Thoughts about chatGPT

I didn’t intend to discuss chatGPT, as I’ve been overwhelmed with the varying opinions on the subject. However, I recognize that the topic is unavoidable and raises many questions and concerns for professionals, especially those in language-related fields.

As such, I’d like to share with you my thoughts and reflections on the matter.

Regarding privacy and data confidentiality, it’s important to remember that if a service is free, we are the product. The data we input into Large Language Models (LLMs) is used to feed them. It’s a give-and-take situation.

With sensitive information, especially in the medical and legal fields, we must take the same precautions as we would with previous search engines. Private and confidential information cannot be shared on these sites. And that is that. Language professionals sign NDAs and have ethical and deontological duties to uphold.

For more general information, we can always ask for a translation, omitting names and addresses and using placeholders like “person X” and “street XYZ.”

Moving on to other questions, Europe has strict laws about data protection. Since these services are based in the USA, it’s essential to understand how they use and store data.

Ah! As with previous systems, I regret to say that the output is in Portuguese, the Brazilian variant. European Portuguese speakers do not produce enough material to feed these models.

Among other features, these LLMs conduct research and present integrated results. Apparently, this is a good thing: we don’t have to search for information across multiple outcomes, and the information is already summarized.

However, we need to pay close attention to the information that is being presented and the point of view it represents. LLMs are trained on American English content, so a US-centric bias exists. It would be valuable for Europe and other regions to have a culturally aware and less biased equivalent model, promoting diversity and inclusivity in AI.

In a recent interview with the New York Times, Noam Chomsky stated that this is the most radical attack on critical thinking, critical intelligence, and particularly science.

I think we need time and distance to understand the implications of all this.

While awareness of these issues is essential, it doesn’t mean we shouldn’t use these new technologies. We need to do so consciously. As Chomsky suggests, we must educate people and raise awareness about these tools to allow people to understand what they can expect.

Have you been able to think about this subject? Feeling worried or curious? Use or reject? A penny for your thoughts.