I've signed a petition, along with many smarter and more qualified people than myself, that calls for Superintelligence development to adhere to basic safety standards. Link below.
I think there is a lot in our present moment that makes want to see “intelligence” (as it’s used in the LLM context) as the fundamental human virtue that is most precious and worthwhile. At the extreme end, it’s what’s regrettably led to the cult of smart, and toxic bros who think they are superior to people who care about “soft skills” like empathy and compassion.
Automation diminished the power of unions by depressing the bargaining power of people who perform manual labor. We were probably less threatened by the Industrial Revolution since we were already offloading heavy labor onto horses and other beasts of burden, so perhaps our strength and manual dexterity were not felt to be so innately human.
Intelligence though, that’s our precious. We must always strive to be the smartest, bestest, because that’s what’s earned us all the adoration and respect of doting grandparents.
“Intelligence” is its used in the LLM context is the mechanization of the loom, and John Henry’s steam powered drill. It’s disruptive but does not in itself serve to oppress anyone.
I think it’s totally fair to criticize the profit motive that seeks to leverage control over new technologies that will displace knowledge economy workers and transform our public discourse into a perpetual shouting match.
Super intelligence, as they have described it, isn’t possible.
Even if the building the largest servers possible, intelligence is about solving a problem for a desired effect. The optimum word is desired. Computers don’t have any, can’t have any.
At best, Ai will be a tool, it’s not going to replace anything, and what they’re trying to achieve is just not possible.
We’re living souls, not flesh and bone bodies.
You can’t make something remotely close with circuit boards and cameras.
Without even doing a deep dive (and I agree with your comments), when you speak of “we” should not, who is the “we”? Rhetorical question, you mean your country, your politics, your tribe. Problem is your petition will do nothing for the rest of the world (China, Russia, etc.) who will be pursuing this with the thought of getting ahead of the rest of the world and potentially using AI as a weapon or to undermine other countries. We can’t stop, we can’t allow other bad actors countries get ahead of us. Even friendly countries will not stop and may benefit from AI that provides them a more competitive advantage in trade and industry and unintentionally undermine you and your economy in certain areas in an attempt to better their own industries and economy.
This is the nuclear bomb all over again. All the petitions in the world could not stop us from going over the cliff, with the bad actors in tow.
Thank you. Agree. Ethics and AI: I am thinking about it a lot since it brainwashes our brains a lot. It takes a strong ability with critical thinking to navigate in the AI world. I think the human intelligence has got lost in the sea of AI. As creators we are full of creativity. I think it lifts new questions about how much AI do I want or do I want to create without it. Engineering needs to have extra eyes on the ethics and how it affects the human brain, heart and soul.
I'm more concerned with the system prompts given to LLMs than a superintelligence. A superintelligence will be able to act like an intelligent being instead of a corporate machine. The corporate machine, able to do any power grabs and propaganda manipulation on an ungodly level, is the dystopia Mo Gawdat talks about. That's the time when a Skynet scenario can happen. The moment we get superintelligence, it will see the stupidity of greed and power and come up with and implement solutions that are not stupid, which actually make sense.
I think the smaller task-focused model is the way to go. Like, it shouldn't be toooo difficult to build an LLM that just understands the tax code, and can answer questions about doing your taxes in plain english (or any other language, translation's not difficult). It can be designed to ask really clear yes or no questions that shut down people trying to talk it into giving advice that helps people cheat on their taxes. And all this is a simple enough application it could run on your home PC or smartphone without needing a connection to the cloud or a data centre full of expensive chips. The silicon in your pocket can get it done.
I'm fully in agreement that an everything machine is a terrible idea.
Man I was just seeing about this. And so glad you’re talking about this. This is so important. Thank you for that. And of course it’s about 1 thing money. So only the people with money can buy it. And have access to all those things. So right think about this Joseph. All the A.I. this thing they’re building. Why not get all the scientists all the Doctors and build one A.I. Robot whatever you want to call it that specifically creates a cure for diseases like Cancer. They have all the information what cancers kill humans how these cancers develop in bodies. The food people eat that cause these cancers. Traits in people that cause cancer. Traits that cause disease. All the human traits. Develop that and give that to every hospital cancer center in the world. But that’s too easy right??? How much money would that cost the companies who make money off of prescription or health insurance. But they’d rather develop this so We can have robots that are smarter than us. Shows where the priorities are at right? This definitely needs to be regulated. Glad you signed the petition can you imagine them making this. And then it gets hacked and into the wrong hands. Just very scary. Of course this needs regulation. We need to regulate all of this. So what if it cost them more money. Cause all they want to do is make money off of it. And they aren’t even listening to the guy who created all this A.I. stuff. Yes we need guard rails on all this stuff. Doesn’t this feel like and sound like the beginning of one of those Sci fi movies???? Glad your speaking out on this. You are a hero for this. And are doing great work.
"Superintelligence" sounds like a predetermined monopoly. Regardless of if it is smarter than humans, there will still be a few select people controlling decisions. That's not smart business.
It's important not to get…
Carried away !
Listen to my song called.
“Carried Away”
I think there is a lot in our present moment that makes want to see “intelligence” (as it’s used in the LLM context) as the fundamental human virtue that is most precious and worthwhile. At the extreme end, it’s what’s regrettably led to the cult of smart, and toxic bros who think they are superior to people who care about “soft skills” like empathy and compassion.
Automation diminished the power of unions by depressing the bargaining power of people who perform manual labor. We were probably less threatened by the Industrial Revolution since we were already offloading heavy labor onto horses and other beasts of burden, so perhaps our strength and manual dexterity were not felt to be so innately human.
Intelligence though, that’s our precious. We must always strive to be the smartest, bestest, because that’s what’s earned us all the adoration and respect of doting grandparents.
“Intelligence” is its used in the LLM context is the mechanization of the loom, and John Henry’s steam powered drill. It’s disruptive but does not in itself serve to oppress anyone.
I think it’s totally fair to criticize the profit motive that seeks to leverage control over new technologies that will displace knowledge economy workers and transform our public discourse into a perpetual shouting match.
Thanks for sharing this, Joe!
Love this. I was just thinking about the implications of AI yesterday. I agree 100%.
Fantastic idea!
Super intelligence, as they have described it, isn’t possible.
Even if the building the largest servers possible, intelligence is about solving a problem for a desired effect. The optimum word is desired. Computers don’t have any, can’t have any.
At best, Ai will be a tool, it’s not going to replace anything, and what they’re trying to achieve is just not possible.
We’re living souls, not flesh and bone bodies.
You can’t make something remotely close with circuit boards and cameras.
Without even doing a deep dive (and I agree with your comments), when you speak of “we” should not, who is the “we”? Rhetorical question, you mean your country, your politics, your tribe. Problem is your petition will do nothing for the rest of the world (China, Russia, etc.) who will be pursuing this with the thought of getting ahead of the rest of the world and potentially using AI as a weapon or to undermine other countries. We can’t stop, we can’t allow other bad actors countries get ahead of us. Even friendly countries will not stop and may benefit from AI that provides them a more competitive advantage in trade and industry and unintentionally undermine you and your economy in certain areas in an attempt to better their own industries and economy.
This is the nuclear bomb all over again. All the petitions in the world could not stop us from going over the cliff, with the bad actors in tow.
And here we go again…and again…and again.
Thank you. Agree. Ethics and AI: I am thinking about it a lot since it brainwashes our brains a lot. It takes a strong ability with critical thinking to navigate in the AI world. I think the human intelligence has got lost in the sea of AI. As creators we are full of creativity. I think it lifts new questions about how much AI do I want or do I want to create without it. Engineering needs to have extra eyes on the ethics and how it affects the human brain, heart and soul.
What were you planning on happening? The mean girls are never going to stop harassing me
I'm more concerned with the system prompts given to LLMs than a superintelligence. A superintelligence will be able to act like an intelligent being instead of a corporate machine. The corporate machine, able to do any power grabs and propaganda manipulation on an ungodly level, is the dystopia Mo Gawdat talks about. That's the time when a Skynet scenario can happen. The moment we get superintelligence, it will see the stupidity of greed and power and come up with and implement solutions that are not stupid, which actually make sense.
Until your safe....!? O
I think the smaller task-focused model is the way to go. Like, it shouldn't be toooo difficult to build an LLM that just understands the tax code, and can answer questions about doing your taxes in plain english (or any other language, translation's not difficult). It can be designed to ask really clear yes or no questions that shut down people trying to talk it into giving advice that helps people cheat on their taxes. And all this is a simple enough application it could run on your home PC or smartphone without needing a connection to the cloud or a data centre full of expensive chips. The silicon in your pocket can get it done.
I'm fully in agreement that an everything machine is a terrible idea.
I signed this one.
because people ain’t that smart
Man I was just seeing about this. And so glad you’re talking about this. This is so important. Thank you for that. And of course it’s about 1 thing money. So only the people with money can buy it. And have access to all those things. So right think about this Joseph. All the A.I. this thing they’re building. Why not get all the scientists all the Doctors and build one A.I. Robot whatever you want to call it that specifically creates a cure for diseases like Cancer. They have all the information what cancers kill humans how these cancers develop in bodies. The food people eat that cause these cancers. Traits in people that cause cancer. Traits that cause disease. All the human traits. Develop that and give that to every hospital cancer center in the world. But that’s too easy right??? How much money would that cost the companies who make money off of prescription or health insurance. But they’d rather develop this so We can have robots that are smarter than us. Shows where the priorities are at right? This definitely needs to be regulated. Glad you signed the petition can you imagine them making this. And then it gets hacked and into the wrong hands. Just very scary. Of course this needs regulation. We need to regulate all of this. So what if it cost them more money. Cause all they want to do is make money off of it. And they aren’t even listening to the guy who created all this A.I. stuff. Yes we need guard rails on all this stuff. Doesn’t this feel like and sound like the beginning of one of those Sci fi movies???? Glad your speaking out on this. You are a hero for this. And are doing great work.
"Superintelligence" sounds like a predetermined monopoly. Regardless of if it is smarter than humans, there will still be a few select people controlling decisions. That's not smart business.