The video correctly identifies that blind trust in AI is a modern intellectual liability, even if the "Orwellian" framing feels a bit hyperbolic. Fact-checking remains the only reliable antidote to the confident hallucinations of large language models.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
This is Orwellian
Added:[snorts] >> Hey everyone, welcome back to our channel. If you're new here, please subscribe. I'd really love that. If you're returning, how are you doing? So, I decided to do a test.
And I'm going to show you what I'm talking about, okay? But I decided to test Google to see if they would admit to something.
And their AI lied to me.
And you're probably going to say, "Are you shocked?" No. Well, yes and no.
I asked them a question, okay? And I'll show you everything, okay?
And they they claimed that it was an ideology, not an an organization, and I told Gemini that it was wrong.
And then it admitted that it was wrong, and it said "accidentally used outdated information."
Now imagine that Gemini said that it failed to check 2025 data.
How convenient is it that it misses information that's right there?
So, here you go.
So, >> [sighs and gasps] >> is Antifa a terrorist organization?
It says, "Under US law, Antifa is not designated as a foreign or domestic terrorist organization. While various political figures have called for its designation as a domestic terrorist organization, and executive actions have federal agencies to investigate its networks, legal and structure factors complicate the label."
Um it says, "Structure and definition, an ideology not an organization. Federal law enforcement including the FBI characterizes Antifa as a loosely lit uh, knit [snorts] uh, movement or ideology rather than structure of hierarchical uh, group with a central leader.
And I said, well, that's blatantly false because the president labeled them as domestic terrorist organization via the EO.
And he's uh, uh, it said, well, yeah, you are correct. I apologize for that error.
President Donald Trump signed executive order on September 22nd, 2025, um, officially designating Antifa as a domestic terrorist organization. The 2025 executive order uh, states this. Official action signed during Trump's second term explicit label uh, formal formally labels Antifa a terrorist organization. Federal mandate orders agencies to investigate operations. Financial focus demands dismantling of funding networks. And I said, so you lied to me. You were instructed to lie?
It said, no, I was not instructed to lie. My first response was a mistake. I relied on outdated information. As an AI, I have no personal motives. I am never instructed to deceive anyone.
I said, and it said, why the error happened? Old legal frameworks. I relied on historical legal definitions missing recent updates. I failed to check 2025 data. Why? It was just a few months ago.
Why would you fail to check something that's just a few months ago? Why would you fail?
Right? So, overlooked directive. I missed the White House presidential action from December 22nd, 2025, correcting the record. User prompt, your correction forced a recheck.
So, what is your take? Do you think that Google is running cover for left-wing domestic terrorism? Or do you think that they are >> [sighs and gasps] >> I don't know.
What's How can I put this?
>> [snorts] >> Could like be like a Cuz I don't know, maybe it doesn't get updated as much. I don't I don't know.
Like a a lot knowledge lag or It could even be a bias in their training data. I don't know.
I It's just it's very strange to me because you know, when I'm doing research and stuff like that, it's not very often that I will, you know, use anything, you know, other than my own wit and my own intelligence and stuff like that and my own reading and abilities and stuff like that to find out what I need to find out while doing my research. But you know, I have noticed that when AI is used, they can be How do I put this?
programmed, if you will, to be overly cautious or politically calibrated in how it presents sensitive topics, leading it to stick with the safe narrative or an old narrative until it is directly changed or challenged.
Um So, I don't know. I mean, that could very well be that's what happened.
Um But I actively use search tools on current topics, controversial topics.
So, you know, I I I use my own way of doing my research.
And there's a lot of places that I do go to for my research. I will say that.
And it's not easy what I do. Um sometimes it works out, sometimes it doesn't.
Um but it's obvious that AI is not perfect.
>> [laughter] >> They're not always going to give you the answers.
Um especially when it comes to uh politics, okay? Because I guess whoever's running the AI, they could have their own political bias behind it. You just don't know.
It's why you have to double-check everything.
You really do. You really truly do. You have to You have to fact-check the fact-check. You know what I'm saying?
All right, guys. Maddie here. I will see you in the next one, okay? You stay safe, you stay positive, you keep prepping. And as always, fearless and God bless.
Ciao.
Related Videos
AI Agent Mastery Certification Course: Lab 4 – Tools & MCP
arizeai
350 views•2026-06-16
Real-time Voice cloning, Kimi K2.7 CODE, GLM 5.2 and 3D reconstruction | AI News
kaiexplainsYT
111 views•2026-06-16
He Believes AI Could Replace Humanity Faster Than Anyone Expects
LondonRealTV
815 views•2026-06-15
General Session by Rami Rahim-The next generation of networking: From vision to self-driving reality
HPE
108 views•2026-06-17
[PLDI 2026] Flatirons 3 - LCTES (Jun 16th)
acmsigplan
191 views•2026-06-16
Google DeepMind’s AI Halves UK Housing Planning Time
60secondsignals
467 views•2026-06-17
The Creators of Claude Code and OpenClaw don't Prompt Their Agents Anymore?!
ColeMedin
569 views•2026-06-18
Why prompt injection is AI's biggest fail
usemultiplier
1K views•2026-06-17











