AI Ethics Weekly – May 4: Do You Trust AI?

AI Ethics Weekly – May 4: Do You Trust AI?
May 3, 2020 LH3_Admin

If you are feeling overwhelmed during these tumultuous times, you are not alone – The Pandemic Class Divide Is Real — But A Lot Of People Are Stuck In The Middle and the data shows that the pressure is even more so for women as the lockdown has exposed the gender divide. “It’s still expected that women will make their jobs fit around childcare and homeschooling, while a man’s job comes first.”

Here’s some sage advice for these on Reinventing Your Career in the Age of Coronavirus “We need not feel defensive about this apparently unproductive time-out at turning points in our lives … In the apparently aimless activity of our time alone, we are doing important inner business.”

GIF Credit: Barbara Pozzi

Start off your week right with our AI Ethics Weekly newsletter.  Sign up to get this exclusive content delivered to your inbox every Monday!

Responsible AI in the Age of Coronavirus

As dozens of tracking apps for smartphones are being used or developed to help contain the coronavirus pandemic, there are serious concerns about privacy and hastily written software. Natasha Singer @natashanyt shares the Scramble for Virus Apps That Do No Harm.

Evan Selinger @EvanSelinger points out that Offices are buying thermal, facial recognition cameras to screen workers to see if they are running a fever. The demand for the cameras is skyrocketing as companies plan to reopen and h/t to Anna @a_bacci Bacciarelli for sharing that Amazon turns to Chinese firm on U.S. blacklist to meet thermal camera needs.

Given the rapid adoption of AI in high-stakes domains, Kay Firth-Butterfield suggests that we consider this critical question Is AI trustworthy enough to help us fight COVID-19?

Marietje Schaake @MarietjeSchaake shares that Coronavirus Contact Tracing Efforts Underway At Google And Apple but “We need an “army” of contact tracers to safely reopen the country. We might get apps instead. Without enough human contact tracers to identify infected people, the US is barreling toward a digital solution and possible disaster.”

Shefaly Yogendra, PhD @shefaly shared that Israel enables emergency spy powers and “although it is shrouded in secrecy, other countries are believed to collect data from mobile phones to be used in mass-surveillance programmes or in specific criminal investigations.”

And as if that wasn’t disturbing enough, Meredith Whittaker @mer__edith draws our attention to the news that SoftBank-backed Palantir competitor and surveillance tech Banjo’s CEO Damien Patton Was Once Tied to KKK and Neo-Nazis.

Sign up to get the full-length version of this content delivered to your inbox every Monday!

Resources & Tools

Maria Axente shares PwC’s Responsible AI Toolkit, a suite of customizable frameworks, tools and processes designed to help organizations use AI in an ethical and responsible manner – from strategy through execution. h/t Janosch Delcker⁦‪ @JanoschDelcker‬⁩ for sharing this AI Ethics Guidelines Global Inventory, a project by AlgorithmWatch that maps frameworks and principles for how systems for automated decision-making (ADM) can be developed and implemented ethically.

Ada Lovelace Institute @AdaLovelaceInst, Examining the Black Box: Tools for Assessing Algorithmic Systems highlights a new report by the Ada Lovelace Institute and DataKind UK clarifies the terms around algorithmic audits and impact assessments, and the current state of research and practice.

ICYMI

Kathy Baxter shared this excellent read on Common Sense Comes to Computers, which highlights the problem of common-sense reasoning that has plagued the field of AI for decades and a new approach that has made important progress.

Human inventors can breathe a sigh of relief for now as Ronny Bogani @BoganiRonny shares the US ruling that AI cannot be recognised as an inventor. “An artificial intelligence system has been refused the right to two patents in the US, after a ruling only “natural persons” could be inventors.”

Watch SciTech Now | Season 6 | Episode 615 for a discussion on the bias in artificial intelligence technology, shared by the great folks over at The Radical AI Podcast @RadicalAIPod

Here is Richard Socher @RichardSocher’s work on The AI Economist Approach to Improving Equality and Productivity with AI-Driven Tax Policies by applying reinforcement learning (RL) to tax policy design for the first time to provide a purely simulation and data-driven solution.

Sign up to get the full-length version of this content delivered to your inbox every Monday!