Gartner once predicted that we would get to a point where the average person would have more conversations with chatbots than with their significant other. Is that really communication, though? By the standard definition of communication, one could say yes. But at a time when it’s impossible to go more than an hour without hearing about or interacting with AI, knowingly or unknowingly, perhaps the question we need to ask is: what does good communication look like in today’s world?
See, the goal of communication isn’t simply to share information. It’s to be heard, understood, moved, influenced. Communication changes us in some way: our actions, behaviors, perspectives – even in the workplace. Therefore, the language we use – intentionally or otherwise, with humans or AI – can have a dramatic impact on the outcomes of our communication objectives.
The Problem with “Good Enough” English
One of the biggest challenges employees face is how to communicate with different audiences. The default assumption is that the same script can generally work regardless of who you’re talking to. The same seems to be true of technology. Systems appear to be designed around assumptions about how people communicate, what they comprehend, what certain words mean, how we interpret subtext, what is innately understood versus what requires greater explanation, and so on.
Native English speakers can take these assumptions for granted. They may not notice the significant linguistic and cultural challenges that multilingual and non-native English speakers have to face every day. These speakers are spending time translating, calibrating their tones, searching for the right register, and figuring out how to enter the conversation to best represent themselves and to avoid being misinterpreted.
According to one study, speakers overestimate listeners’ understanding 46% of the time. Another found that 46% of employees cite miscommunication as the top cause of missed deadlines.
Add AI into the mix, and the issues can amplify. If an AI system is trained or prompted to use idioms, jargon or ambiguous language, the results can become even more difficult for non-native English speakers to navigate.
Going Global
Nothing derails a conversation or presentation faster than if everyone has a different interpretation of what’s being said. The same is true for conversations with AI agents. This is where Global English can help.
Global English is a form of English designed to make communication clearer across cultures and language backgrounds. It favors plain verbs and concrete nouns over jargon and abstractions. It avoids things like local references or sports metaphors. The goal is structurally clear language through a common framework.
When organizations are trained to use Global English, multilingual and non-native English speakers can spend less time code-switching and more time genuinely working. And it’s not just about multilingual speakers – Native English speakers also benefit from increased productivity when they don’t have to spend so much time correcting communication errors. According to a report from Axios, these internal communications errors are costing businesses “upwards of $2 trillion per year.” Imagine what that figure can become when you factor in miscommunications with AI.
Prompting Globally
The principles of Global English are just as applicable to prompting.
A good prompt is not just a list of instructions. It establishes a shared understanding of the task, language, context and desired outcome. Using Global English, especially in the context of AI prompting, requires a mindset shift on the part of the company and its workforce.
For native English speakers, this may mean unlearning expressions that have always seemed obvious.
“Move the goalposts” becomes “change the target.”
“Ballpark figure” becomes “approximate number.”
“Batting 1,000” becomes “consistently successful.”
For multilingual and non-native English speakers, the work is different. Known differences need to be made explicit and resolved in context. For example: “For the U.S. team, ‘launch’ means general availability. In India, it may mean pilot. For this project, ‘launch’ means general availability.” That sentence may feel overly explicit to a native speaker. That is exactly the point.
A prompt might include:
- Directive
- Language rules
- Number and data rules
- What to preserve
- Output format
So, here’s a general rule: make the prompt detailed. Keep the language simple.
Clarity is Capability
There are countless benefits to using Global English. And thankfully, the principles of using Global English are straightforward: say what you mean, mean what you say, and make the language accessible to everyone – including the AI. When companies take the initiative to use clearer language for everyone – native and non-native English speakers alike – AI has a better chance of delivering the intended outcomes more quickly and more effectively.
