@tjrobertson52

So OpenAI just released ChatGPT agent and I wanna talk about what this means for businesses both now and in the future so if you don't know ChatGPT agent is essentially the ChatGPT 0 3 model that can also take actions it can use a browser it can make spreadsheets it can make presentations and it can do this autonomously on very long tasks only been out for a day so there's still a

lot we don't know and I'll make more videos once we've been able to play with it more it's still pretty much unusable at most of the things that it could theoretically do overall it seems like people are pretty impressed with what it can do seems like the best use case for ChatGPT agent right now is doing any kind of work with spreadsheet before this if you wanted AI to help you

with your spreadsheets you had to have some kind of integration like Airtable or G b t for sheets or some kind of n8n automation hooked up to an agent but it seems like ChatGPT agent might make all of that obsolete.

com. But I neglected to mention that AI itself is also going to replace a lot of this. As general AI gets smarter and smarter, a lot of these specific cases just don't need to be built out.

And that's why my advice to most business owners that are asking if they should be automating complex tasks is to just wait. Now if you have the budget and the automation would pay for itself within a year, then it might make sense. But we know all the big model makers are working on agents like this, not just OpenAI, it's Google and anthropic and perplexity.

And from what we've seen so far, it seems like OpenAI is a little bit further down this path than we might have expected. Now in that last video, a lot of people mistook me for saying that building automations never make sense. I still think there are good use cases right now for tools like Nathan.

If you already know how to use the tool and you can build something quickly, go for it. Do you have a process that's very specific to your business with predictable inputs? Building that automation may very well make sense.

I would just be cautious before investing a large amount of time or money into a complex automation. Assume that within a year it's going to be way easier and way cheaper. Instead, spend your time just improving the processes you already have or building processes that you should have built a long time ago.

It might not make sense to automate them just yet, but when it comes time to, you'll be happy that you have well documented processes.