background preloader

Issues

Facebook Twitter

Notes on AI Bias. Machine learning is one of the most important fundamental trends in tech today, and it’s one of the main ways that tech will change things in the broader world in the next decade.

Notes on AI Bias

As part of this, there are aspects to machine learning that cause concern - its potential impact on employment, for example, and its use for purposes that we might consider unethical, such as new capabilities it might give to oppressive governments. Another, and the topic of this post, is the problem of AI bias. It’s not simple. What is ’AI Bias’? “Raw data is both an oxymoron and a bad idea; to the contrary, data should be cooked, with care.”

-Geoffrey Bowker Until about 2013, If you wanted to make a software system that could, say, recognise a cat in a photo, you would write logical steps. With machine learning, we don’t use hand-written rules to recognise X or Y. However, there’s a catch. What does that mean in practice? First, the consequences. AI bias scenarios It gets worse. AI bias management Conclusion. IA responsable - Impact AI. Will a robot take your job? The Revenge of Neurons. Prediction Machines: The Simple Economics of Artificial Intelligence. KF: Escaping the Local Minimum. Escaping the Local Minimum Where AI has been and where it needs to go This report is my final project for the MIT Media Lab Class "Integrative Theories of Mind and Cognition" (also known as Future of AI, and New Destinations in Artificial Intelligence) in Spring 2016.

KF: Escaping the Local Minimum

The paper is available here as a pdf Abstract Artificial Intelligence performs gradient descent. Overview I begin this paper by pointing out a concerning pattern in the field of AI and describing how it can be useful to model the field's behavior. In the first section of this paper, I argue that the field of artificial intelligence, itself, has been performing gradient descent. In the second section, I describe steps that should be taken to prevent the current trends from falling into a local minimum. Finally, I summarize my findings and conclude by reiterating the use of the gradient descent model. Introduction Obviously, any sufficient pattern matching algorithm would predict an incoming set of "bad" points. .

Gradient Descent. AI, Apple and Google. (Note - for a good introduction to the history and current state of AI, see my colleague Frank Chen’s presentation here.)

AI, Apple and Google

In the last couple of years, magic started happening in AI. Techniques started working, or started working much better, and new techniques have appeared, especially around machine learning ('ML'), and when those were applied to some long-standing and important use cases we started getting dramatically better results. For example, the error rates for image recognition, speech recognition and natural language processing have collapsed to close to human rates, at least on some measurements. There are really two things going on here - you’re using voice to fill in a dialogue box for a query, and that dialogue box can run queries that might not have been possible before. Both of these are enabled by machine learning, but they’re built quite separately, and indeed the most interesting part is not the voice but the query.

This is clearly a fundamental change for Google. Andrew Ng: Artificial Intelligence is the New Electricity. What’s Next for Artificial Intelligence. A DARPA Perspective on Artificial Intelligence. AI Principles - Future of Life Institute. AI/Robotics Researchers: Ethique et numérique. DP lancement strategie IA. The Administration’s Report on the Future of Artificial Intelligence. Under President Obama’s leadership, America continues to be the world’s most innovative country, with the greatest potential to develop the industries of the future and harness science and technology to help address important challenges.

The Administration’s Report on the Future of Artificial Intelligence

Over the past 8 years, President Obama has relentlessly focused on building U.S. capacity in science and technology. Public Input and Next Steps on the Future of Artificial Intelligence. Public Workshops on Artificial Intelligence The RFI follows an announcement by OSTP of a series of new actions as a part of a White House Future of Artificial Intelligence initiative.

Public Input and Next Steps on the Future of Artificial Intelligence

We held a series of events co-hosted with academic institutions and non-profit organizations which took place over a period of several months across the country, including four topical public workshops and an event hosted in conjunction with the Global Entrepreneurship Summit. In total, the workshops had more than 2,000 in-person participants, and many thousands of viewers tuned in online from across the world to watch the event livestreams in real time. Videos of the AI workshops are available for streaming: Law and Governance Workshop, May 24 in Seattle, WA — watch here: AI for Social Good Workshop, June 7 in Washington, DC — watch here: The full workshop can be viewed here.

Emerging Topics and Societal Benefit Workshop, June 23 in Palo Alto, CA — watch here: The whole workshop can be viewed here. National ai rd strategic plan. Preparing for the future of ai. MHC AI and AWS FINAL. Artificial Intelligence News - Future of Life Institute.