We are all wrong in the "AI Conversation”

Gabriel Fairman, Founder and CEO of Bureau Works, joins ProZ.com as a blog guest to reflect -- and invite us all to reflect with him-- on AI and the challenges we face when discussing it.

“AI” isn’t the biggest challenge our industry is facing. Talking about AI is.

I log onto LinkedIn every morning to write a post about AI in translation. While I am scrolling around for a minute before I write, I see anywhere from 2 to 10 posts about AI. Nearly every post is some version of the “Humans VS. AI” argument- Quality, Morality, Privacy, Liability, etc.

This is the issue.

Our discussion as an industry focuses on AI like it is a battle to win or a clock that can be turned back. It's not. AI in translation is an inevitable paradigm that we need to curate; a process we need to develop. It isn't "if" AI will change the industry, it is "when" and "how". This is why we need to talk about how to work with it, not against it.

This is my opinion, but it is also indicated by patterns in history. New tech is always controversial at first, then it is common. Recognizing this pattern and adopting the tech early on is the best way to ensure you are part of the evolution of our industry and not a casualty of it.

But the tech itself isn't actually the challenge we are facing right now. The challenge we are facing is how to discuss the tech, the future it will bring, and the fears and uncertainties that come with it.

The current discussions about AI have two fundamental problems: WHAT we are talking about and HOW we talk about it.

Issue #1: "What"

The “What” issue is that the fundamental focus of the conversation is oppositional. We shouldn't discuss "Humans VS. AI" when the reality of the future is "Humans + AI". Humans will use Generative AI to be more accurate, efficient, and effective, and this will introduce new challenges. But there will be no "AI OR Human"; no "this or that".

The productive conversation is about how we can work with AI, not how we can defeat it or how it will destroy us.

Most days of the week, I express that idea on LinkedIn. I publish a post, and then something strange happens.

I start to get comments pointing out every conceivable flaw of Generative AI as it applies to our field: Technical flaws, moral flaws, security flaws, hypothetical cognitive effects, etc. The comments almost always boil down to telling me that AI can’t replace translators and I should stop saying that AI is perfect.

The strange part is that I never say either of those things. I never would, because I don’t believe them.

Now, does this confusion happen because I don’t communicate well, or because people intentionally misunderstand?

I don’t know, but I do know that by not posting “Anti-AI” content I am perceived as an AI fanboy or a tech bro out to make a quick buck. This is indicative of the rift in our industry, and it is stopping us from responsibly discussing how AI fits into our future. Oppositional attitudes are getting in the way of productive conversations.

What people say about me on the internet isn't important, but having the right discussions is. As long as we consider humans and AI to be in opposition rather than collaboration, we will continue to have unproductive conversations. We will need to work with AI in the near future, so we must learn to talk about it now.

This leads to the second issue we are facing- How we talk about AI.

Issue #2- "How"

The second issue with oppositional discussions about AI is that they thrive on imprecision. Broad statements are by nature correct and incorrect at the same time, so the discussion becomes circular when it is based in generalizations. For example, “AI can’t translate” and “AI can translate” are both correct AND incorrect statements. The conversation requires nuance to hash out exactly HOW the statements are correct or incorrect; vague condemnations can be combatted with vague endorsements and vice versa.

Only detailed discussions of the tech's strengths and liabilities can help us create a map to navigate the inevitable. AI is here, it isn’t going away, and figuring out how it fits into our industry is complex. We need nuanced conversation to do this.

Any conversation portraying this as a simple, black-and-white issue is taking space away from the conversations trying to find real answers.

The truth is that our industry is in a moment of evolution, and it is a mistake to think that the changes will stop. But, the beautiful thing about this evolution is that it is not exclusive. We can all adapt, pick up the new tools, and be the force behind the new iteration of our world.

I am optimistic, but I am also not naive- I know that evolution is as brutal as it is beautiful. AI will bring on tough times for many, and some translators will lose their careers. But we all have the agency to participate in this evolution instead of being blindsided by it. We can adapt ourselves to the coming challenge.

We can do this by learning to use Generative AI tools, and we can help others do this by having better conversations.

AI isn’t divine or diabolical, it is a product of humans on earth. This means that it carries with it the potential for good and evil, just like everything else we have built.

Whether it is used for good or evil is up to us, so let’s have good conversations about how to use it wisely.

What do you think? Are you up to discussing how to use AI wisely?