on increasing content production, and for the first time, advertising itself. "As the success grew over 2016, people got more and more bullish on the business and we got the outcome we wanted," Defy Media President Keith Richman said about the funding, led by Wellington Management Co. The New York City start-up, which has a significant base in Los Angeles, publishes articles and videos through online brands including Smosh and Clevver. Licensing videos to streaming apps from companies such as Verizon and Comcast brings in the rest of the cash.

Cases like Edward Snowden's are precisely the reason the president's constitutional pardon power exists. Historically, outgoing presidents have often invoked this power in the last days of their terms -- at times on behalf of people who've committed reprehensible acts -- under the premise that mitigating circumstances outweigh the rationale for punishment. President Obama now has the opportunity to use this power proudly, in recognition of one of the most important acts of whistleblowing in modern history. Since Snowden first disclosed documents in 2013 detailing the National Security Agency's mass surveillance programs, we've seen an unprecedented global debate about the proper limits of government spying. This debate has had a transformative effect: on privacy laws and standards, on the security of the devices we depend on to communicate with one another and store sensitive information, and on how we understand our relationship to the institutions that govern us.

Donald Trump claims to be running for president as an outsider. But his campaign has resorted to one of the oldest tricks in the book in touting "Kate's Law." Named after Kate Steinle, who was allegedly fatally shot by a Mexican national in the country illegally, the law would set a mandatory minimum prison sentence of five years for anyone who returns to the United States after having been deported. Trump says this is the bill he'll send to Congress on his first day in the White House. Bills named after sympathetic victims are the worst form of knee-jerk lawmaking, but it's a surefire political vote-getting device.

Around midnight one Saturday in January, Sarah Jeong was on her couch, browsing Twitter, when she spontane ously wrote what she now bitterly refers to as "the tweet that launched a thousand ships." The 28-year-old journalist and author of The Internet of Garbage, a book on spam and online harassment, had been watching Bernie Sanders boosters attacking feminists and supporters of the Black Lives Matter movement. In what was meant to be a hyper bolic joke, she tweeted out a list of political carica tures, one of which called the typical Sanders fan a "vitriolic crypto racist who spends 20 hours a day on the Internet yelling at women." The ill-advised late-night tweet was, Jeong admits, provocative and absurd--she even supported Sanders. But what happened next was the kind of backlash that's all too familiar to women, minorities, and anyone who has a strong opinion online. By the time Jeong went to sleep, a swarm of Sanders supporters were calling her a neoliberal shill. By sunrise, a broader, darker wave of abuse had begun. She received nude photos and links to disturbing videos. One troll promised to "rip each one of [her] hairs out" and "twist her tits clear off." The attacks continued for weeks. "I was in crisis mode," she recalls.

New York real estate heir Robert Durst has been assigned to an Indiana federal prison, frustrating his defense attorney, who said Sunday that he wants Durst sent to Los Angeles to face a murder charge in the death of his friend Susan Berman. Last December, the Los Angeles County district attorney's office reached an extradition deal with Durst's attorneys. Durst, 73, was due to be transferred by Aug. 18 to a federal prison in Southern California after he agreed to plead guilty to a weapons charge in New Orleans. But Durst has remained in a Louisiana jail. His legal team learned Friday that he was to be relocated to a federal prison with a a specialized medical facility in Terre Haute, Ind. "It is contrary to everything that was agreed upon," attorney Richard DeGuerin told The Times.