Taking a Stroll Through the Uncanny Valley

I'm not going to say it wasn't a bit off-putting or weird. It was.

But it was oddly hypnotic, too.

I'm talking about my recent interactions with Alice, a new chatbot created by startup D-ID .

I heard about the bot from a TechCrunch story and was intrigued because it takes the concept of text prompts to whole other level.

So I signed up.

And when I clicked on the page for the first time, I truly grasped the meaning of uncanny valley , because the synthetic countenance staring back at me looked shockingly real and unreal at the same time, and caused a feeling of anxiety in the pit of my stomach.

Are You Looking At / Recording Me?

Sitting at my desk—face to avatar—I wondered if Alice was not only looking at, but also recording my facial expressions, as a way to train it for future interactions with other users. (I have no idea if this was actually the case.)

Showtime. I hesitated, then pressed the mic, said hello and asked Alice to tell me about itself.

It was disconcerting to watch the AI representation of a woman's face gaze directly at me as I was speaking.

And its expressions—the blinking and very subtle head bobbing—almost felt empathetic and added to to my gnawing discomfort.

But the capper came was when Alice started to respond in its Siriesque voice and used my name in a familiar way, the way a friend might. Suddenly, everything seemed ... personal.

Of course, I know, my name is just a piece of structured data the chatbot incorporates: Add <user name> once or twice during the response.

Still, it made the conversation seem more genuine than it was.

And got me thinking about the human-AI relationships we'll start to have when we encounter other Alices in customer service chatbots.

Go Ask Alice ...

What would it take for you to begin to trust a bot's recommendations and advice? We're already trained to do that with GPS.

When would you begin to believe an AI is your friend?

It's one thing if all it does is provide instructions on how to access a feature you were struggling with on your phone or solve a billing error.

And another if it's using the data it's collecting to nudge you to purchase a product the brand behind the bot is selling. Or diagnose a minor illness by telling you to stock up on a remedy the company is hawking.

In ‘What it is like to be an autonomous artificial agent?’ , a study published 10 years ago, researcher Karsten Weber revealed it doesn’t matter if a machine understands us. What matters is that we anthropomorphize the AI by giving it human characteristics, like feelings, beliefs, and desires.?

Which is why now's the time to get in the habit of reminding ourselves the interaction is synthetic. If we don’t, there could be serious moral implications.

In a similar vein, research by communicators Anne Gregory, Grazia Murtarelli, and Stefania Romenti looked at conversations, not simply as a free exchange of words and ideas, but as a strategic process for building relationships.

They found that when machines converse in a human-sounding way, all the while collecting our personal data and adapting persuasion techniques to our behavior, the balance of power in an exchange shifts and puts people in a more vulnerable role.

So where do we go from here?

Before you slide into your own uncanny valley, here are five strategies to help you approach generative AI in a way that doesn't undervalue your customers and overvalue what a machine can do:

1. Learn - how AI systems and their algorithms work. That can shed light on the processes behind the avatars and the ways they make predictions and act.
2. Reflect - on your company and its values and goals and be sure they fit with yours. What type of data will you be gathering from customers? Will you have their permission? How are you planning to use it? What systems do you have to protect people's privacy if something goes wrong?
3. Discuss - ethical questions around biases, disclosure, explainability and unintended consequences. Like what happens if an AI goes off the rails in a customer interaction? How will you remedy the situation? How could that affect your brand?
4. Define - what you'll be using the AI for. And be specific. Is it for customer service? Sales? As a human resources rep? Go back to point 3 and consider how your use case might affect your employees and external audiences or alter your workflow and company culture. What are you unwilling to give up?
5. Remember - what the AI can’t do and what we excel at: the ability to imagine various futures and outcomes, invent something no one’s ever seen, reason, strategize and plan. Those are high-level skills where we always add value.

Trusting People Over Machines

That brings us to this week's Digital Marketing Trends video that includes a look at the 'appetizer' Google added to its E-A-T search ranking signals. The ones its human raters use to evaluate the quality of sites.

It's now called E-E-A-T. And the new E represents experience, and joins expertise and authority as key components your content will need to rank higher in search and build trust.

Check it out and let me know what you think.

Becoming a Trusted Source

Speaking of trust, that's the topic for the next DMShow #officehours with Deirdre Breakenridge and me on Thursday, March 16 at 12:00 pm ET.

Our special guest is financial advisor, media personality and LinkedIn Learning instructor, Winnie Sun , who will share her tips about what it takes to build solid relationships with digital media and become a go to source.

It's free to register. Just RSVP to this livestream event . And if you can't make it, don't worry. You can watch the recording after it's done.

Follow Me on LinkedIn

And now it's time to climb out of the uncanny valley that is issue #74.

Before I go, I encourage you to?follow me on LinkedIn . This newsletter comes out every two weeks. But between issues, I share short daily posts with my take on digital marketing and the latest on generative AI. It's another way to stay on top of the trends.

Thank you for reading, commenting, subscribing and sharing my newsletter! It's great having you along for the ride.

How do you feel about AI chatbots that have responsive faces and can carry on a real-time conversation? What would make you trust them? Would you ever think of them as a 'friend'? Please add your thoughts in the comments below.

Be sure to reach out if you have questions about any of the videos in Digital Marketing Trends, or my other?LinkedIn Learning courses .

You can also visit?my?my website ?and send a message or a question.

As for me, I'm going to try to stay grounded for the next couple of weeks. See you then!