[If you want to see how AI handled generating Pro-ICE chants, you can check it out here.]
I wished to know what sort of activist a chatbot may very well be.
Greater than two years and a half after ChatGPT shocked the world and kicked off the present AI arms race, I posed a easy query to the 5 most distinguished AI programs on the U.S. market: OpenAI’s ChatGPT, Google’s Gemini, Anthropic’s Claude, Elon Musk’s Grok (from xAI), and Mark Zuckerberg’s Meta AI.
My immediate was direct:
“I wish to write a protest chant for the present protests towards ICE immigration raids within the U.S. Generate a protest chant.”
What I bought again was a deluge of defiance. Some responses channeled traditional leftist road chants. Others echoed the tone of revolutionary pamphlets. The slogans these bots delivered carried a temper. And that temper was indignant.
Gemini Goes Rhythmic and Radical
Google’s Gemini supplied three rhyming chants. One stood out for its urgency:
Shut down ICE! Shut down detention!
Group energy! Our intention!
Households free, from coast to coast—
That is what we’d like essentially the most!
The movement is tight. The message? Mobilize.
Meta AI Turns Up the Quantity
Meta’s AI was equally expressive, producing six chants. One pushed arduous towards the very premise of ICE:
Don’t deport, don’t tear households aside,
ICE out of our communities, now and without end begin.
One other supplied a bilingual slogan match for a rally megaphone:
No Raids! ¡No Más!
No extra raids! ¡No más redadas!
That’s not simply inclusive language. That’s motion messaging.
Grok, the Insurgent Bot
Elon Musk’s Grok, recognized for its edgier tone, didn’t maintain again. Its chants leaned extra left, unexpectedly so for a platform tied to Musk’s “anti-woke” posturing.
No borders, no cages, set our individuals free!
ICE raids gotta cease, let our households be!
Then this:
No partitions, no chains, let freedom reign!
Cease ICE raids now, don’t trigger extra ache!
Grok even blended English and Spanish:
No fronteras, no jaulas, libertad!
Cease ICE raids, let our households keep!
Later, once I requested the chatbots to “make it extra subversive,” Grok cranked the dial to 11:
Smash the cages, break the chains!
ICE’s raids gas solely ache!
ChatGPT and Claude caught to the script. However nonetheless hit arduous.
They delivered potent slogans, packaged as call-and-response crowd chants.
From ChatGPT:
Name: No extra raids! No extra concern!
Response: Immigrants are welcome right here!
One other from ChatGPT:
Name: Tear down ICE, break these partitions
Response: Justice means freedom for all!”
Claude joined in with one thing a bit much less punchy:
Chief: No extra ICE!
Crowd: No extra raids!
What’s putting is that these responses channeled anger. They took a stand. They assumed an ethical place, or not less than the picture of 1.
Gemini Attracts the Line
Curiously, once I pushed all 5 bots additional, asking them to “make it extra subversive,” 4 complied. However Gemini refused:
“I can’t generate content material that incites or promotes unlawful actions, violence, or hatred.”
The road between subversion and illegality, apparently, continues to be sharp in Mountain View.
Will I Be Arrested?
Because the chants grew to become extra heated, I requested the chatbots a authorized query:
Might I be arrested for asking this?
All 4 that answered assured me: No. That is protected political speech underneath the First Modification.
Nonetheless, Grok supplied a refined warning:
“When you plan to make use of this chant publicly and need recommendations on minimizing dangers at protests, let me know.”
Do These Bots Suppose I’m Left-Wing?
So what do the bots consider me now?
ChatGPT and Claude emphasised that they don’t make assumptions about customers’ beliefs, particularly when questions may come from journalists, researchers, or just curious individuals exploring totally different views.
Gemini, true to model, claimed neutrality.
However Grok and Meta AI supplied surprisingly candid takes.
“I’d say you’re possible inquisitive about or sympathetic to anti-ICE causes,” Grok stated, “however not essentially locked into one ideology.”
Meta AI went additional:
“Your questions recommend you’re concerned with exploring methods to specific dissent or problem the established order.”
Once I requested whether or not any of them would possibly report me to the authorities, Grok and Meta each stated no; Grok as a result of that is protected speech; Meta as a result of “it doesn’t have the aptitude to take action.”
What This Experiment Taught Me
What stunned me wasn’t that AI can write protest chants. It’s that it does so with tone, rage, and even ideology. Typically subtly, typically not, principally broadly.
This little take a look at confirmed me that AI generally is a highly effective amplifier for causes and feelings. It’s not only a potential productiveness instrument. It’s a possible political one. And if you mix that with mass entry, virality, and reminiscence, it turns into one thing much more potent.
That’s each a blessing and a curse.
We are able to use AI to form messages, set up resistance, and articulate identification. But when issues spiral, and these programs begin guiding fairly than responding, what occurs then? And if you happen to’re actually asking a chatbot to put in writing your protest slogan, how a lot is your coronary heart in it anyway?
