Not Only Will Your Job Be Automated...

But so will any creative exploits you might have for fun, on the side.

In the wake of the recent U.S. general election, I was party to a conversation in which one of my friends, who works as an editor for a B2B publication, imagined a not-to-distant future in which editorial staff could be completely done away with.

Just point some AI at some sources of relevant data and information having to do with a particular sector, create some parameters around voice, style and such, and the AI does its thing.  To some, it might sound farfetched, but from what we know of content farms and fake news sites, much of the process of churning out countless bits of content designed to drive consumption - including catchy (and often misleading) headlines - is somewhat automated already.

There's another bit that made me feel like content creation isn't all that far from the days where it runs on autopilot with an editor keeping a watchful eye and making adjustments.

Recently, I upgraded my digital audio workstation to Logic Pro X.  This was an inevitable step on a path I started a long time ago, toward getting some of the musical ideas out of my head and turning them into actual songs.

With the upgrade came a bunch of cool new features, like you'd expect in a major upgrade of any type of software.  One particular feature, though, hit me like a ton of bricks.

You can virtually simulate a drummer.  No, I'm not talking about patching in the sound from some old drum machine and running repetitive loops to provide percussion over a song idea.  I'm talking about actually simulating a drummer - that oft-maligned member of the band who seemingly everybody likes to bust on.

Logic Pro X lets you pick who should play drums on your song.  The virtual drummers have names, not to mention styles, influences, and particular drum kits that they like.  The kits are all impeccably sampled, piece by piece, from actual drum kits.  The drummer I have in the song I'm currently working on is named Logan.  He's more of a classic rock guy, sort of heavy-sounding, kind of in the vein of a Keith Moon or a John Bonham.

The thing about Logan is that he's completely obedient.  He always shows up for band practice, and I can control everything about how he plays.  The best part about Logan is that he doesn't sound robotic - I can control how far back he hangs on the beat, whether he rushes the band at any point, how likely he is to play fills, whether he's playing the high-hat or the ride cymbal during a particular passage.  I can tell Logan that I hate his kit, and I can change it around, or swap in tom-toms from a completely different kit if I don't like the way he sounds.  I can even make him make mistakes.  If Logan pisses me off, I can fire him in the middle of a song if I feel like it, keep his recorded drum parts, and swap in another drummer.

Logan's big selling point, though, is that he sounds like a human.  To the casual listener, he sounds like some guy I hired for a recording session.

I've never been able to play the drums.  And now that we've automated the drummer, I'll never be left waiting for a drummer who blew off band practice, or have to argue with one who keeps rushing through the bridge.

This leaves me time to concentrate on other things, like arranging, playing guitar and piano parts, maybe doing some singing.

Speaking of which, my guitar skills have really deteriorated, and there's this song I can hear in my head.  I spent an hour the other day trying to play the main riff and I just couldn't get it - there's one part where the string-skipping pattern is different from the rest of the riff, and I always blow it.

After an hour, I landed on a conundrum.  I could spend the next hour or two trying to nail the riff perfectly, or I could play it once the wrong way, edit the hell out of it so that it sounded like I was playing it the right way, and move on.  Guess which one I opted for.

Whenever I've used recording software over the years, I've been able to justify that songs I wrote with it were entirely my own creation, because at least I could play all the parts.  Now I'm not playing the drums (which is probably a good thing), and I'm fixing my guitar parts that are too complex for me to play without a lot more practice.  Is the song really my own anymore?  Does anybody really care?

I'm sitting here, working on a song, using a computer program that takes the place of hundreds of thousands if not millions of dollars spent in mixing consoles, vintage amplifiers, synthesizers, samplers, sequencers, etc. and I'm composing a song I can't play.  There's no longer a capital investment barrier to doing what I'm doing, and the skill barrier is getting lower and lower every time I upgrade my software.  I'm using virtual musicians that don't exist, but obey my every command.  Explain to me how songwriting and recording is a unique skill again?

Now, let's talk about how I learned to use the software.  Used to be I'd have to take a few hours a week, haul myself off to some music production school in Chelsea, pay a bunch of money and sit through classes.  Now, I'm learning from a subject matter expert through a video series on Lynda.com.  I download these videos or stream them to any one of my devices, watch them at my own pace, and take quizzes after I'm done with each section of the course.  My progress follows me from device to device, so I can watch them seamlessly whether I'm on the train or in bed at home with my iPad.  The quizzes make sure I'm absorbing the material.

So, in learning a software that automates the jobs of a producer, studio engineer and even a couple musicians, I used a learning resource that automates the teacher.  Let's let that sink in for a moment.

Over the years, I've been skeptical of automation, especially as it pertains to creative endeavors.  Here are some statements I formerly believed in, but don't any longer:

  1. A computer can't take the place of a musician and won't ever be human-sounding enough to be convincing.
  2. No digital modeling system will be able to capture the sound of a properly-miked tube amplifier.
  3. It's more important to writing songs to know how to play than to know the range of possibilities that might make a song sound good.
  4. Watching pawn shops and CraigsList for people letting go of great-sounding vintage instruments and equipment is worthwhile if you want to capture the unique sounds that equipment can make.
  5. AI is just incapable of making many of the judgment calls we make as creative people who are capable of writing and performing great songs.

While I'm excited at the possibilities and am looking forward to turning a lot of great song ideas I've had over the years into real songs, I also feel like a career factory guy who just got replaced by a robot.

I took 15 years of classical piano lessons as a kid.  I spent countless hours sitting on my bed with a cheap guitar trying to learn how to play songs I liked.  I spent the better part of my late teens and twenties trying to scrape up enough money to buy Les Pauls, Fender Strats, synthesizers, stomp boxes, amplifiers, vocal processors and much, much more.  I took computer programming classes from the time I was in the third grade all the way up to undergrad.  In summation, I've invested a ton of time and money so that I could have a side pursuit of writing and recording music, and automation has made a lot of it irrelevant.  It's hard to not be infuriated by the notion that someone could decide to take the Lynda.com course today, get really good at using Logic Pro, and produce something incredible-sounding without having to make any of those investments along the way.

So I've been disrupted.  Thankfully, I can adjust.  I can get really good at using the software.  I can put ideas to songs that (hopefully) no one has thought of before.  The possibilities are exciting, but the notion that the barriers of entry have been lowered is scary.  Scarier still is that I can see the day when AI composes compelling works and starts to do it better than humans can.

More broadly, it's becoming a lot easier for me to see a lot of creatively-oriented jobs being eliminated - editors, writers, music producers, artists of all stripes.

What happens to our society when the contributions of the people who work in these fields are devalued?

Let's Return to a Simpler Time

Let's return to a simpler time, when adults didn't ask one another what they planned to do in the voting booth.

Just eight years ago, I would have disagreed vehemently with that suggestion.  I would have considered it every American's responsibility to contribute to the Marketplace of Ideas and get their thoughts out there.  That was when I believed we made decisions as rational beings that consider the facts, weigh them and come to a decision they can support.

But we don't make decisions like that.  At least, most of us don't.

We make decisions emotionally, and then surround ourselves with hand-picked facts that support our position, and challenge things that we consider affronts to our position.  We conveniently ignore facts that don't fit our worldview.  And, worst of all, we judge and demonize people who don't come to the same conclusions we do and wonder how they can come to their conclusions rationally when in possession of the same set of facts we are.

But if we understood that our decision-making process - for most people, anyway - doesn't follow the process we've set out as an ideal, we wouldn't be so judgmental.  We wouldn't be eager to end friendships, or silence voices that give weight to opinions that we don't currently hold.

We fail to understand that no matter who we vote for, some significant percentage of our friends and acquaintances are going to become so emotionally charged and will be so unable to understand your choices, that they will lose respect for you.

They'll block you in social media, cancel plans you have together, silently cross you off the consideration list for a piece of business they were thinking about giving your company, or decide to no longer trust you.

I've not posted publicly in social media about any of the candidates this election cycle.  If anything, I've expressed a certain awe at the strategies being employed - good or bad - and marveled at how they've changed the playing field over time.  (I've always been interested in strategy.)  I've appreciated those things for their strategic value and like thinking about how they'll affect the race.

But I've not publicly committed to a candidate, nor have I gotten involved in the typical arguments you tend to see on social media.  For my political safety valve, I've selected a group of trusted and open-minded friends, and I've kept my comments limited to discussions with them.

Why?  Because if you accept the notion that maybe people don't make decisions the way we might have thought they did, you begin to understand how little effect the discourse actually has.  You begin to look at the risk equation, realizing that every public post, every overheard comment and every opinion that travels about the upcoming election is more of a liability than a credit.

After all, if most people are surrounding themselves with their own supporting facts, and largely ignoring what doesn't fit their worldview - if decisions are made emotionally rather than rationally, trying to change someone's mind is a lot like trying to tell them how they should feel when they fall in love.

I'm not up for that, particularly when showing a bias one way or another would cause me to lose friends I need.

You probably disagree.  You probably think I'm cowardly.  But consider this - I value you and your friendship more than I'm prepared to risk it, simply because we might have different versions of the facts, or different emotional drivers that would cause us to vote one way versus another.

So, let's go back to that period our parents or grandparents lived through, in which people kept their voting booth behavior close to the vest, and we weren't so eager to try to change one another's minds at all cost.

Does Digital Drive Marketing's Need to Violate Privacy? Not Always.

The idea that advertising is wasteful is nothing new.  If you asked anyone on the street to come up with a notable quote about advertising, I’d imagine most would respond with the old John Wanamaker quote – “Half of the money I spend on advertising is wasted; the trouble is, I don’t know which half.”

It’s that kind of thinking that leads marketers to believe that something is inherently wasteful about their advertising, and if they could just employ some sort of method that would allow them to reach only the people who are interested in buying their product, that they would be perfectly efficient with their advertising.

The perfectly-targeted ad campaign, though, is a red herring.  If your advertising reached only the people who were ready to buy your product, your campaign would fail to inform others of your product and you wouldn’t capitalize on mass advertising’s ability to build awareness, such that people who don’t yet need your product won’t turn to a competitor when the need for your product does arise.

I’ve talked about it until I was blue in the face.  Still, this idea persists that micro-targeted ads are the holy grail of advertising.

That belief, though, predates the current trend toward investing in data to target digital ads.

I’ve said it before and I will say it again.  The foundations for digital’s ability to target advertising came from the traditional direct marketing business, the practices of which would completely creep out consumers if they had the slightest inkling of what was going on.  Your purchase behavior, political leanings, personal income, preferred brands and much, much more were being used to target direct mail long before digital ads.  In fact, the first open consumer revolt against DM-style targeting online came when DoubleClick tried to merge with Abacus, a catalog marketing outfit.  People were seemingly okay with offline DM, but didn’t want digital ads to take the same approach.

Don’t mistake this for yet another complaint about the double standards digital has had to endure over the course of its development.  That’s not where I’m going with this.

Last week, though, Doc Searls took issue with addressable cable ads.   I had a couple problems with the piece he wrote.  Laying blame at the feet of digital for addressable cable ads was one of them.  While digital tech can help Big Cable handle the backend intricacies of inserting addressable ads, the method was an idea long before DMPs, DSPs and digital profiles.

If you feel sold out by your cable or telecom provider, know that they’ve been using your information in this way for a long time, and not just in digital ad applications. 

It’s easy to be angry with digital and want to place blame with it, though.  When the aforementioned DoubleClick/Abacus merger caused controversy, digital made the assumption that people didn’t like central databases of consumer information being used to target ads.  So digital just created something that wasn’t centralized.

But the effect is the same that a monster, centralized database of marketing information would have had.  Nowadays, if a company wants to use data to target digital ads, it can simply broker a deal for the information it needs, have a third-party partner match up its own list with cookies or other unique identifiers, and – voila! – the company has its list of ad targets without having to handle any of the Personally-Identifiable Information that would trigger controversy for stepping over the arbitrary line that the industry drew for itself.  I’ve been trying to draw attention to this practice for years, calling it a distributed violation of our collective privacy.

So, Doc asked me on Twitter where I’d place the blame for the enthusiasm for things like addressable cable ads.  First, I’d place them with cable companies.  Secondly, I’d place them with marketers and their agencies that continue to buy into the Church of Wanamaker, and think they’ll achieve marketing nirvana by continually whittling down the size of the audience they think will be interested in what they’re saying (in the interest of “efficiency”).  Thirdly, I’d blame the behind-the-scenes “match partners” that do all the PII dirty work, regardless of the medium being used.

Killing the Tax Cap and Gap Elimination Adjustment

Yes, I'm serious.

Our costs are growing faster than we can raise taxes to make up the difference.  Albany has all the power, since they're making the decisions on state aid.  This is the new normal.

Last month, I saw this.  And it's an example of what the new normal looks like over time.  School districts will be making drastic cuts, but the State Regents will swoop in and save us when key elections approach.  It's an effort to make us forget our struggles on our way to the ballot box.

The re-establishment of local control is a noble goal.  Our kids can't be short-changed because Albany can't balance its budget.  And communities shouldn't be in the position that they have to beg in order to keep their school districts solvent.

Some neighbors and I have something cooking to address this, at a level that's bigger than just our local school district.  And we're looking for people who share the vision of returning control of school districts' destinies to communities.  If you're interested in getting the scoop, hit me up in comments.