3 ms·
I’ve found that Claude is very literal. When I talk to 5.5 it gets what i want it to do, when I talk to Opus 4.8 it does what I say literally and doesn’t get th
by akmarinov 3mo ago
I’ve found that Claude is very literal. When I talk to 5.5 it gets what i want it to do, when I talk to Opus 4.8 it does what I say literally and doesn’t get the intent behind it.
- mingqiz 3mo agoThis. Claude is very good at instruction and treat my mistake in prompt as law. Gpt is just smarter at understanding my intent.