Basis for AI Learning (incl ChatGPT)

Gigo’s law of deduction: Garbage in, garbage out.

The development of AI models like ChatGPT reminds me of the invention of the printing press. While there was the idea that now people become more educated, learn valuable skills, the printing press was also used to print, well, books of questionable value.

Which makes me wonder what the AI models will become when they are fed with all of humanities information and knowledge. And yeah, that is a huge issue when it comes to biases in the AI pipeline. There is a lot of research done. But I think the problem might be a little bit deeper. After all, we are not rational. We never were. We can try to act rational, and science is a method which tries to be as objective, reliable and valid as possible, but even then we are struggling (just look the biases in science).

And this does not even cover outright irrational stuff. Funnily enough, while ChatGPT 3.5 ( ) mentions that things like magic have no scientific basis, it has no problems coming up with a «spell»:

Create an actual and not fictional magic spell that will give someone luck in a future endeavor.

It’s important to note that in reality, there is no scientifically proven method for casting magic spells or influencing luck through supernatural means. However, many people find comfort and motivation in rituals and positive affirmations. If you’re looking for a symbolic ritual to enhance positivity and confidence for a future endeavor, you might consider the following:

Luck Enhancement Ritual:

Items needed:
1. A small charm or talisman (could be a piece of jewelry or a small object with personal significance).
2. A candle (color choice is personal, but green or gold are often associated with luck).
3. A quiet and comfortable space.

1. Set the Stage:
– Find a quiet and comfortable space where you won’t be disturbed.
– Light the candle, symbolizing the illumination of positive energy.
2. Center Yourself:
– Take a few moments to focus on your breath and clear your mind of distractions.
3. Empower the Charm:
– Hold the charm in your hands and visualize it radiating with a warm, golden light.
– Say, “With this charm, I invite luck and positive energy into my life. May it guide me in my endeavors.”
4. Affirmations:
– Close your eyes and speak positive affirmations about your future endeavor. Visualize success and imagine the positive outcomes.
5. Charge the Charm:
– Pass the charm through the candle’s flame (without burning it) while saying, “By the flame, I charge this charm with the energy of luck and success.”
6. Set Intentions:
– Hold the charm close to your heart and state your intentions clearly. For example, “I am confident, capable, and open to the opportunities that come my way. Luck is on my side.”
7. Seal the Ritual:
– Blow out the candle, symbolizing the completion of the ritual.

Remember, the efficacy of such rituals lies in the positive mindset they help create. Believing in your abilities, setting clear intentions, and taking practical steps toward your goals are essential for real-world success.

So, yeah, it might be interesting which monsters we create with these AI models. Well, one positive aspect, we might make our biases explicit and we might look in the mirror made of electrons and see ourselves.

How was it, the most terrifying thing would not be encountering aliens on other planets but other humans.

Then again, if we trained it only with rational, science-based information, it would be inhumane. Reminds me of that quotation:

«Worlds governed by artificial intelligence
often learned a hard lesson:
Logic Doesn’t Care.»
Yin-man Wei, «This Present Darkness: A History of the Interregnum», CY 11956 in Andromeda

Gonna be an interesting future, as long as it lasts.