@tjrobertson52

1 is out, so I figured we'd talk about it and then I'll tell you what model I actually think you should use for work. 0 would do. It's supposed to be better at knowing how long to think for responding and then they also give you more control over the personality of the model.

I think what's really clear is that OpenAI is very much focused on the consumer. They want ChatGPT to be your go to everyday model for common questions and I think they're doing a great job at that. As a consumer myself, it's the one I find myself going to most often, but not for work.

Actually maintain a subscription to ChatGPT, Gemini and Claude because I think each of them is the best at different tasks. I would say about 70% of our work is done inside of Claude. Claude has never really focused too much on the consumer.

It's always been focused on the workplace or enterprise. Historically, they've always been the best coding model, but it does seem like they're starting to lose some ground there. We like Claude because the writing is the most natural among all the models.

It's also the best at handling a large amount of context without missing details and the best at adhering to our instructions. So since most of our workflows are built around creating content, while keeping a large amount of context in mind, Claude just outperforms the other models. They also continue to roll out features that let you customize the response.

They recently came out with something called skills, which looks like it's gonna be really powerful. However, ChatGPT is still the model we used second most. Let me tell you what I think it's best at.

We primarily use ChatGPT to retrieve information. So for writing a blog post, we'll typically do deep research with ChatGPT to collect a bunch of data points and quotes and statistics around the topic. Or just yesterday, we gave it a list of 100 domains and asked it which of those domains was a directory that we might want to submit our client to.

Anytime you need a model to agentically search across the internet for specific details, I think ChatGPT does the best. Now Gemini really should be the best at this. They definitely have access to the better search engine.

As someone who helps companies show up in both Google and ChatGPT, I can say ChatGPT is much easier to game. Their search tool is just much less sophisticated. However, it just seems like ChatGPT tries harder.

As far as I can tell, Google's primary focus with Gemini is integrating it into their existing products. They know they already have the distribution advantage and now they're just focused on making it cheap enough that they can offer it for free without losing money. Right now, we primarily use Gemini for brainstorming, like if we're looking for keywords to focus on or for creating documents like proposals or contracts.

However, if you look at the trajectory of each model, I do think Google is positioned to come out on top in the end. Gemini 3 is supposed to come out any day now, so I guess we'll see.