New ‘Liquid-AI’ Learns Continuously from Its Experience

Posted by Peter Rudin on 19. February 2021 in News No Comments

While most machine learning algorithms cannot hone their skills beyond an initial training period, a  new approach, called a liquid neural network, has a kind of built-in “neuroplasticity.”

The algorithm’s architecture was inspired by the mere 302 neurons making up the nervous system of C. elegans, a tiny nematode (or worm).

At a time when big players like OpenAI and Google are regularly making headlines with gargantuan machine learning algorithms, it is a fascinating example of an alternative approach headed in the opposite direction.

In contrast, in a liquid neural network, the parameters are allowed to continue changing over time and with experience. The AI learns on the job.

Read More

Is Personality Residing in the Brain?

Posted by Peter Rudin on 5. February 2021 in News No Comments

A team of Caltech researchers from the disciplines of neuroscience, psychology and philosophy discuss the long-standing question: What is personality? Most studies measure personality in various ways, and they are often ambiguous about what personality really is.

The researchers believe that genes and environment are causes of personality and that behaviour results from personality, but personality itself is located in the brain.

The researchers propose a method to discover where personality resides in the brain, and how it relates to other psychological functions including memory and emotion and they propose ways of testing those models using the tools of neuroscience.

Read More

AI to Mimic the Brain’s Prefrontal Cortex for Learning

Posted by Peter Rudin on 22. January 2021 in News No Comments

The biological brain has served as an inspiration for AI-machine-learning designs, such as artificial neural networks used for deep learning. Now AI is being deployed as a tool to help unravel how the brain works.

For example, how and why is the brain capable of adaptive lifelong learning? The precise neural mechanisms on how the brain achieves this has not been entirely clear.

In a recent neuroscience study, researchers from the Salk Institute and the University of Massachusetts Amherst created AI machine-learning model to provide new insights on how the brain’s prefrontal cortex operates when it comes to lifelong learning.

Read More

How to Make AI More Democratic

Posted by Peter Rudin on 8. January 2021 in News No Comments

A new type of learning model uses far less data than conventional AIs, allowing researchers with limited resources to contribute. One such recent advance is called “less than one”–shot learning (LO-shot learning), developed by Ilia Sucholutsky and Matthias Schonlau from the University of Waterloo.

Allowing AIs to learn with less plentiful data helps to democratize the field of artificial intelligence. Not only does LO-shot learning make the barriers to entry lower by reducing training costs and lowering data requirements, but it also provides more flexibility for users to create novel data sets.

By reducing the time spent on data and architecture engineering, researchers looking to leverage AI can spend more time focusing on the practical problems they are aiming to solve.

Read More

Beethoven and the Power of Joy for Christmas

Posted by Peter Rudin on 25. December 2020 in News No Comments

Beethoven’s music is the sound of human freedom at its core – the freedom of our minds. Unsurprisingly, a composer who could capture the very essence of human freedom in sound would himself come to be cast in the image of liberating music from convention.

Today, Beethoven’s music remains deeply connected with a true humanism, which has the principles of freedom and self-determination at its heart. The composer’s music grew out of the age of European Enlightenment, which located human reason and the self at the centre of knowledge.

And as we are at the threshold of exploring and eventually mastering Artificial Intelligence, reaching new frontiers of knowledge, we must uphold the value and joy of our mind of freedom.

Read More

Crazy Dreams make Sense of our Memories

Posted by Peter Rudin on 11. December 2020 in News No Comments

The leading theory about what we are doing when we dream is that we are sorting through our experiences of the last day or so, consolidating some experiences into memories for long-term storage, and discarding the rest. That does not explain, though, why our dreams are so often so exquisitely weird.

A new theory, the Overfitting Brain Hypothesis (OBH), proposed by Erik Hoel of Tufts University, suggests  that perhaps the brain’s sleeping analysis of experiences is akin to machine learning.

The goal of machine learning is to supply an algorithm with a data set, a “training set,” in which patterns can be recognized and from which predictions that apply to other unseen data sets can be derived.

Read More

Synthetic Biology: The Xenobot Future Is Coming

Posted by Peter Rudin on 27. November 2020 in News No Comments

A new field of science called “synthetic biology” aims to digitize genetic manipulations. Xenobots are living robots, made up of masses of cells working in coordination, that can help unlock the mysteries of cellular communication.

These living robots can undulate, swim, and walk. They work collaboratively and can even self-heal. They are tiny enough to be injected into human bodies, travel around, and—maybe someday—deliver targeted medicines.

Soon, how we reproduce, repair ourselves, prevent disease will be the result of intentional choice, not chance. But the ability to recode cells, de-extinct species, and create new life forms will come with ethical, philosophical, and political challenges.

Read More

AI trying to Diagnose Depression from Brain Waves

Posted by Peter Rudin on 13. November 2020 in News No Comments

An estimated 17.3 million adults in the U.S. have had at least one major depressive episode, according to the U.S. National Institutes of Health. But with 1,000 possible symptom combinations, depression manifests differently in different people. Today’s assessments mostly rely on conversations with clinicians or surveys.

X, Alphabet’s experimental R&D lab, recently detailed Project Amber, which aimed to make brain waves as easy to interpret as blood glucose. The Amber team sought to marry machine learning techniques with electroencephalography (EEG) to measure telling electrical activity in the brain.

It took three years for the Amber team to create a low-cost, portable, research-grade system designed to make it easier to collect EEG data.

Read More

The Evolution of Modern Intelligence

Posted by Peter Rudin on 30. October 2020 in News No Comments

For about 300,000 years after Homo sapiens first appeared, tools and artifacts remained surprisingly simple. Starting 50,000 to 65,000 years ago, more advanced technology started appearing: complex projectile weapons, sewing needles, ceramics etc.

This sudden flourishing of technology is called the “great leap forward” supposedly reflecting the

evolution of a fully modern human brain. But fossils and DNA suggest that human intelligence became modern far earlier.

Culture can evolve even if intelligence does not. Humans in ancient times lacked smartphones and spaceflight, but we know from studying philosophers such as Buddha and Aristotle that they were just as clever. Our brains did not change, our culture did.

Read More

AI has Limits because it lacks a Physical Body

Posted by Peter Rudin on 16. October 2020 in News No Comments

We are witnessing the emergence of a new commercial industry with tremendous potential. There are no areas that are beyond improvement by AI – no tasks that cannot be automated, no problems that cannot at least be helped by an AI application.

But is this strictly true? Research in the new field of developmental robotics is now exploring how robots can learn from scratch, like infants. The first stages involve discovering the properties of passive objects and the “physics” of the robot’s world.

So, while disembodied AI definitely has a fundamental limitation, future research with robot bodies may one day help create lasting, empathetic, social interactions between AI and humans.

Read More