Let $\theta_0,\theta_1 \in \mathbb{R}^d$ be the population risk minimizers associated to some loss $\ell: \mathbb{R}^d \times \mathcal{Z} \to \mathbb{R}$ and two distributions $\mathbb{P}_0,\mathbb{P}_1$ on $\mathcal{Z}$. We pose the following question: Given i.i.d. samples from $\mathbb{P}_0$ and $\mathbb{P}_1$, what sample sizes are sufficient and necessary to distinguish between the two hypotheses $\theta^* = \theta_0$ and $\theta^* = \theta_1$ for given $\theta^* \in \{\theta_0, \theta_1\}$? Making the first steps towards answering this question in full generality, we first consider the case of a well-specified linear model with squared loss. Here we provide matching upper and lower bounds on the sample complexity, showing it to be $\min\{1/\Delta^2, \sqrt{r}/\Delta\}$ up to a constant factor, where $\Delta$ is a measure of separation between $\mathbb{P}_0$ and $\mathbb{P}_1$, and $r$ is the rank of the design covariance matrix. This bound is dimension-independent, and rank-independent for large enough separation. We then extend this result in two directions: (i) for the general parametric setup in asymptotic regime; (ii) for generalized linear models in the small-sample regime $n \le r$ and under weak moment assumptions. In both cases, we derive sample complexity bounds of a similar form, even under misspecification. Our testing procedures only access $\theta^*$ through a certain functional of empirical risk. In addition, the number of observations that allows to reach statistical confidence in our tests does not allow to "resolve" the two models -- that is, recover $\theta_0,\theta_1$ up to $O(\Delta)$ prediction accuracy. These two properties allow to apply our framework in applied tasks where one would like to \textit{identify} a prediction model, which can be proprietary, while guaranteeing that the model cannot be actually \textit{inferred} by the identifying agent.

Amid declining sales and evidence that smoking causes lung cancer, in the 1950s tobacco companies undertook PR campaigns to reinvent themselves as socially responsible and to shape public opinions. They also started funding research into the relationship between health and tobacco. Now, Big Tech companies like Amazon, Facebook, and Google are following the same playbook to fund AI ethics research in academia, according to a recently published paper by University of Toronto Center for Ethics PhD student Mohamed Abdalla and Harvard Medical School student Moustafa Abdalla. The coauthors conclude that effective solutions to the problem will need to come from institutional or governmental policy changes. The Abdalla brothers argue Big Tech companies aren't just involved with, but are leading, ethics discussions in academic settings.

Three Californians say that the video game publisher Electronic Arts is secretly manipulating them. On Nov. 9, they filed a class-action lawsuit accusing EA of surreptitiously using a patented A.I. technology known as dynamic difficulty adjustment in its FIFA, Madden, and NHL games--three of the biggest sports games on the planet. The lawsuit claims EA is using the technology to unfairly increase the difficulty of multiplayer mode online matches in order to encourage players to spend real-world money to boost their chances of winning. EA has denied ever implementing the technology and has called the lawsuit "baseless." For years, players have been stewing over ideas of fairness and balance in games, feeling taken for granted at best and taken advantage of at worst. The class-action complaint, Zajonc et al. v. Electronic Arts, doesn't contain any evidence for its claim, but that's fairly typical for this sort of class-action complaint.

Is the cheer and mirth of the work from home transition beginning to fade away because of the lack of support system you once enjoyed at the office? When could you simply ask for a file, a case note, case research, and stay on top of all the matters all day long simply by interacting with your interns, paralegals, staff, or associates? Well, it certainly would have been great if you could get an exclusive resource to help with those valuable yet mundane repetitive tasks of passing along the information to clients, generating quick templates, or imagine if someone could address all those common concerns of the clients at odd hours, so you could give your undivided attention to the important legal issues that need your expert inputs! After all, there are things the sentient beings will always be better at, whether it be advocacy or being business leaders with a legal perspective. However, artificial beings can certainly take a load of all those tasks off your shoulders that take you less than one second of thought.

Asimov was essentially an optimist, but he realised that future AI devices, and their designers, might need a little help keeping on the straight and narrow. Hence his famous Three Laws, which have influence in science and technology circles to this day. Now, almost 80 years later, legal academic and artificial intelligence expert Frank Pasquale has added four additional principles. He's given RN's Future Tense podcast the lowdown. Here's what you need to know.

Alok Choudhary is Chief Scientist at Mediaocean, the essential platform for omnichannel advertising. In 2020, the digital walled gardens have more working in their favor than ever before. When you add Google's effective death knell to third-party cookies to the passage of CCPA's enforcement date, we see a dramatic acceleration of a trend that was already well underway: Closed ecosystems are garnering an insurmountable edge in the world of online advertising. Against the backdrop of the Covid-19 pandemic, mega-players like Google, Facebook and Amazon have only widened their moats. The advantages of closed ecosystems were indisputable well before the abolition of cookies, the go-live of new privacy laws and pandemic-driven consumer behavior shifts.

In 2020, the digital walled gardens have more working in their favor than ever before. When you add Google's effective death knell to third-party cookies to the passage of CCPA's enforcement date, we see a dramatic acceleration of a trend that was already well underway: Closed ecosystems are garnering an insurmountable edge in the world of online advertising. Against the backdrop of the Covid-19 pandemic, mega-players like Google, Facebook and Amazon have only widened their moats. The advantages of closed ecosystems were indisputable well before the abolition of cookies, the go-live of new privacy laws and pandemic-driven consumer behavior shifts. That said, their advantage was built on technology that is rapidly becoming more democratized across the media landscape: artificial intelligence. As AI applications become more common and sophisticated among a diverse set of industry players, we're going to see the power within the media landscape continue to shift.

There aren't enough game consoles in the world for our upcoming locked-down holiday. As Nintendo similarly struggles to keep up with demand, the number of people searching iFixit for Switch repair guides has more than tripled since last year. Traffic to our Joy-Con controller repair page started growing dramatically on March 14--the day after President Trump declared a national emergency. It's been surging ever since. At a time when so many of us are turning to games for fun, stress relief, and social connection, it is imperative for our collective sanity that we press every game console into service.

Google's dismissal of a top artificial intelligence researcher vocal about the company's failures to address the lack of diversity in its workforce has drawn sharp new scrutiny of its treatment of Black employees, particularly women. Timnit Gebru says she was fired via email after refusing to retract a research paper that asked tough questions about a type of artificial intelligence, including Google's use of it. Jeff Dean, Google's executive in charge of AI, told employees in an email that Gebru's paper did not follow the rules for work published externally. Some 2,000 Google employees signed a petition protesting the company's handling of the situation. Academic researchers called out Google on social media in a rare and widespread rebuke.

AI is, loosely speaking, a form of machine learning and Wang believes it needs to be a part of education in schools.