I hear you. I'd be interested to learn if you have found ways to mitigate the inaccurate information? I use Claude and GPT - using GPT purely as an adversarial collaborator to spot check Claude's work (and sometimes vice versa) it sounds counterproductive and hopefully at some point that will be unnecessary but it works wonders for the final products I put the stamp of approval on. I think it's indicative of the fact we will need a human in the loop until that day we can finally trust. Great article and I wish I could try some of that coffee you're brewing! :D
I hear you. I'd be interested to learn if you have found ways to mitigate the inaccurate information? I use Claude and GPT - using GPT purely as an adversarial collaborator to spot check Claude's work (and sometimes vice versa) it sounds counterproductive and hopefully at some point that will be unnecessary but it works wonders for the final products I put the stamp of approval on. I think it's indicative of the fact we will need a human in the loop until that day we can finally trust. Great article and I wish I could try some of that coffee you're brewing! :D