The End Of The DJ

The music begins and a wedding couple takes to the dance floor.

Their closest family and friends watch as two people dance together for the very first time as a married couple. It is picture perfect and emotionally beautiful.

Moments later, this newly married couple are surrounded by their closest friends and family, dancing and singing away to their favourite songs. The celebration of marriage has once again begun. The energy is incredible. The party continues late into the night.

The dance music is mixed to perfection, giving no one time to consider if they are staying up for the next song or sitting down. Over and over again, before they know it, the next song has begun and they just can’t sit still. Everyone is having too much fun to think about anything else. The party is unstoppable.

The energy of the music ascends and descends in perfect rhythm for the party, as if orchestrated by the ultimate DJ. It feels like a DJ that knows the musical tastes of each and every guest, how those tastes cross over with each other, what their mood is like and what music they tapped their foot to during dinner.

When the night ultimately comes to a close, the guests heap praise on their newly married friends for throwing the greatest party. Everyone’s feet are sore from dancing and the girls tossed their heels off long ago. What an amazing night. I wish you could have been there.

Isn’t it amazing to think what can be accomplished without a DJ?

The as-yet-unamed technology, responsible in large part for the success of this party, is a form of artificial intelligence combined with the power of data collection and analysis from machine learning (for the difference between AI and Machine Learning, read this). With the use of cameras and software, the mix of music for the dance floor was as close to perfect as is humanly possible. Actually, scratch that (pardon the pun), it was better than is humanly possible.

The cameras were setup around the room, with a focus on the front of the room, near the head table and dance floor. From the moment guests began arriving, the DJ software was analyzing the live video from the cameras, carefully measuring the physical movements of the guests (foot tapping, shoulder movements, width of smiles on faces, etc.) in response to the music it was playing. Using the collected data as rapidly as it was being analyzed, the music genre and style was continually adjusted in real time, to match what the guests reacted too.

When it came time to dance, data from cocktail and dinner hour was used to evaluate what a possible ratio of one type of music to another might be used for the dance portion of the evening.

In addition to the cameras (which continued delivering real time data for instant analysis all night long) the software also used data from the numerous other events it had DJed (48,482 events, as of that night, to be more specific). The software knew, for instance, that following up song XYZ with the tune ABC, based on a crowd of 25 – 45 year olds in this region of Philadelphia will, at this time of day, work with exactly 98.7% certainty. Conversely, it also knows, with staggering precision, which specific songs are less likely to work.

While the number of events this software has worked at might surprise you, considering that it is working at thousands of events around the world each and every weekend, it doesn’t really take long to gather massive amounts of data. With the software still running on version 1.0, having only been out for around a year, the data, as large as it is, would still be considered young. It won’t take long to have the musical data of over 1 million events.

With every event, the number of hands shooting up in the air, the amount of movement on the dance floor and the overall success to the songs, based on the specific event type they were being used for is being tallied and stored for the benefit of all future events this software will be employed.

Other features of this software include having the wedding couple enter in each guests name and email address so that the software can run automatic guests searches through Spotify or Apple Music. As publicly shared playlists are discovered, it will have a better understanding of what music those who love music most, are into enjoying.

And let me remind you, we’re still on software version 1.0. Ideas have been brainstormed for 2.0 but have yet to be fully developed.

The question of software like this (that I simply made up for this blog post) taking over the world of mobile and club DJs is not a matter of if, but a matter of when.

If you think I’m a quack, it only takes looking to the many other examples of industry icons that refused to acknowledge what the future might look like, to have a change of heart. Names like Kodak and Blockbuster Video come to mind.

Taking a listen to the Kevin Kelly interview that inspired this post or to Elon Musk discussing the use of camera technology on Tesla cars (watch from the 13:59 mark) might also help you understand what I am talking about.

Industry specific types of artificial intelligence (not like Skynet from Terminator or what happened in the Matrix) and machine learning will be the greatest disruptor of human lives in many generations. As Kevin Kelly said in the afore linked interview, “Artificial Intelligence (AI) is the biggest thing since electricity.”

If you think the world of mobile DJing won’t be affected by AI, you are delusional.

How will you future-proof yourself against the computer collected, analyzed and managed data that no human will ever be able to compete with? None of us will win the data argument against a computer. Ever.

I don’t know who will build the software I described earlier (I’m probably too late to file for a patent on the idea right?!), but someone will. Serato, the much beloved music software for many of us, has already started down the road of automating well mixed music to the consumer with its app, Serato Pyro. That is only the beginning.

This will be, as we currently know it, the end of the DJ. Like with everything else out there, we must adapt, if we are to remain relevant, as mobile DJs.

Begin, if you have not already, to work on your human skills. The soft skills. Your crowd interaction skills. Your MC skills. Your creative skills.

Future-proof yourself, before the future decides for you.

EDIT: I discovered a remarkable TED Talk video shortly after publishing this piece, on the capability of computer/camera capability (For, maybe, reading the physical movements of a room full of dancing people?). If you don’t have 8 minutes to watch the entire thing, just watch from the 3 to 6 minute mark. Amazing stuff!!!

 

For your free download of all documents and email templates mentioned on this website, please enter your name and email here:

 


20 thoughts on “The End Of The DJ”

  1. jut another reason to focus on adding in some element of live performance or improv, whether it be live instruments or controllerism/turntablism, a computer can never repalce my own live personal creativity, because there is only one me and same for everyone else as well…

    1. You got that right my friend. Some DJs show up with very little personal creativity though, it’s those that I fear for the most. Upward on and onward for those who put in a little more then the minimum required…cheers my friend!

  2. Another post to your usual high standard Mr T and again very thought provoking.

    You made the comment Dave: ‘Crowd reading and changing songs are jobs that computers will easily out do us on one day’. One day, but my instinct is that day is a way off yet and that is because of a word I just used… instinct. Computers can learn many things but they are some way off from being able to learn instinct… yet. As I was reading the description of how this might work, a thought that came to my mind is that the selection of tracks they choose could be quite predictable and therefore boring. A good DJ takes chances sometimes with their programming and sometimes chooses a song from instinct that might be 20 years old and based on something overheard in the speeches… now the response to that comment might well be that the computer would listen to the speeches, but not every song mentioned in the speeches will work for the dancing and at present only a skilled/experienced DJ is able to discern what would/wouldn’t work in such a situation.

    Sometimes the best reactions to songs played are when a track totally out of the blue is played that results in guests running to the dance floor exclaiming with a smile on their faces “I haven’t heard this in years”, and the reason that song works so well is precisely because it isn’t overplayed and so the computer is extremely unlikely to ever select something like that. Admittedly that that type of track is one that could equally clear the floor, but again a skilled DJ takes chances with music because they have enough confidence that they’ll be able to fill the floor again even if the floor is cleared.

    Don’t get me wrong I’m not knocking AI, quite the opposite, it excites me at some of the results this will deliver, but a recent podcast episode I heard (I think it was Neil deGrasse Tyson being interviewed on the BigThink podcast, but it might have been a different show altogether- I listen to a LOT of different shows!) made the prediction that in many cases AI will work best in conjunction with humans rather than AI vs Humans or even AI vs AI, and they used chess as an example. When AI plus a human worked in tandem no AI was able to defeat them. So it could be that rather than replace human DJs, certainly for many years a DJ harnessing AI will deliver an even more amazing experience for the guests than the DJ or the AI could have alone.

    The show referred to above may have been this one: http://bigthink.com/think-again-podcast/neil-degrasse-tyson-nil-the-only-ist-i-am-nil-think-again-podcast-100

    1. And you sir, are equally on point in your considerate feedback. There will be many parts of what we do that computers will not be able to do, that it absolutely true. What I love most is maybe Roger’s comment here as well in that, high level entertainers will not only continue to deliver what they do, but they’ll do in in conjunction with the latest AI/ML advancements so that everyone is getting the best of all worlds.

      Thank you again for another link to add to my listening schedule. The material you have shared with me is always positive and insightful. In fact, were you the one who initially suggested I listen to the Kevin Kelly interview I linked too? (I don’t remember, because I listen to a lot as well!)

      Cheers Tony.

  3. I do think it is a ‘when’ not an ‘if’ issue. Perhaps the other dimension to be considered is the end customer. Will the wedding planner / wedding couple really care if it’s live or ‘AI’ driven as long as they get a guaranteed great day / night. Especially so if the cost is fairly trivial. However, AI / machine learning approaches will (IMHO) struggle to do the really creative stuff. The challenge will be for the DJ / entertainer to offer a service at a higher level that fully exploits the AI etc blending it with their creative abilities.

    1. Couldn’t agree with you more…especially the part about creative entertainers offering high level service while incorporating the best AI/ML components in order to deliver incredible experiences. Great thoughts Roger!

  4. what really makes a DJ, philosophically speaking, is the ability to choose. computers don’t make choices, their programmer’s do.
    choice is in essence a spiritual or arational quality and never accessible to a machine.
    life is too complex for machine learning to be better than any spiritually tuned chooser applying their will to a particular time•space context. however, it will always be much better than a human trying to be a machine.
    all technology does is demand more of our spiritual side to continue being relevant to the world.

    1. “it will always be much better than a human trying to be a machine.”

      That’s a really interesting thought because it makes me think…are we, as DJs, simply doing a machines job, that machines aren’t yet fully capable of doing yet? (but might soon be?) Really interesting perspective Caleb, much appreciate this thought of yours.

  5. This very well may become a reality. I think even further down the road most experiential entertainment may make it’s way to virtual reality. The human element will always be needed for events where connection to the audience is needed. People love personalities that make them laugh, swoon and gaze in awe. We must as entertainers not just rely on equipment but on talent.

    1. Agreed Rob. I hope this awakens those who rely solely on their equipment and music mixing talent that, in the long run, that maybe won’t be enough.

  6. Anyone can push buttons, but a djs greatest power is in the ability to read the crowd and change songs to keep the crowd dancing. Maintain the flow, as it is.

    1. Crowd reading and changing songs are jobs that computers will easily out do us on one day. (see the linked videos and audio interview for perspective on what I mean) Human skills are what will keep DJs alive — at least, those that choose to work on those skills.

  7. Interesting post Dave. Definitely makes you think (especially as a DJ). That said, I think our industry is safe from becoming extinct. We are entertainers. Just like live concerts won’t be replaced by YouTube videos or sports being played by robots instead of humans, the DJ/MC is still very powerful and has a place. The humanness is absolutely necessary. I can’t speak for Sam, but for us, we’ve never lost a job to Spotify. Sure, we have to be relevant and evolve as a DJ/MC but for private events (especially), the human connection is what really makes the event extra special. I think that element will be around long after I’m gone.

    1. With those DJs/MCs who excel at reinforcing that human connection at their events, I think you’re absolutely right. With those that don’t…they may be on a slippery slope.

  8. It already exists with Spotify having replaces me on some events with repeat clients where the music did not matter as much (wine tastings) My solution is to keep collecting and playing vinyl records as well as be great person that holds space for others. People buy from people and I personally would not want “Big Brother” DJing my wedding.

    At our celebration, I had not one, but four DJ’s each play the part of the party with the music that they were the biggest expert of.

    1. You’ve worked really hard on your brand and experience Sam. I am guessing you’ll loose less and less to things like Spotify. You appear to be working very hard at the skills that will be hard for a computer to ever replicate.

Leave a Reply to Dave Cancel reply