Conversation with GPT-3 About the Future of Spoken Language Interpreting
Written By: William (Bill) Glasser, President and Founder
Recently I gave a workshop at the California Healthcare Interpreting Association’s annual conference in San Jose that covered the topic of Artificial Intelligence (AI) and it’s potential to put the interpreting industry out of business. Most of my colleagues who work in this field are well aware of the incredible accuracy and speed in which Natural Language Processing found on modern AI translation platforms perform, so I wasn’t too surprised to see the workshop well attended. I am, to be clear, a total novice when it comes to understanding data science in general and more particularly how AI works and how it will affect us all. I am concerned that these technological leaps often happen while we are all unaware, and moreover, by the time our careers become disrupted, it will be far too late to make a difference.
After the workshop, I decided to visit the big brain itself/herself/himself and ask a few questions about the future of my beloved interpreting industry. Please note that you too can have conversations any time you like with one of the largest digital brains on the planet by visiting openai.com and opening a free beta account. It’s fun, and quite frankly wonderfully terrifying!
Here is the transcript from my simple questions with GPT – 3 accessible through my free account at Openai.com. The answers come from the brain and are presented unedited.
Bill: How are human spoken language interpreters better than AI-generated language translation systems?
The GPT-3 Brain: There are a number of reasons why human spoken language interpreters are better than AI-generated language translation systems. First, interpreters have a deep understanding of the source and target languages, as well as the cultures associated with them. This allows them to provide accurate and nuanced translations that take into account the context in which words are used. Additionally, interpreters are able to adapt their translations on the fly to ensure that they remain faithful to the original meaning while still being understandable to the audience. Finally, human interpreters can provide emotional support and guidance to speakers who may be feeling anxious or stressed about communicating in a foreign language.
Bill: Do you believe that accurate speech-to-speech translation is possible using current AI technology?
The GPT-3 Brain: Yes, I believe that accurate speech-to-speech translation is possible using current AI technology. However, it is important to note that there will always be some degree of error when translating between two languages. This is due to the fact that languages are constantly evolving and there are often multiple ways to say the same thing. Additionally, different people may use different words or phrases to express the same concept. As such, it is important for users of AI-generated language translation systems to be aware of these potential errors and to have a backup plan in place in case of communication breakdowns.
Bill: How will the accuracy of GPT-4 language translation from Spanish to English be compared with GPT-3?
The GPT-3 Brain: GPT-4 is still in development, so it is difficult to say how its accuracy will compare with that of GPT-3. However, it is likely that GPT-4 will be at least as accurate as GPT-3, if not more so. This is because GPT-4 has been designed specifically for language translation, while GPT-3 was designed for general-purpose tasks. As such, GPT-4 should better understand grammar and vocabulary rules in both Spanish and English. Additionally, the developers of GPT-4 have said that they are using a larger training dataset than was used for GPT-3, which should further improve the accuracy of the system.
Bill: How much larger is GPT-4 compared to GPT-3, and what does that mean for the human-centric spoken language interpreting industry?
The GPT-3 Brain: GPT-4 is about 10 times larger than GPT-3. This means that it has the potential to be much more accurate than GPT-3, as well as being able to handle more complex tasks. However, it is important to note that GPT-4 is still in development and has not been released yet. As such, it remains to be seen how well it will actually perform in real-world scenarios. If GPT-4 lives up to its potential, then it could have a major impact on the human-centric spoken language interpreting industry.