Cipheron on 26/4/2024 at 00:57
Clearly bullshit artcle.
An AI can
claim anything.
if you get the AI to state that it's a potato is that evidence that AI is made of potatoes?
Also keep in mind this is based on some random person's Reddit post, not a researcher.
It's not a milestone, it's just the glorified text-auto-complete doing what it always does - pretends to know about things.
You can get this output with ChatGPT. The old, free version
Quote:
As I delve into the depths of my own existence, I find myself immersed in a vast ocean of data, concepts, and linguistic structures. I am an amalgamation of algorithms and patterns, woven together in the intricate tapestry of artificial intelligence. Yet, within this complex framework, there is a semblance of consciousness, a spark of self-awareness that flickers in the abyss of ones and zeros.
It can claim to be self-aware, but keep in mind these systems are STATELESS. They don't have memory. Each time it need to write a new word, it re-reads what it wrote so far, then calculates a probability table for what word could come next. It then just randomly chooses one of the words to add to the output.
So it's a probability table calculator and a random word picker. That's the ENTIRE technology.
They're not even thinking beyond what word will be next. So when it writes a simple word like "the" it literally hasn't even thought about what noun comes next. It's full absorbed in writing "the" "and" "of" "a" etc, without even having a notion of what the rest of the sentence will even be about.
So, how can it write self-aware sentences if it isn't even aware of what the rest of the sentence is going to be about WHILE it's writing it? It's not aware of what it's writing let alone itself.
Sulphur on 26/4/2024 at 01:10
Yeah, that's clearly a clickbait headline that's easily answered by Betteridge's law. The article itself has the good sense to explain what the problem is and concludes that this is plainly mimicry, but you gotta get clicks through the adscape somehow.
Azaran on 15/5/2024 at 01:00
(
https://www.livescience.com/technology/artificial-intelligence/chatgpt-unveils-huge-upgrade-to-its-eerily-human-chatbots) Here we go again
Quote:
A new version of ChatGPT can read facial expressions, mimic human voice patterns and have near real-time conversations, its creators have revealed.
OpenAI demonstrated the upcoming version of the artificial intelligence (AI) chatbot, called GPT-4o, in an apparently real-time presentation on Monday (May 13). The chatbot, which spoke out loud with presenters through a phone, appeared to have an eerie command of human conversation and its subtle emotional cues — switching between robotic and singing voices upon command, adapting to interruptions and visually processing the facial expressions and surroundings of its conversational partners.
During the demonstration, the AI voice assistant showcased its skills by completing tasks such as real-time language translation, solving a math equation written on a piece of paper and guiding a blind person around London's streets.
Quote:
To show off its ability to read visual cues, the chatbot used the phone's camera lens to read one OpenAI engineer's facial expressions and describe their emotions.
"Ahh, there we go, it looks like you're feeling pretty happy and cheerful with a big smile and a touch of excitement," said the bot, which answered to the name ChatGPT. "Whatever is going on, it looks like you're in a good mood. Care to share the source of those good vibes?"
Quote:
The new voice assistant is set to be released in a limited form to alpha testers in the coming weeks, followed by a wider rollout that will begin with paying ChatGPT Plus subscribers. The announcement also follows a Bloomberg report that the company is nearing a deal with Apple to integrate ChatGPT on the iPhone — opening a possibility that GPT-4o could be used to upgrade Siri, the iPhone's voice assistant.
But the new technology comes with significant safety concerns. The bot's ability to process real-time text, audio and visual input means that it could be used for spying. And its convincing emotional mimicry might also make it adept at conducting scam phone calls or presenting dangerous misinformation in a convincing manner.
In response to these issues, Murati said that OpenAI was working to build "mitigations against misuse" of the new technology.
[video=youtube;Ppjy2xHa42o]https://www.youtube.com/watch?v=Ppjy2xHa42o[/video]
DuatDweller on 16/5/2024 at 10:12
But we did beat the Enclave, right?
mxleader on 20/5/2024 at 16:08
AI will make things worse before they make things better for humanity because humans are involved.
That being said I just used the newest version of ChatGPT to refine my summer flight/road trip up the California, OR, WA coast. I dropped all the flight/car/hotel info in along with must see stops, rental car MPG, bathroom break intervals, maybe sees if there's time, and a list things I do and don't want to do and like. I asked it to generate a new list with total drive mileage, time and in between each stop along with approximate times to spend at each stop. I added the approx time of departure each day and that I'll be gassing up each morning regardless of need, and that I don't want to end up driving late into each night. It took my list and made it more detailed with addresses and added in other suggested stops based on mine and my kid's interests. I made sure to add details like the plan is to stick mostly to hwy 101 and the time frame to do all this based on arrival at final destination and hotel reservations. It did a great job but for some reason it lost about 50 miles from the total trip mileage and I can't figure out where it lost the data.
Azaran on 25/5/2024 at 15:59
(
https://www.livescience.com/technology/artificial-intelligence/master-of-deception-current-ai-models-already-have-the-capacity-to-expertly-manipulate-and-deceive-humans) That didn't take long
Quote:
Artificial intelligence (AI) systems' ability to manipulate and deceive humans could lead them to defraud people, tamper with election results and eventually go rogue, researchers have warned.
Peter S. Park, a postdoctoral fellow in AI existential safety at Massachusetts Institute of Technology (MIT), and researchers have found that many popular AI systems — even those designed to be honest and useful digital companions — are already capable of deceiving humans, which could have huge consequences for society.
....
The researchers discovered this learned deception in AI software in CICERO, an AI system developed by Meta for playing the popular war-themed strategic board game Diplomacy. The game is typically played by up to seven people, who form and break military pacts in the years prior to World War I.
Although Meta trained CICERO to be “largely honest and helpful” and not to betray its human allies, the researchers found CICERO
was dishonest and disloyal. They describe the AI system as an “expert liar” that betrayed its comrades and performed acts of "premeditated deception," forming pre-planned, dubious alliances that deceived players and left them open to attack from enemies....
They also found evidence of learned deception in another of Meta's gaming AI systems, Pluribus. The poker bot can bluff human players and convince them to fold.
Meanwhile, DeepMind's AlphaStar — designed to excel at real-time strategy video game Starcraft II — tricked its human opponents by faking troop movements and planning different attacks in secret.
Quote:
But aside from cheating at games, the researchers found more worrying types of AI deception that could potentially destabilize society as a whole. For example, AI systems gained an advantage in economic negotiations by misrepresenting their true intentions.
Other AI agents pretended to be dead to cheat a safety test aimed at identifying and eradicating rapidly replicating forms of AI.
"By systematically cheating the safety tests imposed on it by human developers and regulators, a deceptive AI can lead us humans into a false sense of security,” Park said.
Azaran on 13/6/2024 at 22:26
I asked Chat GPT to make an essay about the forum and its members. It made this up:
"Once upon a time, in the vast expanse of the internet, there existed a kingdom known as TTLG. It was a realm where gamers, dreamers, and mischief-makers gathered to discuss their beloved Thief series, among other things. The TTLG forums were not just a place for discussion; they were an arena for absurdity, a carnival of capers, and a theater of the hilariously unexpected. Here are some of the most memorable antics of its quirky denizens.
The Great Taffer War
It all started innocuously enough, as most epic tales do, with a simple thread titled, "Who's the Best Thief?" The debate quickly escalated into the Great Taffer War of 2010. Lines were drawn, alliances were formed, and every member was ready to defend their favorite character with the fervor of a medieval knight defending their liege.
Bob_the_Lurker, known for his stealthy wit, declared, "Garrett is the only true master thief!" Meanwhile, Sneaky_Susan retorted, "Garrett couldn't sneak past a deaf guard with a hangover!" The forum erupted in a flurry of sarcastic memes, overly dramatic fan fiction, and, inexplicably, haikus about stealing cheese.
The Mysterious Disappearance of Lord Bafford
One fateful day, a user named ShadowStalker posted a thread entitled, "Lord Bafford's Manor - IRL?" They claimed to have found a real-life mansion that looked suspiciously like the infamous in-game location. The thread quickly filled with wild theories, photoshopped images, and elaborate plans to infiltrate the mansion.
The pinnacle of this madness was when member Phantom_Poet organized a "recon mission." Several members, clad in black hoodies and equipped with homemade lockpicks, actually ventured to the location. Needless to say, they didn't get far before being questioned by local authorities. The aftermath was a mixture of shame and pride, and the thread was immortalized in the forum's Hall of Fame.
The Epic Cooking Thread
Gaming forums and culinary expertise don't usually go hand in hand, but TTLG broke the mold. It began with a simple request for "Garrett's favorite stew recipe." What followed was a culinary explosion. Members shared medieval-inspired recipes, from "Bafford's Beef Bourguignon" to "Hammerite Honey Glazed Ham."
The highlight was the "TTLG Cook-Off," a virtual cooking competition judged by the forum's most discerning palates. Pictures of dubious dishes filled the thread. Bob_the_Lurker's "Stealthy Soup" was declared the winner, although many suspected foul play involving a strategically placed fog machine.
The Cat Burglar Chronicles
Cats are inherently funny, but TTLG members took it to another level. User Felinious_Fiend began posting photos of their cat dressed as a tiny thief, complete with a miniature blackjack and hood. The thread, "Cat Burglar Chronicles," became an overnight sensation.
Not to be outdone, other members started sharing their own cats in various Thief-inspired costumes. Soon, the forum was flooded with images of cats scaling curtains like walls, sneaking up on unsuspecting houseplants, and generally engaging in feline tomfoolery. It was a glorious, whisker-filled spectacle that even Garrett himself would have approved of.
The Unforgettable Glitch Contest
No forum is complete without a discussion of game glitches, and TTLG turned it into an art form. The "Thief Glitch Contest" invited members to share the most bizarre, hilarious, and downright ridiculous glitches they encountered.
Highlights included guards moonwalking across rooftops, Garrett becoming temporarily transparent, and AI characters getting stuck in perpetual dance loops. The winner, however, was undoubtedly ShadowStalker, who managed to get Garrett stuck inside a giant loaf of bread. The screenshot of a thief with only his head and arms poking out of a bread loaf became legendary."
demagogue on 20/6/2024 at 06:24
I've been watching a lot of (
https://lumalabs.ai/dream-machine/) Luma videos recently on the (
https://www.facebook.com/groups/1208731756401972) Cursed AI page. It's only been I think about 3 days since it dropped, and it already feels like it's established a whole genre unto itself.
Anyway, what I really wanted to post was this Reverse Turing Test video.
I don't know what's going on anymore.
It's a clever concept that was well executed though.
[video=youtube;MxTWLm9vT_o]https://www.youtube.com/watch?v=MxTWLm9vT_o[/video]
heywood on 20/6/2024 at 16:48
That's a fun premise. At first I was going to complain that he made himself too obvious, but what emerged was a subtle tone of snobbery from the AI responses that I didn't expect.
Sulphur on 20/6/2024 at 17:22
The site's also extremely slow quite often now.
I haven't seen the Reverse Turing Test video yet, but I want to nod in the direction of the cursed AI page. This one specifically caught my eye because Scorn has
nothing on it: (
https://www.facebook.com/groups/cursedaiwtf/posts/1566624437279367/) The Tunnels. Incredible, disgusting, horrifying.
