Updates:
- After the Hugging Face incident last week, we've moved on to Occupy OpenAI.
- The Pacing the Frontier letter asked for tools to pace AI development, but without governments enforcing a slowdown, this is not enough.
We'll be outside OpenAI (550 Terry A Francois Blvd) every day until Sam Altman (@sama) calls for a binding AI Treaty.
In physics, an observable is a physical quantity that can be measured. Online, observable is a notebook that's split up into blocks and executed continuously. Observables must be named after their ability to be observed. #ProgrammingParadigms@NortheasternCA
according to the founder of @observablehq , compsci is fundamental to communicating, and observable's asynchronous and reactive, making it interpretable in real time #ProgrammingParadigms@NortheasternCA
WebSockets let the client send a single HTTP request to the server to open a connection and then TCP connection is left open, we can build real time features by establishing a 2way connection between a client and server. #ProgrammingParadigms@NortheasternCA
Observable looks like a JS notebook but it has much different features. One difference is that cells are separate scripts and don't run in order. The cells run when their reference values change. Users may import cells across notebooks. #ProgrammingParadigms@NortheasternCA
JS testing package Jest works with react and we can check coverage with yarn jest --coverage. compared with pylint (pytest https://t.co/CvWTEPzlS6) #ProgrammingParadigms@NortheasternCA
a web worker makes running a script in a background thread of a main execution thread of a web app possible. the project https://t.co/s4wkRNDGlW runs a WebAssembly that sketches a user-input genome assembly in the browser #ProgrammingParadigms@NortheasternCA#genomics
React was first released in 2013 and it makes code more elegant because it allows us to manage the complexity of codes by breaking them into components of class fields allow you to work with them separately and at the same time together. #ProgrammingParadigms@NortheasternCA
a typical image recognition convolutional network looks at a lot of images and slowly corrects itself after seeing more examples of images. TensorFlow can receive user input now because the js library works with JS apps
@NortheasternCA#ProgrammingParadigms
TensorFlow js was published in 2018 as a library of math functions that compute webdev. a basic model is based on stochastic gradient. a tensor is an array. @NortheasternCA#ProgrammingParadigms
variables declared with const can’t be changed but they can be updated b/c const creates read-only references to a variable that cannot change. But the variable itself can change @NortheasternCA#ProgrammingParadigms
OOP is not necessarily better than functional but reformatting input as a string can be done both ways and they also differ because functional data is immutable while OOP isn't always. #ProgrammingParadigms@NortheasternCA
Constructor functions start with a capital letter by convention. extension is a form of inheritance that creates a new class from an older one. Extension is a way to reduce redundancy in code. #ProgrammingParadigms@NortheasternCA