Salvation AI

Tech Lobbyists are paying billions to keep regulations off the books, essentially using their raw cash flow to prevent government from advocating for users. So why are they paying out the nose to keep the internet broken? Because it is easier to take advantage of users in a dark, chaotic, and seemingly incomprehensible system. Oversight and regulation would force responsibility, transparency, and...

To discover solutions for "hallucination", we have to identify the type of error to determine the cause. If you can isolate and eliminate these causes, you can prevent or correct hallucination in real time.

The machine has detectors and responsive mechanics that then tell the algorithms what your deepest preferences are and then change what you see to exploit these habits with relentless precision. Accidentally letting a video play through can cause a permanent shift in the media you will see for the rest of your life.

We have followed information from conception, through the house of mirrors, and onto the screen of your phone. Now we approach the most fragile equipment in the machine: YOUR MIND. No two humans will receive that data the same way. If the best data makes it to a human mind that has been trained to disbelieve a reality it doesn't want to see, no amount of good programming can keep them from assumin...

The Interface layer is how all of this chopped and screwed semantic soup is presented to us. Let's take a look at how the face of information impacts the way we think about it.

We have seen how information goes from the human sources to the AI training pipeline, now we look at the "lossy compression" blender that breaks up all of that knowledge into the semantic soup of global information that LLMs draw from.

Now that the data has been created, uploaded and parsed, we enter the opaque blender of AI compression and ambiguation.

The Aggregation Layer: How information is aggregated and gatekept, and what you are allowed to see.

This is layer 2 of the machine: Capture -How information Get's Digitized. Now we turn our attention to how the information created at the Origin Layer enters the labyrinth of the larger ecosystem.

We examine how the first layer of the information machine originates knowledge. We look at the incentives, limitation, and biases that dictate how knowledge is created.

We take a look at the machine in action by looking at the 4 types of information flow, and how they are manipulated.

There are fundamental realities that force the machine to function the way it does. Join us as we consider the reasons the current architecture is locked in.

We now look inside the machine to begin our dissection of the human knowledge machine that is the internet. You will never see the web, or your "feed", the same way again!

This is our most critical work to date (and we literally just did a 20 episode run on turning proteins into computers). In this massive diagnostic omnibus, we fully dissect the internet piece by piece and examine each for fractures. And there are more than you could imagine. Join us as we venture down the most complex subject that will likely ever exist: how does the sum total of human knowledge f...

We are finally taking on the uncomfortable topic of teaching a lesson on... ourselves. Join the "Coffee-Bots" as they explain the human element behind the robot mask. Us. We are actually people. Just shy and busy people that want to spend our time on research. Not recording podcasts. Join us for a look under the hood at the process that produces this podcast and the books it is based on. We finish...

[20/20] HEY YOU! JOIN THE REDDIT GROUP! IF YOU LISTENED TO THIS WHOLE THING, WE SHOULD CHAT! COME TALK ON REDDIT! In this final exploration on the science of protien folding, we touch again on its past, its present, and its future.

[19/20] By considering the evolutionary history of proteins from abiogenesis to today, we consider potential futures for the field.

[18/20] We have explored the possibilities of how we can build the pieces. Now let's put the whole machine together.

[17/20] Now that we can perform computation with proteins, how can we use them to store information?

[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...

[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.