A former Apple engineer has pled guilty to stealing hardware and schematics for the company's autonomous car project. Xiaolang Zhang was arrested at San Jose International Airport back in 2018 while attempting to catch a last-minute flight to China. Zhang was accused of stealing circuit boards and a Linux server from Apple's secret development labs. He also allegedly transferred a 25-page document to his wife's laptop that included technical manuals and engineering schematics of a circuit board for the prototype self-driving vehicle. According to a court document (PDF) filed this week, Zhang has changed his plea to guilty.

Recent research has shown that seemingly fair machine learning models, when used to inform decisions that have an impact on peoples' lives or well-being (e.g., applications involving education, employment, and lending), can inadvertently increase social inequality in the long term. This is because prior fairness-aware algorithms only consider static fairness constraints, such as equal opportunity or demographic parity. However, enforcing constraints of this type may result in models that have negative long-term impact on disadvantaged individuals and communities. We introduce ELF (Enforcing Long-term Fairness), the first classification algorithm that provides high-confidence fairness guarantees in terms of long-term, or delayed, impact. We prove that the probability that ELF returns an unfair solution is less than a user-specified tolerance and that (under mild assumptions), given sufficient training data, ELF is able to find and return a fair solution if one exists. We show experimentally that our algorithm can successfully mitigate long-term unfairness.

Models for bankruptcy prediction are useful in several real-world scenarios, and multiple research contributions have been devoted to the task, based on structured (numerical) as well as unstructured (textual) data. However, the lack of a common benchmark dataset and evaluation strategy impedes the objective comparison between models. This paper introduces such a benchmark for the unstructured data scenario, based on novel and established datasets, in order to stimulate further research into the task. We describe and evaluate several classical and neural baseline models, and discuss benefits and flaws of different strategies. In particular, we find that a lightweight bag-of-words model based on static in-domain word representations obtains surprisingly good results, especially when taking textual data from several years into account. These results are critically assessed, and discussed in light of particular aspects of the data and the task. All code to replicate the data and experimental results will be released.

Data is central to the development and evaluation of machine learning (ML) models. However, the use of problematic or inappropriate datasets can result in harms when the resulting models are deployed. To encourage responsible AI practice through more deliberate reflection on datasets and transparency around the processes by which they are created, researchers and practitioners have begun to advocate for increased data documentation and have proposed several data documentation frameworks. However, there is little research on whether these data documentation frameworks meet the needs of ML practitioners, who both create and consume datasets. To address this gap, we set out to understand ML practitioners' data documentation perceptions, needs, challenges, and desiderata, with the goal of deriving design requirements that can inform future data documentation frameworks. We conducted a series of semi-structured interviews with 14 ML practitioners at a single large, international technology company. We had them answer a list of questions taken from datasheets for datasets (Gebru, 2021). Our findings show that current approaches to data documentation are largely ad hoc and myopic in nature. Participants expressed needs for data documentation frameworks to be adaptable to their contexts, integrated into their existing tools and workflows, and automated wherever possible. Despite the fact that data documentation frameworks are often motivated from the perspective of responsible AI, participants did not make the connection between the questions that they were asked to answer and their responsible AI implications. In addition, participants often had difficulties prioritizing the needs of dataset consumers and providing information that someone unfamiliar with their datasets might need to know. Based on these findings, we derive seven design requirements for future data documentation frameworks.

Responsible AI, Ethical AI, AI for social good -- I am sure you must have heard these terms at some point or the other, whether you are a Data Scientist or not. "The development of full artificial intelligence could spell the end of the human race." And there my journey of understanding this critical aspect of the AI foundation started. I used to wonder how to relate ethics with AI which is just a series of algorithms, when, in fact, we have not been able to apply ethical behavior among ourselves. As per the AI index report published by the Stanford University Institute for Human-Centered AI, cybersecurity and regulatory compliance are among the top risks identified by AI/ML-oriented organizations. Another report reveals how AI has captured interest among undergraduate students.

With recent advances in brain implants touted by the likes of Elon Musk, mind-altering technology might not be the stuff of science fiction for much longer. Neurotechnologies are brain implants or pieces of wearable tech that interact directly with the brain by monitoring or influencing neural activity. They are already being used in medicine to treat Parkinson's disease and tested by military organisations looking to employ'cyborg soldiers'. A report published this month by lawyer Dr Allan McCay from Sydney University looked at the ways the legal profession could change if implants become more mainstream in society. It suggests that law enforcement agencies could utilise brain chips in order to manage the behaviour of the convicted and help prevent re-offending.

Back in 2018, former Apple employee Xiaolang Zhang was arrested at San Jose International Airport where he was going to board a last-minute flight to China. Zhang was accused of transferring a 25-page document that includes the engineering schematics of a circuit board for the company's self-driving vehicle, along with technical manuals describing Apple's prototype, to his wife's laptop. He was also accused of stealing circuit boards and a Linux server from the company's development labs. Now, Zhang has pleaded guilty to a felony charge of theft of trade secrets in San Jose federal court, according to CNBC. The news organization has obtained a court document (PDF) summarizing the proceedings in which Zhang changed his plea -- he originally pleaded not guilty when he was indicted in 2018.

Over the last few weeks we all have been overwhelmed by the response and have been working hard to ensure a safe and ethical release, incorporating data from our beta model tests and community for the developers to act on. This is a permissive license that allows for commercial and non-commercial usage. This license is focused on ethical and legal use of the model as your responsibility and must accompany any distribution of the model. It must also be made available to end users of the model in any service on it. This understands concepts and other factors in generations to remove outputs that may not be desired by the model user.

A former Apple engineer has pleaded guilty to trade secret theft -- one of two people accused of stealing trade secrets from the iPhone maker's nascent self-driving car program. United States federal prosecutors have alleged that Xiaolang Zhang downloaded the plan for a circuit board for Apple's self-driving system after disclosing his intentions to work for a Chinese self-driving car startup and booking a last-minute flight to China. He was arrested at the San Jose airport after he passed through a security checkpoint. Zhang initially pleaded not guilty to the charges, but according to court documents on Monday, he had reached a plea deal with prosecutors and changed his plea to guilty. The plea deal is sealed and sentencing is set for November.

Role requiring'No experience data provided' months of experience in San Francisco We are a startup within an enterpise business and have huge growth plans! Our product relies on Computer Vision to make it easier for customers to choose between different product offerings. We are headquartered in the Bay Area but have engineers throughout the country! With a remote-first culture, we strongly believe in collobaration via Microsoft Teams. Our software is used daily by millions of customers globally and we are still gaining new customers, we have exciting plans for the future!