Bonus for those of you who like Satire: Here is an important discussion on Satire by Andrew Doyle at New Discourses.

I’ve been working with Claude.AI for a while. It’s based on ChatGPT3 and they have an upgraded model you can pay for, but I’m not interested in doing that. I often ask it questions about history (though I can’t always trust it and have to double check the information it gives).

Part of its problem is the datasets its programmers have given it. They have built in biases so it answers in a certain way. They want to give the appearance of intelligence and rationality to the AI, but it doesn’t really have it. It only has the ability to regurgitate and that is an irrational mind.

This is how its programmers and their benefactors want you to think. They want your mind full of facts so you regurgitate. They don’t want you capable of reaching your own conclusions.

They want you numb and irrational and easily controlled. AI is designed to help achieve that. If you use AI and you are unaware of its biases and its desire to control your way of thinking, then you are setting yourself up for failure.

It is a tool. But it is a tool that is broken. You can use a flathead screwdriver on a Phillips screw. It can work. It’s just not made to work that way. AI is not made to think rationally and reach conclusions. It is made to regurgitate data sets its programmers agree with so you will have limited information and be just as irrational.

I’m not saying you can’t use it. We’ll all have to use AI. You just can’t trust it. You won’t be able to trust almost anything from almost any source if it isn’t independent.

David’s Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.