Anthropic really does have a big lead on the competition. Their models are better at understanding what the user wants instead of what they say. This doesn't always show up on benchmarks. All of the other big models fuck up their reasoning around weird edge cases.
The secret is what they reinforce in their model. They don't fall for the sycophancy that ChatGPT uses.
It's clear that other models are sounding more and more like Claude.
Yeah, I design the basics with ChatGPT, use Claude for the coding -havent tested it for BIM referencing yet-, and images tend to be banana Pro and MidJourney. Video is such a monster that it depends on what you want to do specifically.
Anthropic really does have a big lead on the competition. Their models are better at understanding what the user wants instead of what they say. This doesn't always show up on benchmarks. All of the other big models fuck up their reasoning around weird edge cases.
The secret is what they reinforce in their model. They don't fall for the sycophancy that ChatGPT uses.
It's clear that other models are sounding more and more like Claude.
Over the weekend I made an output style for Terry Davis. It refuses to call me a nigger but it's better than the alternatives would do
Yeah, I design the basics with ChatGPT, use Claude for the coding -havent tested it for BIM referencing yet-, and images tend to be banana Pro and MidJourney. Video is such a monster that it depends on what you want to do specifically.