AI News, Massachusetts Institute of Technology artificial intelligence

MIT finally gives a name to the sum of all AI fears

Now we know what to call it, that vast, disturbing collection of worries about artificial intelligence and the myriad of threats we imagine, from machine bias to lost jobs to Terminator-like robots: 'Machine behaviour.'

While there are no conclusions here about any of this, it's a nice, ambitious effort to give some direction to studying AI's role in society rather than just worrying about it.  Also: AI pioneer Sejnowski says it's all about the gradient Published in the journal Nature this week, the paper, Machine Behaviour, calls for a joint effort of 'the fields that design and engineer AI systems and the fields that traditionally use scientific methods to study the behaviour of biological agents.'

Specifically, the authors propose to study not just how machine learning algorithms works, but how they are affected by, and affect, the environment in which they function.  It's 'akin to how ethology and behavioural ecology study animal behaviour by integrating physiology and biochemistry -- intrinsic properties -- with the study of ecology and evolution -- properties shaped by the environment.'  Questions large and small MIT proposes asking about A.I.

The authors are really exploring here the problem of the 'objective function' in machine learning, namely, what exactly these algorithms are supposed to be achieving.  The fourth aspect, evolution, isn't as simple as you might imagine from the name: There's several aspects, including the propensity for assumptions by creators of neural nets to promote certain kinds of algorithms versus others, but also the prospect of 'mutations' propagating in unexpected ways.

And the choice to try and observe algorithms in an empirical fashion in the wild brings its own set of ethical concerns.  Even the very term 'agent,' which they use repeatedly to refer to AI innovations, brings all kinds of problematic assumptions of human and animal parallels, they acknowledge.  'Even if borrowing existing behavioural scientific methods can prove useful for the study of machines, machines may exhibit forms of intelligence and behaviour that are qualitatively different—even alien—from those seen in biological agents

Marvin Minsky

Marvin Lee Minsky (August 9, 1927 – January 24, 2016) was an American cognitive scientist concerned largely with research of artificial intelligence (AI), co-founder of the Massachusetts Institute of Technology's AI laboratory, and author of several texts concerning AI and philosophy.[12][13][14][15]

At that point in time, it was known to be the simplest universal Turing machine–a record that stood for approximately 40 years until Stephen Wolfram published a 2,5 universal Turing machine in his 2002 book, A New Kind of Science.[23]

This book is the center of a controversy in the history of AI, as some claim it to have had great importance in discouraging research of neural networks in the 1970s, and contributing to the so-called 'AI winter'.[24]

Minsky says that the biggest source of ideas about the theory came from his work in trying to create a machine that uses a robotic arm, a video camera, and a computer to build with children's blocks.

In November 2006, Minsky published The Emotion Machine, a book that critiques many popular theories of how human minds work and suggests alternative theories, often replacing simple ideas with more complex ones.

Clarke's derivative novel of the same name, where he is portrayed as achieving a crucial break-through in artificial intelligence in the then-future 1980s, paving the way for HAL 9000 in the early 21st century:

and argued that a fundamental difference between humans and machines was that while humans are machines, they are machines in which intelligence emerges from the interplay of the many unintelligent but semi-autonomous agents that comprise the brain.[36]

He cautioned that an artificial superintelligence designed to solve an innocuous mathematical problem might decide to assume control of Earth's resources to build supercomputers to help achieve its goal,[38]

In 2006, he was inducted as a Fellow of the Computer History Museum 'for co-founding the field of artificial intelligence, creating early neural networks and robots, and developing theories of human and machine cognition.'[45]

Artificial intelligence meets neuroscience at MIT

Live-broadcast at the day of the event: Time: 9:30 – 10:00 Talk: Artificial intelligence and its relationship to increase our understanding of the brain Speaker: Dr. Omar Costilla Reyes – Miller Lab, Brain and Cognitive Sciences, MIT Abstract: During the past few years, we have experienced an impressive explosion in the development and deployment of artificial intelligence systems to solve tasks such as image recognition and autonomous driving.

Time: 10:00 – 10:30 Talk: Steady As She Knows: Invariant Representations of Facial Emotion and Identity Speaker:Kathryn C O’Nell – Saxe Lab, Brain and Cognitive Sciences, MIT Abstract: Every day, your brain works constantly to help you perform the tasks vital to your life.

You can, with relative ease, recognize whether someone else is happy or sad based on their facial expression, regardless of their identity or what angle from which you’re viewing their face.

Sensory feedforward pathways carry information up the brain's processing chain and internal information is thought to be carried by feedback pathways.

Time: 11:30 – 12:00 Talk: Using AI to understand how brain regions “talk” to each other Speaker: Mengting Fang, Anzellotti Lab, Psychology department, Boston College Abstract: Most cognitive tasks are not completed by a single brain region, but by many regions working together.

I will talk about how artificial neural networks can help us understand the “language” with which brain regions communicate with each other, and show examples of how brain regions that respond to faces and scenes interact with the rest of the brain (while participants were watching Forrest Gump!).

Joy Buolamwini

She founded the Algorithmic Justice League, an organisation that looks to challenge bias in decision making software.[3]

In 2011, she teamed with the Trachoma program at Carter Center, to develop an Android-based assessment system for Ethiopia and aide eradication of the disease worldwide.[15][6]

As a Fulbright fellow, in 2013 Buolamwini worked with local computer scientists in Zambia to empower Zambian youth to become technology creators.[16]

During her research, Buolamwini showed facial recognition systems 1,000 faces and asked them to identify whether faces were male or female, she found that software found it hard to identify dark-skinned women.[22][23]

Her program, Algorithmic Justice League, aims to highlight the bias in code that can lead to discrimination of underrepresented groups.[35]

In 2017, Buolamwini was awarded the grand prize in the professional category in the Search for Hidden Figures contest, tied to the release of the film Hidden Figures in December 2016.[38]

The contest, sponsored by PepsiCo and 21st Century Fox, was intended to 'help uncover the next generation of female leaders in science, technology, engineering and math,' [39][40]

1. Introduction and Scope

MIT 6.034 Artificial Intelligence, Fall 2010 View the complete course: Instructor: Patrick Winston In this lecture, Prof. Winston ..

MIT reshapes itself to shape the future

MIT has announced a $1 billion commitment to address the global opportunities and challenges presented by the prevalence of computing and the rise of ...

Marc Raibert: Boston Dynamics | MIT Artificial Intelligence (AI) Podcast

This is a talk by Marc Raibert for course 6.S099: Artificial General Intelligence. He is the CEO of Boston Dynamics. This class is free and open to everyone.

What AI have MIT been creating? - BBC Click

Click heads to MIT's CSAIL to check out its weird and wonderful robotic research, and visits Wimbledon to see how AI is set to transform the tennis tournament.

Vision-free MIT Cheetah

MIT's Cheetah 3 robot can now leap and gallop across rough terrain, climb a staircase littered with debris, and quickly recover its balance when suddenly ...

MIT Investing $1 Billion in Artificial Intelligence, Computing

The Massachusetts Institute of Technology is heavily investing to study computing and AI technology. The school announced its bold plan to reshape its ...

Building Machines that Learn & Think Like People - Prof. Josh Tenenbaum ICML2018

Recorded July 13th, 2018 at the 2018 International Conference on Machine Learning Joshua Tenenbaum is Professor of Cognitive Science and Computation at ...