Exploring OpenAI’s ChatGPT – Building Apps in Multi-Languages with ChatGPT

ChatGPT is an amazing AI service created by OpenAI. It has ushered us into the next generation of search engines that is powered by LLM (Large Language Models) and NLP systems.

Let us explore this awesome service. First of all you will need to create an account (email and phonenumber as well as the basic username) in order to use this service. This makes sense to collect all these details for security purposes and to avoid abuse.

General Overview

ChatGPT Overview

Testing ChatGPT to Design a Streamlit App

Let us actually try out the code given by ChatGPT

import streamlit as st
import numpy as np
import matplotlib.pyplot as plt

st.title("Streamlit App")

# Create a slider
slider = st.slider("Select a value", 0, 100)

# Create a plot based on the value of the slider
x = np.linspace(0, 10, 100)
y = x**2

fig, ax = plt.subplots()
ax.plot(x, y)


After following the instructions on how to run it above

This is quite impressive.

Building 5 Apps with ChatGPT

Let us give it a bit harder but useful task – A Sentiment Analysis App. We will first start with streamlit, then in flask, nodejs and then in a more difficult option such as Carbon language or Julia.

In running the code generated by ChatGPT, it actually works.

We will then repeat the same in Flask and Julia

ChatGPT even gave me the html code for the frontend

Let us try it in Express or Nodejs

ChatGPT on a more recent programming language

Let us try it on Carbon language

For those who are familiar with Carbon here is the code as copy and paste. Let us know if it actually works.

import "textblob"

fn main() {
  form {
    textarea name="text"
    submit "Analyze"
  let text = form.text
  let analysis = TextBlob(text)
  let sentiment = analysis.sentiment.polarity
  if sentiment > 0 {
    result = "Positive"
  } else if sentiment == 0 {
    result = "Neutral"
  } else {
    result = "Negative"
  p "Sentiment: " + result + " (" + sentiment + ")"

Finally we will try it with Julia

using Flux
using MLJ
using MLJModels
using MLJSentimentAnalysis
using TextAnalysis

function sentiment_analysis_app()
    form = @form(:post, "/") do

    if form.submitted
        text = form.params.text
        sentiment_model = SentimentAnalysis.SentimentAnalysis()
        sentiment = sentiment_model(text)
        if sentiment > 0
            result = "Positive"
        elseif sentiment == 0
            result = "Neutral"
            result = "Negative"
        p("Sentiment: $result ($sentiment)")

using Flux.Tracker: @epochs


@epochs 1 Flux.train!(sentiment_analysis_app, Flux.ADAM(), loss)

To stretch the limits of ChatGPT, I asked it to compile each source code into a downloadable zip file and it did, however the links were not functionally which makes sense, since it is a free service.

We have seen how ChatGPT can be used to do some pretty incredible things. With great power comes great responsibility.

Happy ChatGPTing

Jesus Saves

Jesse E.Agbe(JCharis)

Leave a Comment

Your email address will not be published. Required fields are marked *