Hi All,
My question to everyone here is, do we need to develop chatbots to help people with various topics like finance, mind therapy, trading knowledge, website development, and more?
We already have huge models like ChatGPT, BRET, GEMINI, CLAUDE, and new models coming daily. Anyone willing to use our developed chatbot can just visit these big models and get their work done.
Then why should we spend time to work on these projects and develop chatbots? Should I spend my time on these?
IMHO, it depends on the problem you are looking at and wish to solve. Part of Deep Learning espoused here (and the instruction is great !), kind of presumes we have like, a million+ data points. Many, but not all, small companies have that.
Developing a NN would be like, ‘overkill’ for the problem at hand. However, I still find it useful that one should study the basics of the models, if nothing else to understand how we got here.
And then maybe you or someone else can design a model that is even better ?
While large models are powerful, there is significant value in developing custom chatbots for specific applications. It’s all about identifying the gaps in existing solutions and focusing on areas where specialized, fine-tuned models can make a meaningful impact. I think it’s sparking our creativity.
@lukmanaj thanks for the response, i have been into data science since many years but mostly I have worked on python, sql, pyspark these kind of tools.
Could you please tell me the tech stack that I should learn to develop such apps/bots?
I am eager to learn these tools and develop an actual working bot
@ankurjain he was highlighted as one of Prof Ng’s interviews and one of his students, so I like AK’s Neural Networks: Zero To Hero.
But I would say be careful if you go over there, it gets ‘a little wild’ sometimes. Whereas we are all respectful here.