On Slack and Brainwaves

Alan Nichol
3 min readFeb 24, 2016

--

Ever since my co-founder Alexander Weidauer and I decided in January to kill off the product (treev) that we’d been working on for quite a while, we’ve been exploring new directions. We ran a little experiment using a consumer brainwave sensor called Muse, and thought it would be fun and interesting to share our thoughts.

Reading Minds

Neither of us have any special knowledge of EEG, neuroscience, or psychology, so we started from a super naive question: how far away are we from being able to read people’s thoughts? TechCrunch et al. keep saying that no UI is the new UI — so why not take that to the logical conclusion and build software you control with your mind?
It turns out there are a few ‘brain scanners’ you can buy for a few hundred dollars online. Anyone can use them (you don’t even need to apply gel for the electrodes to work) and some have SDKs you can play with. EEG is mostly a medical/research tool, but people have used it to do things like ‘neuro-marketing’ and it’s pitched to consumers as a meditation aid.

Available Hardware

We reached out to David Vivancos, who has been advising companies on EEG for a long time and helped us understand what the different devices were capable of. We considered the emotiv headset, which we’d seen used to control that little BB-8 robot, but it would have taken 5–6 weeks to deliver. We bought a Muse headband partly because Amazon had it in stock and we were impatient. It’s a slick piece of hardware and trying meditation for the first time was a cool experience. But now that we had the hardware, what should we build?

close-up of a muse EEG headset. you wear it with the band across your forehead.

Idea

Improving productivity at work is something we’d put a lot of thought into before, so we settled on a really simple idea: trying to build an anti-distraction device. It’s frustrating when you’re working on something that requires your full attention (like programming) and notifications keep coming in and breaking your flow. It annoys me enough that I usually quit Slack and put my phone on 🌙 mode— hardly an elegant solution.

But what if we could monitor your brainwaves and protect your focus when you really need it?

As it happens, interpreting variations in the voltage in different parts of your brain is nontrivial. We’re quite far from being able to ‘read’ thoughts, although a recent paper showed that some feature engineering and simple classifier can differentiate between the EEG signals of a person looking at two different images (a face and a house, for example). Fortunately for us, our headband already reports a ‘concentration’ measure. It’s not perfect, but we reliably saw our friends’ concentration scores shoot up if we asked them to spell their name backwards or to count in prime numbers.

Alex’s concentration score (dark grey) during a series of tasks. Sections highlighted in green are ‘active’ times. The correlation between activity and high concentration scores is far from perfect. Key: A-reading an article, B- checking facebook and email, C-writing some emails, D- Chatting, E- watching a movie trailer, F- playing an online quiz.

MVP

We built a very simple iOS app which tracks your concentration. It talks to the Slack API, and switches on ‘Do Not Disturb’ as soon as it starts. When your concentration level drops for a few minutes, we switch it DnD off again so your Slack messages come through. No need to quit Slack, and no need to keep to a strict pomodoro schedule.

Lessons

We’ve been trying it out and it’s pretty neat, but probably as far as we’ll take this idea. It’s nice to have and super fun, but we just don’t see a business there. We’ve posted the (pretty hacky) source code on GitHub in case anyone wants to buy a Muse and try this for themselves.

We’re still very much in exploration mode, so if you know a way to create a business out of this, want to work with us on it or have any other ideas you’d like to try, be sure to send an email our way. Want to just chat? You can find us doing so on Twitter..

--

--

Alan Nichol

Computer Simulation, Machine Learning, Startups. Co-founder @Rasa_HQ