Survival according to ChatGPT

I wanted to test what ChatGPT was like so I told it:
"lecture about how to start a fire with a bow drill"
This is what I got back:
_______________________________________________________________________

Sure, I can provide a brief lecture on how to start a fire with a bow drill.

Starting a fire with a bow drill is an ancient method of making fire that has been used for thousands of years. It involves using a bow and a spindle to create friction between two pieces of wood, which generates heat and eventually produces an ember that can be used to ignite a fire.

Here are the basic steps to start a fire with a bow drill:

Gather materials: You'll need a few key materials to make a bow drill fire. These include a bow, a spindle, a fireboard, and some tinder. The spindle should be a straight, dry stick that is about as thick as your thumb, and the fireboard should be a flat piece of dry wood that is about the size of your hand.

Prepare the fireboard: Use your knife to carve a small notch in the fireboard, near the edge. This will be where the spindle sits and spins. Next, carve a slight depression next to the notch. This is where the ember will form.

Make the spindle: Cut the spindle so that it is longer than your arm, and about as thick as your thumb. Sharpen one end to a point, and make the other end slightly wider so that it can be held in the bow.

Create the bow: Find a flexible branch or stick that is about as long as your arm. Tie a piece of cord to each end of the bow, and then tie the cord around the spindle at the wider end.

Start drilling: Place the spindle into the notch in the fireboard, and hold it in place with one hand. Place the bow over the spindle, and wrap the cord around the spindle. Use your other hand to hold the fireboard steady. Begin moving the bow back and forth to spin the spindle, and create friction between it and the fireboard. This friction will create heat, and eventually an ember will form in the depression you carved earlier.

Create a coal: Once you have an ember, carefully lift the fireboard and tap it to release the ember into a small pile of dry tinder. Blow gently on the ember to help it grow and ignite the tinder.

Build the fire: Once the tinder has ignited, add small sticks and twigs to the fire to help it grow. Gradually add larger pieces of wood as the fire gets stronger.

Remember, starting a fire with a bow drill takes practice, patience, and persistence. It can be a rewarding and empowering experience to start a fire in this way, and it's a valuable survival skill to have in case of emergency.
_______________________________________________________________________

I can now see how people are making a living 'writing' bogus articles about subjects they do not know about.

I can also see how this could be dangerous to have integers into search engines, as it can give very realistic how-to articles that are wrong...
For example, a Ham Radio blogger did this test: Blogging With AI (ChatGPT)
It reads very well, but says that VHF is better for long distance communications than HF! It also misidentified the 70-centimeter(440 MHz) band as VHF rather than UHF...

What do you'all think of all this???

Ultimately AI is only as good as the training data that it was fed.

There are numerous examples that highlighted this, like getting an AI to diagnose skin cancer. Only afterwards did they figure out that the AI had associated a ruler being present in the photograph of the melanoma with a positive diagnosis of cancer.

There's other stuff like x-ray images of hands, and AI seems to be able to predict the ethnicity of the person.
Another is the ability to determine the sex of person just from a picture of their eye.

No one knew there things were possible, but seems the AI has "learnt" something, but what we cannot know.

I would say that we cannot hope for AI to be better than the training data that it is fed. I've seen it be much worse.

While I am not a data scientist, I've spent many years working for companies that employ them. Training data quality, the algorithms and heuristics used, and particularly the attributes of the training data (usually selected with a great deal of experimentation over time) are all critical to any kind of artificial intelligence.

Also, machine learning and "deep learning" are closer, in my opinion, to marketing terms than useful distinctions.

There's no such thing as AI and never will be. Machines don't have minds. All they can do is what people tell them to do.

qjs