Tag: machine-learning

33 links

ai.googleblog.com > Lora Aroyo and Praveen Paritosh
Uncovering Unknown Unknowns in Machine Learning
11 feb. 2021 - The performance of machine learning (ML) models depends both on the learning algorithms, as well as the data used for training and evaluation. The role of the algorithms is well studied and the focus of a multitude of challenges, such as SQuAD, GLUE, ImageNet, and many others. In addition, there have been efforts to also improve the data, including a series of workshops addressing issues for ML evaluation. In contrast, research and challenges that focus on the data used for evaluation of ML models are not commonplace. Furthermore, many evaluation datasets contain items that are easy to evaluate, e.g., photos with a subject that is easy to identify, and thus they miss the natural ambiguity of real world context. The absence of ambiguous real-world examples in evaluation undermines the ability to reliably test machine learning performance, which makes ML models prone to develop “weak spots”, i.e., classes of examples that are difficult or impossible for a model to accurately evaluate, because that class of examples is missing from the evaluation set.
 · adversarial-ai · crowd-sourcing · generative-adversarial-networks · machine-learning · research

joanna-bryson.blogspot.com > Joanna Bryson
Three very different sources of bias in AI, and how to fix them
13 jul. 2017 - Since our Science paper came out it's been evident that people are surprised that machines can be biased. They assume machines are necessarily neutral and objective, which is in some sense true -- in the sense that there is no machine perspective or ethics. But to the extent an artefact is an element of our culture, it will always reflect bias.
 · accountability · algorithmic-bias · artificial-intelligence · black-struggle · machine-learning · racist-technology · tools-for-justice

medium.com > Jacob Metcalf
“The study has been approved by the IRB”: Gayface AI, research hype and the pervasive data ethics gap
30 nov. 2017 - Just as it has changed the methods of science and engineering, the tools of large scale data analytics have caused major shifts in how we judge the ethical consequences of scientific research. And our current methods are not keeping up. Historically, research ethics has been animated by a core set of questions, such as how do you decide if a scientific experiment is justified given the potential risks and benefits to the people being studied, or to society at large? How do you track who has to bear those risks and who gets the benefits?
 · data-ethics · data-science · gay-rights · machine-learning · project-hva · research-ethics

locusmag.com > Cory Doctorow
Past Performance is Not Indicative of Future Results
2 nov. 2020 - In “Full Employment“, my July 2020 column, I wrote, “I am an AI skeptic. I am baffled by anyone who isn’t. I don’t see any path from continuous improvements to the (admittedly impressive) ‘machine learning’ field that leads to a general AI any more than I can see a path from continuous improvements in horse-breeding that leads to an internal combustion engine.”
 · antropology · artificial-intelligence · correlation-causation · machine-learning · project-hva · recruitment · statistics · thick-description

www.eff.org > Jillian C. York and Svea Windwehr
One Database to Rule Them All: The Invisible Content Cartel that Undermines the Freedom of Expression Online
27 aug. 2020 - Every year, millions of images, videos and posts that allegedly contain terrorist or violent extremist content are removed from social media platforms like YouTube, Facebook, or Twitter. A key force behind these takedowns is the Global Internet Forum to Counter Terrorism (GIFCT), an industry-led initiative that seeks to “prevent terrorists and violent extremists from exploiting digital platforms.” And unfortunately, GIFCT has the potential to have a massive (and disproportionate) negative impact on the freedom of expression of certain communities.
 · anti-terrorism · censorship · content-moderation · false-positives · freedom-of-expression · gifct · islamophobia · machine-learning · youtube

www.infoq.com > Vivian Hu
The First Wave of GPT-3 Enabled Applications Offer a Preview of Our AI Future
12 aug. 2020 - The first wave of GPT-3 powered applications are emerging. After priming of only a few examples, GPT-3 could write essays, answer questions, and even generate computer code! Furthermore, GPT-3 can perform algebraic calculations and language translations despite never being taught such concepts. However, GPT-3 is a black box with unpredictable outcomes. Developers must use it responsively.
 · artificial-intelligence · data-science · gpt-3 · machine-learning · not-read