Maximizing productivity with ChatGPT & the next generation of AI/ML Tools

When I first heard about ChatGPT, I was intrigued. AI/ML has been around for a very very long time. I’ve spent a fair amount of time working with data, models & sentiment analytics in school and professionally before I moved into product management. The world has seen countless chat bots & AI assistants, so I was a little skeptical that this was just another buzz word or trend. ‘What are they doing differently this time?’, I wondered. My feed has been flooded with countless articles about how ChatGPT is going to replace everyone’s jobs. So I decided to experiment with it myself and see if it lived up to the hype.

What ChatGPT is NOT

Let’s start with busting some myths. ChatGPT, in its current form, is not going to completely replace human jobs that require critical thinking. Like most AI/ML models & tools, there’s really no effective way of handling complex problem solving - There’s an incredible amount of bias that the model is prone to, the model lacks context on what is ethical or culturally appropriate, & without defined intent, context or a pre-defined ‘standard’, the tool will not be able to function autonomously. We still need to define what this training data should be, define ‘ground truths’, and that by nature, means there is subjectivity involved in the process. ‘Who decides what fact is and what fiction is?’, ‘Who do we give the power to decide what’s truth and what is not?’, ‘What’s the basis of deciding this power?’ As you can imagine, the dangers of misinformation are plenty. Even the ChatGPT website talks about how the training data is only till year 2021, which makes a lot of the ‘facts’ it talks about outdated.

But here is where it gets interesting.

ChatGPT is an excellent interface when it comes to subjective or creative outcomes.

When answers are not so black and white, when possibilities are many, when choices are several, when there’s no clear ‘right way’ to do something, it absolutely is a fantastic tool to use. I’m excited about this development, because finally we’re opening up new possibilities when it comes to creative fields.

I tried out a few use cases of this myself. I asked the tool for ideas on a new book.

When I wasn’t happy with the outcome, I shared more context about my situation and skills, and I was pleasantly surprised with the results. What would normally take me hours of research & then brainstorming, was presented in a summary for me to then take action upon.

The chat also continued to remember context from earlier in the conversation & continuously refined the answers to match my preferences. As someone who is constantly looking for ways to be more productive, I’m definitely going to be trying out this tool for more ideas & places where I do not need to rely on factual information. I definitely appreciate that the tool remembers past context, which other AI tools simply lack today & the conversational interface is what has made the tool stand out.

There have been several examples of users using ChatGPT for creative pursuits over the past few days. Some have even used the tool to generate 100s of line of python code and build things with this.

It’s clear that the possibilities of this are endless. ChatGPT has already crossed 1 million users and has clearly done a good job in terms of launching this product. They’re also benefiting from the feedback received, something that’s key to data products in order to improve accuracy. The best part is that OpenAI was started as an open source initiative & by enabling others to access their API and their tools, they’re enabling many others to innovate in this space and collaborate. It would be great to see more products integrate these capabilities & leverage this. We’re only skimming the surface & there’s still a long way to go, but I’m excited that we’re heading in the right direction.

Disclaimer: All ideas & views here are my own & do not represent my employer, past or present, in any way or form.