Narz
keeping it real
Translation, Muslim fundies will murder my makers if I joke around about Mo
The AI Thread
- Thread starter Truthy
- Start date
Josu
Emperor
Yesterday I had some free time so I signed up in chatGPT.
I asked to recommend a historical novel set on the basque country.
The first two recommendations were a books that actually don't not exist.
I asked for the ISBN, and gave me a number with wrong format
Third recommendation was a book set in Afghanistan
Next recomendation was Basque history of the world by Mark Kurlansky, which is a good book, but not novel.
Next one was Obaba, by Bernardo Atxaga, probably biggest best seller novel in basque languaje, but not historical.
Asked for another and recommended The accordeonist's son, again by Bernardo Atxaga, set in 1950 to present.
As I have allready read these three books, I asked for a historical novel not by Bernardo Atxaga, and recommended Basque history of the world by Mark Kurlansky, as said before, not a novel.
I told the AI that it was not a novel and recommended Obaba, by Bernardo Atxaga.
Yes, the AI entered in a loop with this three books, and I continued with the conversation to see if it was able to exit for there, I gave up at the third iteration.
I asked to recommend a historical novel set on the basque country.
The first two recommendations were a books that actually don't not exist.
I asked for the ISBN, and gave me a number with wrong format
Third recommendation was a book set in Afghanistan
Next recomendation was Basque history of the world by Mark Kurlansky, which is a good book, but not novel.
Next one was Obaba, by Bernardo Atxaga, probably biggest best seller novel in basque languaje, but not historical.
Asked for another and recommended The accordeonist's son, again by Bernardo Atxaga, set in 1950 to present.
As I have allready read these three books, I asked for a historical novel not by Bernardo Atxaga, and recommended Basque history of the world by Mark Kurlansky, as said before, not a novel.
I told the AI that it was not a novel and recommended Obaba, by Bernardo Atxaga.
Yes, the AI entered in a loop with this three books, and I continued with the conversation to see if it was able to exit for there, I gave up at the third iteration.
Hygro
soundcloud.com/hygro/
ahahaha rekt
courtesy bing
courtesy bing
1000 science nerds have signed a letter asking for a 6 month pause in making AI stronger than Chat-gpt4
futureoflife.org
Once Skynet learns how to write code, it is off to the races I'd say.
Three laws of robotics should slow it down by a few seconds.
Actually, if they could articulate what exactly they think is dangerous, that could be helpful.
The open letter reads like it was written by AI.
Pause Giant AI Experiments: An Open Letter - Future of Life Institute
We call on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4.
futureoflife.org
Three laws of robotics should slow it down by a few seconds.
Actually, if they could articulate what exactly they think is dangerous, that could be helpful.
The open letter reads like it was written by AI.
Last edited:
A few of the people are famous.
It might be a pitch to get government involved more, to hire more AI people and spend more money regulating stuff.
The cynical view is always around.
Then again, maybe they really are concerned for the future of all humans.
*Edit*
I guess the biggest concern is that evil people will all get a genius-level personal assistant with no morals some day in the future?
Still confused about what they want government to do exactly.
It might be a pitch to get government involved more, to hire more AI people and spend more money regulating stuff.
The cynical view is always around.
Then again, maybe they really are concerned for the future of all humans.
*Edit*
I guess the biggest concern is that evil people will all get a genius-level personal assistant with no morals some day in the future?
Still confused about what they want government to do exactly.
Last edited:
Hygro
soundcloud.com/hygro/
I 100% disagree with all of them. OpenAI should do what it wants. It had demonstrated itself to be trustworthy and shouldn’t give its competitors 6 months to dethrone it.
That said, in the next 6 months is it even working on the successor to gpt-4? Or are they spending the next while just making gpt-4 better since that’s their flagship product and they can’t even keep up with demand?
That said, in the next 6 months is it even working on the successor to gpt-4? Or are they spending the next while just making gpt-4 better since that’s their flagship product and they can’t even keep up with demand?
Hygro
soundcloud.com/hygro/
I started talking to it about all my theories that no one seems to “get” unless they already came up with something similar on their own. It handles it really quite well.
I’m definitely going to be obsolete soon.
I’m definitely going to be obsolete soon.
r16
not deity
- Joined
- Nov 10, 2008
- Messages
- 12,142
finance is fiction , let miniscule amounts of real money flow out of the system and do nothing and it will be multiplied over the years . Enough to get America and China to actually think about war . Some Swiss bank just last week or so apparently stole 17 billion dollars from its customers . The Kerkük oil arbitration in favour of Iraq was agreed in mid 2022 and is announced only today , because that 1,4 billion compensation reaching to 3,5 billion dollars is like real money getting into the system . As if ransom . If not , then the elections will be a surprise . Greed of mankind would tax any computer to emulate ...
Thorgalaeg
Deity
Translation: Elon and apple are way behind and want Open AI and Microsoft to stop six months so they have time to develop his own language models.1000 science nerds have signed a letter asking for a 6 month pause in making AI stronger than Chat-gpt4
Pause Giant AI Experiments: An Open Letter - Future of Life Institute
We call on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4.
Once Skynet learns how to write code, it is off to the races I'd say.
Three laws of robotics should slow it down by a few seconds.
Actually, if they could articulate what exactly they think is dangerous, that could be helpful.
The open letter reads like it was written by AI.
Only possible philosophical and juridically fundamented answer to the letter: cry me a damn bloody river.
Last edited:
Thorgalaeg
Deity
First "atempts" to get closer to some kind of AGI (Artificial General Inteligence or strong AI). It is not any real strong AI of course but an automation of GPT (we are probably decades away of real strong AI, or at least some months 
), however it may show the path to follow. It is amazing to see the AI reasoning and taking decisions to reach the proposed goals. If some github open source projects are already at this point who knows what OpenAI or Google have already developed in secret.
), however it may show the path to follow. It is amazing to see the AI reasoning and taking decisions to reach the proposed goals. If some github open source projects are already at this point who knows what OpenAI or Google have already developed in secret.
Last edited:
Decades, or an eternity according to some prominent scientists. At least if we are talking about digital machines 
But imo real ai isn't more useful than the type of ai we are seeing with the current models. Although their factual errors (in math/science) have to go down still.
But imo real ai isn't more useful than the type of ai we are seeing with the current models. Although their factual errors (in math/science) have to go down still.
Hygro
soundcloud.com/hygro/
I guess we considered terminator a sentient-enough AI, and that’s basically the level that exists if we made these models behave that way, packaged them into one piece of hardware.
Hygro
soundcloud.com/hygro/
I remember T1 much better than T2. What in T2 proves he has more sentience than a well tuned, rogue-goaled LLM attached to a vision AI and a movement AI?
Hygro
soundcloud.com/hygro/
Haha Samson and I on the two pronged attack, coming at it from opposite sides
