Which is better? OpenAI vs Claude 3 - An Introduction

Introduction to AI in No-Code Apps

If you're building a no-code app and you're wanting to add AI into it, then for a long time the service that we were all talking about was OpenAI or ChatGPT. But now we really need to be taking Anthropics, Claude and especially Claude III seriously. Now if you're building a no-code web app, there's no better place to learn how to do that than at PlanetNoCode.com. And you can follow a link down in the description where you can find hundreds of tutorial videos presented by me where you can basically build 90% of startup web app ideas. You can build them with Bubble.

Exploring Claude's Latest Release

Let's dive into this latest release from Claude. So it comes in three different flavors. We have Haiku, we have Sonnet and we have Opus. And similar to how we've become used to OpenAI GPT 3.5 versus GPT 4, we've got Speed and Price on the spectrum here, but also the Intelligence Index. Now the scholar in me says you always have to be wary of a chart that has no labeling and it doesn't have a label on the y-axis. But I think it's just meant to clearly express to you that Opus is vastly more intelligent than Haiku. But also it's going to be a little bit more expensive because we do get that scale on the x-axis.

Comparing AI Models

Now remember this is reporting from Anthropics themselves comparing their model to GPT and to Gemini. But we get some very impressive data in here specifically when you look at Opus. But across the board, we're getting claims here that the Haiku outperforms GPT 3.5 for graduate level, undergraduate level. I mean something that OpenAI chat GPT, all of these AI models have struggled in the past, is the ability to do maths. So we've seen a huge improvement now. But we've got GPT 4 coming out above Sonnet and Haiku. We've got all of these benchmarks. Obviously this is their launch page, this is their press page. They're wanting to make their offering, their being Anthropic, I want to expect their offering look better than OpenAI.

Improvements in AI Models

We get some comments here about Speed saying that. And we've got previous videos. If you go onto our website, search our YouTube channel, you can find videos about Claude2. And they're saying that Sonnet is up to two times faster. They've added in vision models here, less refusals, improved accuracy, long context. Oh yeah, look, 200k context window. I mean that is just huge. What we're seeing with AI models and the LLMs as they grow and develop, we're basically seeing every few months huge advancements. Which just kind of blow the door on what is possible.

Expanding Context Windows

Because it wasn't that long ago that we were all having to summarise our conversations. You had to really keep the context window low to like a few paragraphs and then a few pages. And now we're onto books and soon we'll be on multiple books. So it's really incredible the amount of data that you can provide to a model in order to generate a response. We then get loads about safety here and finally pricing on each of the models.

Conclusion and Future Plans

So yeah, there's a quick summary. Yeah, just reading through all that. I'm going to do some videos about how to integrate this into a Bubble soon. So do like and subscribe to our channel to not miss those. But this is really exciting. If you've got any questions, you've got any thoughts. If you are also comparing OpenAI and Anthropics, Claude, do leave a comment down below and let us know how you're getting on. What difference have you noticed between the two models? Because yeah, like I said, it's really, really exciting time to be working with AI and building no-code apps.