The AI steelmans my case better than any therapist or friend I've spoken to in the past 30 years. Better than all of them together.
If it's true that healing trauma can't happen until after effective, empathetic validation, then the experience I've just had could turn out to me one of the most useful I've had in this sphere. It doesn't matter to my purpose whether the AI _actually_ understands when it performs understanding so well.
I'm going to enjoy this, Tom. Quite adventurous of you to offer this to us. One thing it will give me is a deeper understanding of your arguments. It is in its way a modest tool, not pretending to know what it does not know. There is something freeing about being able to ask a machine questions without wondering how the question represents me. Many thanks for getting this out to your readers and fans.
Thanks Allen. It is amazing the way that AI can take in information and then distribute it based on questions. I have played with it a bit and it does a pretty good job. I will be curious to hear what you think. Maybe you can do one on your book?
I never thought of this, Tom. I'll have to devise an experiment. It feels a little like "mirror mirror on the wall," a least to me, i.e., don't ask if you don't want to know.
Jesse - I think you are missing the point. The custom GPT is only based on the books I have written. It does not respond based on anything else. So far I have seen it take small liberties but it seems to stay true to the uploaded material of my three books. Ask it a question and see if it gives a male friendly response.
Yes, the danger of AI is that it is dependent upon whoever wrote the program. The beauty of the custom gpt's is that you get to control the data that gets fed to the GPT. In the instructions settings you can tell it how to respond, what tone to use, etc.
Thank you, Tom, and well done.
Glad you found it helpful.
The AI steelmans my case better than any therapist or friend I've spoken to in the past 30 years. Better than all of them together.
If it's true that healing trauma can't happen until after effective, empathetic validation, then the experience I've just had could turn out to me one of the most useful I've had in this sphere. It doesn't matter to my purpose whether the AI _actually_ understands when it performs understanding so well.
I'm very grateful, Tom. Well done!
Glad you found it of use Mark.
I'm going to enjoy this, Tom. Quite adventurous of you to offer this to us. One thing it will give me is a deeper understanding of your arguments. It is in its way a modest tool, not pretending to know what it does not know. There is something freeing about being able to ask a machine questions without wondering how the question represents me. Many thanks for getting this out to your readers and fans.
Thanks Allen. It is amazing the way that AI can take in information and then distribute it based on questions. I have played with it a bit and it does a pretty good job. I will be curious to hear what you think. Maybe you can do one on your book?
I never thought of this, Tom. I'll have to devise an experiment. It feels a little like "mirror mirror on the wall," a least to me, i.e., don't ask if you don't want to know.
The last thing I'm interested in is AI's "thoughts" about men and boys emotionality. Gender queer trash.
Jesse - I think you are missing the point. The custom GPT is only based on the books I have written. It does not respond based on anything else. So far I have seen it take small liberties but it seems to stay true to the uploaded material of my three books. Ask it a question and see if it gives a male friendly response.
Maybe I will, but I have zero trust in AI.
Yes, the danger of AI is that it is dependent upon whoever wrote the program. The beauty of the custom gpt's is that you get to control the data that gets fed to the GPT. In the instructions settings you can tell it how to respond, what tone to use, etc.