Follow us on Twitter
Question, remark ? Contact us at contact@myengineeringtools.com
2023 has seen the development of new AI models that are now accessible to most of the public. 3 large players that have entered this field are ChatGPT (by OpenAI), Google Bard and Microsoft Bing Chat (enhanced with ChatGPT AI). Which of these tools is the best ? We are proposing here a review that can help you decide when to use which tool.
Top 5 Most
Popular
1. Compressor
Power Calculation
2. Pump Power Calculation
3. Pipe Pressure
Drop Calculation
4. Fluid Velocity in pipes
5. Churchill Correlation
(friction factor)
Those tools are free, or have a free version but remember that those company will collect all your data, so you must not enter anything private, at least using the free plans.
Let's start by asking all of these AI to make a Python program, from the following description :
Prompt : "Act as a programmer and financial analysis. Please write a Python program to pull stock data from the GAFAM and calculate the return of each of them. The program must look for data over a period that will be programed as variable"
ChatGPT is replying by giving a Python program. The code is nicely formatted and can be copied directly. ChatGPT is also giving explanation about the programing to be able to understand the code and improve it.
We have tested this code to verify it does what we want. If we analyse it, we actually see it is programmed to pull data from Yahoo, calculate the cumulative return, and actually show them on a graph (kind of a bonus)
It works, it does not crash and is indeed showing a graph of cumulative return over the period specified in the program.
It is not exactly what we want be it is a very good workable basis !
Let's see the reply of Bard :
Ok, that's a piece of code (Bard did not provide much other details), let's see if it works, we are using Python Portable with Pyscripter to test :
Hey Google, it does not work... The issue here is the usage of the reserved instruction return !
Actually, don't stop her with Bard. The AI from Google has indeed the interesting function to be able to generate different drafts you can select :
So let's see if the 2nd draft is better ?
Actually it did not work again, the coding is quite complex.
Then the 3rd draft ?
Yes ! This one is at least working, it is calculating daily returns.
We were not expecting Bing to give a reply on this as it is more marketed as an improved search engine, but yes, Bing gave a piece of code, with some reasonably helpful information to continue coding.
So we have tested the code, once again with Python Portable and Pyscripter.
Hum, it did not work. Actually we suspected that the issue with was the pandas_datareader library which can be tricky. The beauty of the AI model is that we can "naturally" converse with them, so we asked Bing to adapt the code using yfinance instead.
And Bing, it worked.
It correctly gave the cumulative return for each stock in portfolio.
1. | Chat GPT |
2. | Bing Chat with AI |
3. | Google Bard |