Can AI Actually Enhance Communication?
Is it normal to automate the words you choose? For humans, I mean. I saw these robots pictured above many years ago at the 2002 Whitney Biennial and never thought I'd be writing a blog post about them 16 years later.
I guess artists (Ken Feingold) have a funny way of predicting the future, because nowadays I feel like I am not only talking to robots, I am letting robots talk for me.
I had a LinkedIn conversation recently which was initiated by my automated outreach message (via DuxSoup), but which got an equally automated response. I thought this was hilarious and commented on it. The response to that comment was, you guessed it, another automated response.
So the entire thread, minus my one real remark, resembled the conversation depicted above.
What is normal communication
Now is that normal? Normal is a funny word because it makes the bizarre acceptable.
For example, it used to be that it was normal to pick up the phone whenever it rung, but that's far from normal behavior now. Just because someone calls you, doesn't obligate you to answer.
It used to be normal to send a carrier pidgeon with a note written on a scroll. Or a raven, if you're a GoT enthusiast. But that's asychronous communication, which is becoming the preferred method.
Smartphones have predicting your messaging intention for a long time, but based on what you write, not based on someone else thinks you write.
What's normal changes every day, even if only a little bit. So let's look at what's becoming normal today, little by little, and make sure we're OK with it.
Gmail completing your sentences
I'll admit, sometimes Gmail gets it right. And yes, I have let Gmail complete some sentences on my behalf. I've let a computer program probably written by an engineering consultant in San Bruno, CA, speak for me.
The faster emails are written, the faster Google can run against them, so maybe there a financial incentive for them. But that's not why Google introduced the feature; it introduce the feature because it thinks it's becoming normal to let software speak for you.
You're probably thinking, "no problem, I determine what gets sent out. It's saving typing and if it's wrong, I'll just fix it.".
But that's now how language works. You are forever influenced by what you see.
The collective result is we're all communicating a little bit more like the Silicon Valley brogrammers, or just normal nerds. And not just because of Gmail.
LinkedIn speaks for you too
LinkedIn's communication automation is less subtle than gmail (and always seems to include an 👍 emoji). Nothing wrong with thumbs up - or any of these choices. The problem starts, though, when you stop thinking about what you say.
Each of these options seem to close the loop on this conversational thread. Do you want to close the loop? Nothing else to add?
And if you do want to close the loop, do you want to sound and feel like everyone else on LinkedIn?
As with the Gmail example, maybe this just saves typing; gets you through your day faster. Lke letting a car drive for you, so you can take a work call. But your unique expression is at stake here in a way that isn't part of letting a software program take over driving your vehicle.
A more intelligent example of communication AI: Crystal
Cystal App takes the automation of your self-expression one step further and suggests customizations to the way you write based on the DiSC personality profile of your recipient.
In the example pictured, the recipient has type "I", one of sixteen major personalityarchetypes that the DiSC system slots humanity into: "I", for Influencer.
Apart from being very interesting reading, knowing someone's personality profile is meant to help you organize your communication in the most effective possible way. This person, for example, prefers greetings like:
Not, to take an extreme example:
By the way, I was had an employee who communicated with everyone this way, Name:
"Rowan: thanks for the download." "Rowan: what are your long terms plans for the company?"
Efficient maybe, but certainly not personalized to either personality or context.
I'm all for personalization.
Crystal uses DiSC to advise you on every aspect of communication. Though as I allude to up above, you must strike a balance between your personality and that of your recipient. I would never use the sign-off, "Cheers", with or without an exclamation point.
To thine own self be true.
If you Google DiSC Personality assessments, you will find all kinds of stuff, including many tests.
And some businesses coaches administer the test. I personally have taken it in 4 different context, including through a business coaching program and through Crystal app, and the results are surprisingly consistent; not always the same type, but in a close range.
What sets Crystal apart is that instead of giving a test, it analyzes your digital presence (social media) to derive its DiSC profile assessment. Is it accurate? I think it's about 80% accurate based on my personal ability to assess personality profiles and on my read of digital presence content that I know Crystal can't access.
So as with LinkedIn and Gmail, Cystal might help. And it might actually be much more helpful.
But you can't let it write for you. The problem with automating language is that it's not a mechanical function; it's an expression of who we are.
As an aside, that's why it's impossible to separate out good identity copywriting (ie brand messaging) from strategy. Describing who you are as a business,=, the problem you solve, whom you solve it for – there's one, single most accurate way to do that. And it's going to read the same whether it's in a vaulted internal document or on the homepage of your website.
The idea of automating brand messaging then, is ridiculous. Thinking about the DiSC profiles of the average prospect might help. But if are humans and not robots, then we must be careful to think and speak for ourselves.
How about you, are you happy to let a robot speak for you?