The number one tip of all if you really want to improve communication is: Eliminate Weak Language such as ‘Filler Words’. In English, the main filler word is ‘um’ – every language has some version of the hesitation form.
Why do we say filler words like ‘um’?
We say it because we’re uncomfortable. We want to use it as a crutch while we’re thinking of what we’re going to say next.
A word like ‘basically’, if you had to define it for a non-native speaker, what does it mean? The simplest form for ‘basically’ is a fancy way to say ‘um’. It means nothing, it adds nothing to your sentence.
Weak language is any word or phrase that does not add value to your message.
There is no neutral. Everything that comes out of your mouth makes your message stronger or makes it weaker.
Try to think of someone in your life who takes very long to say very little. When that person opens their mouth, what does your brain do? It never shuts off, but it raises its filter. Even when you like the person and you like what they’re saying, the filter goes up because it takes them so long to tell their story.
The opposite effect is true as well. Your filter goes down when they open their mouth because what they say is concise. They don’t use a lot of wasted words.
Whenever you open your mouth, everybody else’s filter moves up or down based on the reputation you have in their subconscious brain for weak language.
Your mission when improving communication is to identify and eliminate those words that you use that don’t add any value. However, we know it can be hard to self-identify your filler words as you are talking. That is why we built a cutting-edge Weak Language AI Engine, implemented on our LifeHikes® mobile application.
Can a computer understand weak language?
Human beings can easily recognize weak language in conversations since we understand the contextual meaning behind the sentences. But for computers, it’s a different story. Words such as “like,” “well,” and “right” all have different meanings, depending on the context. Some context makes them weak — but in other cases, those words have actual meaning. Does this mean it’s impossible for computers to recognize weak language? Although there are several obstacles, the Own The Room team figure out a solution: A.I. Language Model.
What makes our technology the best?
- Labeled Data from Professional Coaches – A.I. needs labeled data; the process is called Machine Learning. The outcome of that process depends on who’s the teacher. Our world-class professional Coaches provided more than 10,000 labeled sentences, which makes our model unique and full of “Own The Room style.”
- 500 Hours + Model Fine-tuning – Our team has spent over 500+ hours fine-tuning the parameter and structure of the neural network behind the A.I. engine. We trained more than 50 models with different approaches and picked the best one. As a result, the AI model reached 85% accuracy on “weak language” detection and outperformed our rivals’ rule-based model (yields 64% accuracy).
- Performance and stability – After optimizing the performance of our A.I. service, the average response time for weak language detection is currently about 0.12 seconds per sentence.
You can improve communication skills by eliminating weak language! Download the LifeHikes app (it’s free) and become the strong, concise speaker everyone likes to listen to.