The difficulties of Remote Machine Learning Work

The difficulties of Remote Machine Learning Work

Remote Working ยท Sept. 3, 2018

At RemoteML we celebrate Remote Work and all its advantages. Working remotely is proven to enhance your happiness and productivity, but it’s not all just good. These are some of the difficulties of Remote Machine Learning Work.

Company Culture & Loneliness

Office workers know it: Coming to the office is not just always about work. In a healthy environment, a good team forms a company culture. Inside jokes, the chat about your weekend and friendships mostly form in the break room.

When you are working remotely, you don’t have a break room to attend and no seating partners to chat with. Obviously there are team chats like Slack, but initiating a casual conversation there is a lot more difficult.

So what can you do about it? Employers can fight this by introducting a virtual breakroom with video chat and remote freelancers can try becoming part of a community. Hint: The Machine Learning Chat is one of those.


When talking about remote work, people often jokingly ask me: ‘Oh, so how much do you work? Do you even work? Like two hours?’. The problem with remote work in most cases is not working too little, it’s working too much.

In offices you have a physical reason to end your work for the day: You have to head out at some point, get dinner and have no obligation - in some cases don’t even a way - to work from home. When your office is in your own appartment, you don’t have a reason not to work. You are home, you don’t have nothing to do, so might as well work right? Try to fight this by physically limiting yourself from working, for example by separating your work and personal notebook or by having a dedicated room as an office.


This difficulty falls into the same space, but is a little different. Sometimes if you are working for a full day, it might be that you don’t have anything to show off for the day. Who knows: Maybe a computer made some problems, maybe you didn’t get the data in time, maybe you just wanted to read some papers today.

In a physical office, you see people putting in the hours at their desk. Even if you don’t have anything to show, you were present. If you are remote, this feeling quickly vanishes. Your boss might think so, but the chances are high that you feel very unproductive on these days. In the end, if you don’t show anything, there is no proof that you actually worked, right?

The solution for this one is just to get out of the mindset that “presence = productivity”, since it is simply not true. If you put in the work, you should feel good about yourself, even if you don’t feel so yourself.

About the Author

Dominic is the founder of RemoteML. Additionally to working as a (Remote) Deep Learning Engineer and working on his own sideprojects, He is also a vivid writer, having written for Towards Data Science, HackerNoon, MakerMag and O'Reilly before).

When I started working remotely, I had a lot of questions. Now, I hope to help you with yours!

Find a Remote ML Job today!
Remote Machine Learning Jobs in your inbox as they get posted.