KevinSimper

It is funny how every AI company has copied ChatGPT interface (Claude, Deepseek, Mistral, all history on the left, a small text box in the middle and a conversation interface once you begin writing. It is boring.

Here is what I would suggest doing better:

Favorite conversation

Make it possible to favorit certain conversations, some conversations you want to come back to.

Cursor and Windsurf has shown that AI can be an incredible enabler. You can quickly ask the AI and add context to your prompts.

Windsurf is new in that it is the most agentic, in that a agent can call functions and follow its own lead.

Ideas to make it better

1. Multi Agent

   I want and should be able to create multiple agents inside my repo. I want to create an expert agent that knows how to edit my prisma schema. I want an agent that is an expert in creating visual components. I don't want to instruct the agent to do that, it should be based on the folder it is trying to edit in, then it should delegate responsibility to that agent and wait for it to finish.

I quit my job recently because I wanted to start something new, jump into the unknown.

It can feel scary to start something new, how is it going to end, what are the next steps. I have been comfortable with not always knowing the way, I have multiple times had times where I worked on my own thing or did something else.

Compute is everywhere and companies like AWS and Google Cloud are some of the most profitable companies in the world and they keep growing. They make it really easy and affordable to get compute.

But getting GPU compute is still new and unsolved, GPU compute is expensive and we are very much tied to a single provider called Nvidia if you want to do something.

The easiest way to get started is definitely your own computer, but most times if not all the time your computer is simply not powerful enough, so many things are not even possible to run locally in relation to machine learning.

What is also easy is using Google Colab, it is a easy web interface that allows you to run Python through what is called a notebook, concretely a Jupyter notebook. It is really good for proving a point and visualising the output of your machine learning or data processing.

Today I want to show how streaming an AI prompt response from a CloudFlare worker. We have been used to know that ChatGPT streams the response of the large language model, it is pretty mesmerizing and adds to the fealing of a thinking mashine that you can see the dialog as it is coming.

Simulating the response

We don't want to call an AI model during development just to simulate the LLM sending the events.

We can work with streams in Node.js by creating a TransformStream, that will allow us to both write and read the stream.

let { readable, writable } = new TransformStream();