LL38

 This week I took a course on AI. I learned about how to collect and refine data and how to train a neural network. I also learned about newtons fractal, which uses newtons method for finding roots on the complex plane and coloring each point based on what root it lands up on.

LL36

 I watched a video on the physics of car crashes, it was very interesting. The video talked about the many ways car manufatres protect the people in the car. The goal for the car manufacturer is to make the frontal area crush, but the drivers box stay intact, when the frontal area is crumpled it slows the driver down a lot slower than it would if the frontal area was rigid. 

LL35

 This week I took the ARML math test, I got 2/10 which is decent I guess. One of the problems I got right was a combinatorics problem that no one else got right.

LL 34

 This week I wrote an incomplete essay on the technological singularity, here it is:

The technological singularity is the idea that if humans could build an AI smarter than themselves,

it would be a better designer and hence be able to make an even more intelligent AI. This process would

then continue and possibly produce an infinitely intelligent being.

It would be nearly impossible to predict what life would be like for the average human if a superintelligent

being existed; however if the AI is motivated to create better and better versions of itself, it may try

to take over the world and use the world's resources for more computing power.

A self improving computer does not however necessarily produce a superintelligent being. If, for example,

it becomes harder and harder for the computer to improve itself it may take an ever increasing amount

of time for the computer to double in intelligence.

If the computer takes a fairly consistent amount of time to double in intelligence the graph

would look more like this.

While this would lead to exponential growth it does not approach infinity, if however it takes less and less time to double in intelligence it will in theory reach an infinite amount of intelligence in a finite amount of time.

LL 33

 This week I learned about AI and chat GPT, it uses an algorithm called the backpropogation algorithm to try to build what's called a neural network. The neural network tries to figure out what word it should do next.

https://www.3blue1brown.com/topics/neural-networks

I also found a cool AI based guessing game called akinator. It tries to guess the person you're thinking of, it's been able to guess all of mine. The most guesses it took for me was 50 for Terrence Tao, I haven't won any games yet.

https://en.akinator.com/

LL 32

 This week I learned about tetration, which is repeated exponentiation, wich is repeated multiplication, which is repeated addition. Repeated tetration is called pentation and repeated pentation is called hextation. tetration is so hard to compute that for all we know pi terated to itself could be a whole number.

I also found a cool website called neal.fun, it's got really cool educational games. This is one of them, it's called the size of space.

https://neal.fun/size-of-space/