Uncomfortable Truths About AI in Translation Nobody Talks About

AI is changing everything. Fast.

In the world of translation, the impact feels even stronger.

But most people don’t want to talk about what’s really happening. Gabriel Fairman, CEO of Bureau Works, does.

“Translation isn’t a niche anymore it’s a problem everyone’s trying to solve.”

In this video, Gabriel breaks down the hard truths we often ignore. Here's what he had to say.

“Automation” Is Mostly Just a Buzzword

Many companies say they’ve automated their processes. But what does that really mean? In most cases, it just means someone pushed a button.

“It’s very simple to send an email to 30 translators when a project is approved. That’s not real automation.”

True automation means:

• Matching tasks to the right translator using real data
• Tracking quality and deadlines automatically
• Moving work forward with little human input

Most companies aren’t doing this. They’re still stuck with tools from 10–20 years ago.

Post-Editing Is Unfair and Exhausting

Translators are often asked to fix machine-generated text.

They're paid less because “the work is done.” But the reality is harder.

“It’s easier to write a sentence from scratch than to fix 50 tiny mistakes under time pressure.”

Here’s why it doesn’t work:

• Machines don’t learn from changes
• Translators must repeat edits over and over
• It’s mentally draining but paid like light work

Plus, translators move faster under pressure but that speed comes at a cost: burnout.

‍

Nobody Wants to Change

The tech is evolving. But the industry? Not so much. Many language companies still:

• Use outdated relational databases
• Rely on manual input
• Avoid real-time AI tools

“There’s this group of people trying to pretend everything is tidy while the kitchen is on fire.”

Instead of embracing change, companies freeze. Clients wait for agencies to act. Agencies wait for clients. And nothing moves.

Cultural Ownership Still Matters

Even if AI reaches “parity” with human translators, Gabriel says there’s still a role for people. Not just to fix things but to own them.

“We may want people to own the text. Not just verify it own it.”

This means:

• Holding responsibility for voice and tone
• Keeping content aligned with company values
• Avoiding errors that AI can’t catch

Without that, you lose connection to your brand’s identity and your audience.

‍

Real Change Means Real Discomfort

Most businesses try to “look automated” without changing anything major. Gabriel calls that a mistake.

“You can’t have real automation without a lot of change. And change is painful.”

Here’s what change could look like:

• Project managers shift to data and quality roles
• Agencies invest in AI verification services
• Translators focus on creativity, not repetition

Yes, it’s uncomfortable. But it’s also the only way forward.

Conclusion: Efficiency Starts With the Big Stuff

Gabriel’s advice? Don’t start small. Fix the big inefficiencies first. Even if they hurt.

“The game people are playing is: how do I adopt AI without disrupting anything? That’s not how this works.”

To thrive in this new world:

• Question what “automation” really means
• Value the human side of language
• Embrace change before it’s forced on you