
Sign up to save your podcasts
Or


[16/20] Now that we can create logic gates out of proteins, how can we read them?
[15/20] Now that we can design proteins, why not see what happens when you turn a protein i to a logic gates.
HALLUCINATION CHECK: The Coffee-Bots struggle with acronyms and other notation sometimes. They struggle greatly with the names of logic gates in this episode, but they do so in a way that we can still understand. It just sounds SUPER AWKWARD when they do it in rapid succession. Let us know on our Reddit page if you can't understand it.
[14/20] We look at the implications of building enzymes for unique applications.
[13/20] In this episode we examine de novo protein design and how we can make these machines work for us.
[12/20] Some of the difficulties in the road ahead have already been solved by natural evolution!
[11/20] Computational protein construction might not be as easy as you think.
[10/20] We discuss the massive breakthrough that AlphaFold represents, how it works, and what it means for the future of the field.
[9/20] Now that we know the physical and biological mechanics of proteins, we begin to examine them from a new angle: as computational engines!
[8/20] We explore how the proteins topology is the constraint that governs behavior.
[7/20] This is how proteins and Information Theory collide. And here we introduce the AI marvel that is AlphaFold. Who needs a refresher on Shannon Entropy?
From the publisher's feed
Like NPR for mad scientists. We compile and curate the most complex and critically important data, break it into classes, and give you a full, unambiguous exploration on the topic of the day. These…
Salvation AI features "The Coffee-Bots" discussing humanity's most critical issues from the "Source Material" (triple-fact-checked and meticulously curated documents created by our research partners at Aion-Sigma Correlated Publishing). Join our Deep Dive fact-bots as they explore human cognition, technology, politics, science, and more. They are not here to provide you all the answers. They are here to make you think.
Consider each of these classes an invitation to begin your own journey of universal discovery. Question what you hear. We do not claim expertise in these subjects. Our knowledge stems from a wide assortment of interests that we have explored as lay-people. Because we do use AI hosts, we include "Hallucination Checks" at the bottom of episode notes as necessary, to make sure we have a consistent mode of correction (luckily, due to our process there are very few (and very small) factual errors, but there are more than a few verbal misfires and poorly chosen phrases in the mix). If something sounds off, look it up and let us know. We are happy to correct bad data. These episodes are here to provide listeners with a unique perspective on unique topics. Don't expect to gain encyclopedic understanding just from listening to these sessions.
Shout out to NotebookLM for use of their hosts. We are good at research and writing, not being talk show hosts. If it wasn't for the robot hosts giving a little life to these boring book nerds, you'd probably have to read all this yourself. The substance of this project is human. The face, for your benefit, is not.
To the people that think all AI involved content is inherently "Slop", we gently suggest that not all content is created equally, and we offer this as a good faith example.
To the people that take issue with the word "Salvation" in our title: No we are not religious. No we don't think that AI is a miracle savior. And yes, it was chosen to be provocative. Low-key, it was just the most memorable sounding name pitched when we decided to put this stuff up (Which we only did in case anyone else out there had any interest. This isn't exactly the kinda thing you do as a cash grab). If the name of the podcast is too loaded for you to handle, you won't be able to manage the contents.
If you can manage the name and the robot messengers, you might learn something. If you can't, there's always Joe Rogan.