Peering Into the Mind of a Machine with Sam Friedman

<p>Dr. Fred and co-host Sam Morris sit down for a mind-blowing conversation with Sam Friedman, an “artist and scientist working with electricity, trying to find a starry synthesis of the mysterious speedy electron and the soft slow human.”</p> <p>A Machine Learning Scientist in the Data Sciences platform at the Broad Institute of MIT and Harvard, Sam is keenly interested in how machine learning can help advance our understanding of cardiovascular disease and make an impact in the clinic. As tech lead for the Machine Learning for Health group, his days are spent developing model architectures, interpreting learned features and trying to fuse multiple modalities into meaningful representations.  </p> <p>If you haven’t explored machine learning before, prepare to be astounded. Here are some of the topics Dr. Fred and Sam Morris explore with today’s guest:</p> <ul><li style="font-weight:400;">How Sam’s artist background prepared him to behold the beauty of machine learning</li> <li style="font-weight:400;">The difference between machine learning and AI</li> <li style="font-weight:400;">Risks posed by bias in working with neural nets</li> <li style="font-weight:400;">What can humans learn from machines?</li> <li style="font-weight:400;">Can machines become self-aware?</li> <li style="font-weight:400;">How machines can have “beautiful thoughts and experiences”</li> <li style="font-weight:400;">What it means to be “carbon chauvinistic”</li> <li style="font-weight:400;">How do neural nets dream?</li> <li style="font-weight:400;">The incredible intersection of machine learning and psychedelics</li> <li style="font-weight:400;">Interoception and exteroception vs. ego dissolution</li> <li style="font-weight:400;">Where does the experience of ego dissolution map in the brain? </li> </ul> <p> </p> <p>Episode Length: 00:52:05</p> <p> </p> <p>SAM FRIEDMAN’S RESOURCES</p> <p>Twitter > <a href='https://twitter.com/lucidtronix'>https://twitter.com/lucidtronix</a></p> <p>GitHub > <a href='https://github.com/lucid