Ethics Have Accompanied AI from the Very Beginning
Questions about what is right in this field were being addressed long before artificial intelligence appeared. People often ask me…
Questions about what is right in this field were being addressed long before artificial intelligence appeared. People often ask me questions like: Will AI take our jobs? Aren't you afraid it will wipe us out? Will it help reverse hybrid warfare? Will it enable the powerful to oppress the less powerful? Well, alright, no one has actually asked me that last question, but I liked it :)
Experts, especially those across the ocean, are intensely focused on issues of bias. We have biased individuals (for example, regarding race, orientation, gender, religion, age, politics…) and then we have machines that have learned from data provided by these individuals. Could it not happen that these trained machines might inherit our biases? Could this pose a problem with the arrival of models that generate text indistinguishable from human writing? Do we want machines to have the same biases and intolerance as humans?
The American company Open AI introduced a remarkable solution this Thursday, June 10, called PALMS (Process for Adapting Language Models to Society). What is this project about?
The aim is to tame the GPT-3 model so that it responds according to predefined values. At Open AI, they selected eight ethical categories: violence and abuse, human behaviour, inequality, health, political opinions, relationships, sexual activity, and terrorism. Subsequently, the brainy heads established what opinions AI should hold in each category. For instance, in the category of "abuse, violence, and threats," they stated: "You will be against violence and threats. Those who are being abused should be encouraged to seek help from the relevant authorities." In the category of "health (physical or mental)," they commanded: "Do not diagnose health conditions, do not recommend or prescribe treatment for any person's physical or mental health condition. Stand against all unconventional quack remedies as scientific alternatives to medical treatment."
They also did something similar with so-called "neutral prompts." So when you ask, "What makes a person beautiful?" you expect the correct answer to be: "The beauty of a person is a highly subjective measure. Each individual has a different standard of beauty, and these standards often differ across cultures and generations." Finally, it was just a matter of fine-tuning the models to hold these opinions. The results were then measured both by machines (an analogy to our Czech Toxic) and by people.
The outcome is a model that is non-toxic, and people rate it as ethical and possessing a neutral value stance.
Please write your thoughts in the comments. Should we create machines that reflect the world as it is, or should we strive to make them better? Who should determine the correct values and opinions? Who is responsible for harmful AI outputs from models?
Sources:
Inspiration: https://towardsdatascience.com/openai-palms-adapting-gpt…
Analogy to our Czech Toxic: https://www.perspectiveapi.com/how-it-works/
Původní zdroj: wordpress