Videos you watch may be added to the TV's watch history and influence TV recommendations. To avoid this, cancel and sign in to YouTube on your computer.
Share
An error occurred while retrieving sharing information. Please try again later.
> Create fake democrat/republican > Make them say outrageous shit > Have real people go apeshit over it with consequences > This dude never existed, but this problem now exists.
The concept of being gang-stalked by AI and being bullied by them is absolutely terrifying and hilarious at the same time. Is this a Black Mirror episode?
It's funny remembering that kojima basically predicated all of this in MGS2, Ai generated content, unfiltered and random, filtered and specific, all made to numb and confuse the masses. The perfect weapon, Metal gear is an AI!
Imagine being a YouTuber or streamer and decades after you're dead, AI is used to parade a digital version of your corpse around in an attempt to create entertainment. Gives me the chills.
When ai art is indistinguishable from photos photograph evidence will no longer be useable in court and we won't be able to know anything with certainty.
Everyone please take a moment to think about how sophisticated some of these bots, and A.I generated content are, and then remember the abysmal state of the A.I in Star Wars Outlaws. It's so fucking sad!
So, I just got a really fucking TERRIFYING idea...Imagine when fully autonomous androids become available. Now imagine those same people who employ bots to harass people, get black market androids to do that in real life, except their eyes have live-streaming cameras, a voice box which can deepfake mimic any voice, they don't need to sleep or eat, probably have a battery life that can last for several days, and they can just stand outside your house in the middle of the night, waiting till you go to sleep, to just sneak inside your house with perfect ease, and then end you.... Yea, I'm getting deadbolts for every single door in my house.
THAT'S NOT DESTINY. Destiny's body was just discovered at his home 3 days ago. We didn't know because his codec link with the server was cut off, but Mei Ling said his transmission signal was coming from *inside the base*.
Egon Chalokian claims to be a particle physicist and a tax analyst and an educator of intelligence agents. He's claimed to have studied at Harvard and worked at both NASA and the European Centre of Nuclear Research, as well as US government agencies. You'd think someone with such a CV would have more of a footprint than he does.
I saw this shit happening live. It was legit horror, it started with the Twitter reactions that seemed too artificial, then came the posting of ai generated images and then he went into the entire Igon Chelokian stuff and trying to figure out wheter he was real or not. It was eery as fuck. It was like you are sitting in a public place but you're 100% sure that everyone you see is not real and they're imitating human behaviour by reacting of eachother to try and fool you specifically. Actual schizo situation but in this case it's real.
The biggest giveaway that the 3 hour video is AI, is that it's a 70yo grandpa sitting still for literally 3 hours not moving. Also talking straight for 3 hours without drinking. No human can do that.
I am quickly losing confidence in the ability for us to recognize the advancements in AI. These errors of the repeating figures is NOWHERE near as noticeable as this stuff was a few years ago. This is terrifying
To add on the misspelling part. When most AI languages parse data and responses they usually convert/encode it into number sequences and search for similar results and mash them together to get semi-coherent answer. There is the whole "Strawberry" thing where most AI Language Models could not tell you how many R's are in the word even though they can easily spit the word back at you spelled correctly. So basically your question could be composed of well thought out string of words but the AI sees [35432, 321 ,86515, 4841, 5468, 88499541, 8548, ...etc.] and inputs it into their models and spits back out similar code to then get decoded again. There IS Information loss even though most of it may seem coherent. It is fairly easy to see in how AI's use diffusion models to create Images. It's why people end up getting extra goofy fingers, off shape extremities, or weird bends. You also see finer detail being brushed over or a lot of inaccurate or just missing text. Very similar to the beginning of the video where there were missing or lack of legible text of the books and goofy shaped people in the pictures. Sauce: I am a Software Engineer with very minimal knowledge about AI Machine Learning.
"You can't read when you dream" "In dreams you have extra fingers when you count them one by one" "When is a door, not a door?" - Your reality is a-jar.
0:10 Audio engineer here. You can use two mics in that configuration to create what is called a Cardioid Polar Pattern with two Omnidirectional microphones. This makes the opposite end of the mic’s capsule reject noise by using correct phase and time alignment doing it this way. It’s a technique used quite often. It does have the added benefit of being a backup mic.
When the FBI planted wires on the mob back in the 80s on targets who knew they were being monitored, they'd ensure some of the wires would be found to give false assurance, and prevent the target from looking for the ones that weren't meant to be found.