Should We Be Concerned About AI-powered Subliminal Messages?

Some studies in the past have reported that subliminal messages possess the power to exert a subtle, yet quantifiable influence over human behavior.

What would you say if I told you that the text and even some of the images in this post may convey a subliminal message? Subliminal messaging involves using stimuli that are presented just below the threshold of conscious perception for humans. This means that the audience may not be consciously aware of the underlying message, but their brain might still process it. It’s a bit like a message hidden within a message.

Interestingly, many of you may not be aware that subliminal messaging has been employed in advertising, marketing, and even politics for quite some time, although it has been banned in certain countries like the US.

This topic is undeniably reminiscent of something you’d find in a covert operation conducted by a spy agency. It’s a controversial subject, with some individuals arguing that this form of communication is unethical and manipulative.

When Did Subliminal Messaging Hit Mass Consciousness?

The concept of subliminal messaging first gained significant attention during the 1950s, thanks to market researcher James Vicary, who claimed to have boosted popcorn and Coca-Cola sales by flashing subliminal messages on a movie screen. Although Vicary’s assertions were later discredited, they triggered a surge of interest in the field of subliminal messaging.

Over the years, extensive research has been conducted on the subject. Some of these studies have identified subliminal messages’ modest yet measurable impact on human behavior. For instance, one study revealed that individuals exposed to subliminal food-related messages were more inclined to consume snacks later in the day. Another study found that exposure to subliminal messages promoting a particular brand increased the likelihood of choosing that brand when making purchase decisions. However, it is essential to emphasize that the effects of subliminal messaging are typically subtle and not consistently replicable.

So why am I telling you all of this here? Subliminal messaging is not a secret. Any half-decent advertiser shall tell you it’s old news. But there are a couple of reasons why it has excited people again. Of course, since the stated mission of this newsletter is reporting from the crossroads of content, marketing, and tech, I thought of talking about this trend here.

SM has been talked about now because of artificial intelligence (AI); more specifically, generative AI. This tech is now being used in creating subliminal messages in a number of ways. For example, AI can be used to create subliminal messages that are tailored to the individual viewer. This can be done by analyzing the viewer’s past behavior and preferences.

EXAMPLE:

Baskin Robbins’ original 31 flavors are well-known. Observe how the conclusion of the “B” and the commencement of the “R” within the logo above artfully shape the numerical representation “31.”

We find ourselves once again at the start of a buzz over subliminal messaging. I came across a couple of articles that alerted me to it.

This one tells you how to create an SM using AI.

A second piece in Vice.com talks of why all the concern now.

The Vice report talks of a particular technique that’s captured the World Wide Web’s attention, allowing users to create optical illusions through AI-generated images “subliminal messages”. This technique, says the report, is known as ControlNet, and offers users a greater degree of control over the generated images by allowing them to specify additional inputs. Essentially, it empowers individuals to craft images or embed words within other images, leading some to characterize it as a means of conveying “hidden messages.” These concealed messages can take the form of subtle visual cues, such as a McDonald’s “M” logo subtly integrated into the contours of a movie poster.

Reportedly, the buzz surrounding ControlNet escalated rapidly on social networks, including X (formerly Twitter). Some hailed it as a revolutionary innovation, while others viewed it with suspicion, considering its potential for nefarious applications. Some also began sharing various images featuring these intriguing “hidden” messages. Others have cautioned against sensationalism.

ControlNet utilizes Stable Diffusion, an AI image-generating tool, and one of its initial applications involved the creation of elaborate QR codes, according to this article. A user’s experimentation led to the discovery that using the QR Code ControlNet and substituting a QR code with different black-and-white patterns could produce captivating optical illusions. Shorn of all the mumbo jumbo, it basically means that a pattern can be slipped into the main image without detection.

So, one can now understand what all that fuss is about. Visions of an Orwellian State or a dystopian society because of the misuse of such subliminal messaging have already started doing the rounds.

On a more down-to-earth note, there’s also the growing public concern about the potential negative consequences in the marketing or the sale of goods. For example, some people are concerned that subliminal messaging could be used to manipulate people into buying things they don’t need or want.

Here are some specific examples of how AI can be used to create subliminal messages:

• Targeted advertising: AI can be used to create subliminal messages that are tailored to the individual viewer. This can be done by analyzing the viewer’s past behavior and preferences. For example, an AI-powered advertising platform could show viewers a subliminal message about a new product they are likely to be interested in.
• Deepfakes: AI has the capacity to produce deepfakes, which encompass altered videos or images portraying individuals engaging in actions or uttering statements they never genuinely performed or said. Deepfakes can serve as a vehicle for concealing subliminal messages, wherein clandestine messages are seamlessly incorporated within the video or image content.
• Virtual reality (VR) and augmented reality (AR): VR and AR can be used to create immersive experiences that can be used to deliver subliminal messages. For example, an AI-powered VR app could create a virtual environment that is designed to make the viewer feel happy or relaxed. The app could then embed subliminal messages in the environment to influence the viewer’s behavior.

Positives and Negatives of Subliminal Messaging

On the positive side, subliminal messaging could be used to promote healthy behaviors, such as eating healthy foods or exercising regularly. It could also be used to educate people about important social issues, such as climate change or racism.

For example, one study found that people who were exposed to subliminal messages about the dangers of smoking were more likely to quit smoking.

On the negative side, subliminal messaging could be used to manipulate people into buying things they don’t need or want. A study found that people who were exposed to subliminal messages about violence were more likely to behave aggressively.

Can Subliminal Messaging Be Used to Take Control of Someone’s Emotions?

It is unlikely that SM can be used to take control of someone’s emotions completely. However, there is some evidence that subliminal messaging can influence people’s emotions to a certain extent.

For example, one study found that those people shown subliminal messages designed to make them feel happy were more likely to report feeling happy afterward. Another study found that those shown subliminal messages designed to make them feel angry were more likely to report feeling angry afterward.

However, it is important to note that the effects of subliminal messaging on emotions are often very small, and they are not always consistent. Additionally, people are more likely to be influenced by subliminal messages if they are already predisposed to feeling a certain way.

Want to be more engaged with AI? Subscribe to our 2nd newsletter “AI For Real”.

Is the Concern Over AI-generated Subliminal Messages Justified?

Let’s not forget that the use of AI to create subliminal messages is still in its infancy. So, let’s not panic. At the same time, let’s also not lose sight of the fact that AI has the potential to be used to create subliminal messages that are more effective and more difficult to detect than ever before. That’s why so many people have started talking about subliminal messaging now. The concern and the justification do exist. It could also be used to spread misinformation or propaganda. For example, one study found that people who were exposed to subliminal messages about a particular political candidate were more likely to vote for that candidate.

In Conclusion

Subliminal messaging is not only a complex but also a controversial topic. With the introduction of AI, it is natural for concerns to be raised, once again. There is some evidence that subliminal messaging can have a small but measurable effect on people’s behavior and emotions. However, it is important to note that the effects of subliminal messaging are often very small, and they are not always consistent. Additionally, some people are more likely to be influenced by subliminal messages if they are already predisposed to feeling a certain way. The introduction of gen-AI in the mix needs to be monitored, and perhaps then regulated, to guard against misuse.

(A confession: Some help was taken from a machine to write/re-write bits and portions of this newsletter.)

References

• “Subliminal Messaging: A Review of the Evidence” by Wilson, Smith, and Graham (2013)
• “The Effects of Subliminal Messages on Consumer Behavior: A Review of the Literature” by Van Rompay, Van den Bergh, and Moons (2010)
• “Subliminal Messaging: Can It Control Our Minds?” by the BBC (2017)
• “The Power of Subliminal Messaging: How It Works and How to Protect Yourself” by Psychology Today (2022)