I wish I had AI-powered headphones many years ago at a bar in Dublin, Ireland. It would have saved me some embarrassment.
I remembered a story from 2011. I was working at GitHub at the time, and many of us attended Funconf in September. It was my first time in Ireland, and one of the few trips I had taken abroad since 2011, right after I had joined GitHub.
I had many colleagues from GitHub joining Funconf, along with many of their friends from other famous Silicon Valley startups. I was generally alone, not having been too long with GitHub. Working from Greece remotely, didn’t help either, as I never had the opportunity to form a closer relationship with any of my colleagues yet.
My goal, of course, was to mingle and socialize, but all that was happening in loud Irish bars. I had a problem in those environments. A problem I had already started to realize from other interactions in settings like GitHub company summits, that I had already attended earlier. I couldn’t understand what people were saying in English in loud environments. Especially, folks that weren’t raising their voice a bit to compensate for the noise.
I had that issue cause although I knew English, and I hadn’t been exposed to in person conversations with native English speakers, each with different accents, in loud environments before. I had only been exposed to TV, cinema, and other media. And of course, written English. Joining GitHub in 2011 created all these opportunities for me to travel to various places and interact with English speakers in person.
I really wanted to participate in the various conversations, but I couldn’t. I was embarrassed when people were asking me something, and I had to say “sorry, could you repeat that?” 2-3 times, until someone besides me would transfer what was being said close to my ear.
I would get away like that most of the times, but that day in that bar in Dublin, I was sitting at a table with more than 10 folks. There was a lot of noise. I couldn’t really understand the various conversations going on. There was one fellow, from another startup that was quite famous among my colleagues. He was sitting across where I was sitting. At one point he probably noticed I wasn’t participating in the conversation and asked me something. He probably wanted to include me, which was great. The problem was I couldn’t understand a single word of what he was saying. I think it was the extra bar noise, his accent and volume of his voice, that made things a bit worse than usual.
After a couple of tries, I think he must have been very confused by my look. I remember his look, haha. He was partly confused, partly annoyed. As I wasn’t able to answer to what he was trying to say or ask, at some point he abandoned the effort and returned to his conversations with the other folks.
As the years passed and I had more interactions in similar situations, and more exposure to different accents, my ears started to become a bit better. However I wish I had the following technology at the time:
Researchers at the University of Washington developed AI-powered headphones that allow users to focus on a single speaker in noisy environments by looking at the person for a few seconds. Called “Target Speech Hearing,” this system cancels out other sounds and plays only the selected speaker’s voice. It uses machine learning to enhance voice clarity and is tested to be nearly twice as effective as unfiltered audio. The technology, not yet commercially available, aims to improve hearing in noisy places and might extend to earbuds and hearing aids.
It would have saved me from the big embarrassment in that bar in Dublin.
Here’s what I am doing
At Workbrew, I help our customers succeed, while working on docs, fixing bugs, and developing internal tools. At Amignosis, I pour my heart and skill into crafting slowly brewed software, one thoughtful line at a time. I am craftsman in a world of complexity and low-quality solutions. I am a shoemaker. I take the time to create simple, timeless software built to last. Check what I am doing now and talk to me.
Leave a Reply