I think the concept of AGI is a little flawed itself in comparing it to a human. If we are talking about strictly reasoning and even emotional intelligence, it might be comparable, however when you factor in the crystalized knowledge it has access to, it immediately becomes smarter than a person just by existing with equal reasoning. Imagine how many things you could figure out and do if you had immediate access to all the data an AGI would have access to, without even increasing your ability to reason. It would be like having the knowledge of a PHD in nearly all subjects. AGI would also have to have be able to interact with, or at least experience its environment. Then factor in if we build more than one, which seems quite likely. Narrowing down what it can do is almost comical at that point; it could transform every field of research, development, manufacturing etc. Then the problem becomes how we treat it, because at that point, sentience must be considered. We spend so much time teaching ourselves that crystalized knowledge, where an AGI doesn't have that limitation. Overall, it just seems nonsensical to define AGI in a way that compares it to a human. The standard seems flawed.
I'm just curious that where is IBM in the AGI development? I think some companies you know who already passed the stage where IBM may can't see that yet.
Exacty. For example, Open AI has secret models that seem to be breakthroughs in extended red team mode. AGI is not far off, according to my understanding and knowledge of current models as well as models to come.
Main advantage of the AGI is to act as life coach and mentor to each person. The main disadvantage is that once you hand your life over to the machine you don't get it back.
number 1 wont really help much since people want to talk with people, unless AGI can be held responsible which cannot. and any customer services that can be solved by the AGI should be able to be solved by classic GUI based App.
I absolutely agree about the eight AGI-benefits: Machines will better simulate empathy than old-school call-center agent, write code only machines will be able to maintain, cover-up car accidents, find reasons to raise your health insurance, substitute these awful 'hugging' teachers, kick the blue-collar asses, widen the social gap via SE, and lower the already low income of scientists.
The autonomous transportation topic. I think it is cool technology and should be a life saving innovation as an _assistant_ to humans for defensive driving or seeing something that we miss like a child running after a ball in the road if driver is distracted. It would improve life, safety, peace of mind and insurance rates. Just all and all a very good technology advancement. The elon complete human driverless auto mobile with no one to be there to control the car, in case of some sort of emergency I find to be a bad idea for a list of reasons and frankly a “stupid idea”. It attempts to solve a problem that doesn’t exist. There will be no robotaxi.
right now everyone uses 'scans' in the stock market to find the runners or the under valued-- it would be nice if AI did some/all of this work-- but I dont have the time or skills to learn AI. so I have to do the work by hand.
I feel very sad because even I as a good software engineer, I don't find a job in France. I am just stuck in this country where skills doesn't matter... Will I be able to work as an engineer? Or have I done all the years of effort for nothing? But I love the topic.
Stop using "use case" this way. You know this isn't what it means. "Use case" is a term from UML referring to which parts of your program should have access to other parts., not a synonym for "way to use".
![The moment we stopped understanding AI [AlexNet]](http://i.ytimg.com/vi/UZDiGooFs54/mqdefault.jpg)
I think the concept of AGI is a little flawed itself in comparing it to a human. If we are talking about strictly reasoning and even emotional intelligence, it might be comparable, however when you factor in the crystalized knowledge it has access to, it immediately becomes smarter than a person just by existing with equal reasoning. Imagine how many things you could figure out and do if you had immediate access to all the data an AGI would have access to, without even increasing your ability to reason. It would be like having the knowledge of a PHD in nearly all subjects. AGI would also have to have be able to interact with, or at least experience its environment. Then factor in if we build more than one, which seems quite likely. Narrowing down what it can do is almost comical at that point; it could transform every field of research, development, manufacturing etc. Then the problem becomes how we treat it, because at that point, sentience must be considered. We spend so much time teaching ourselves that crystalized knowledge, where an AGI doesn't have that limitation. Overall, it just seems nonsensical to define AGI in a way that compares it to a human. The standard seems flawed.
Bro is taking about gpt5 not agi
Looking forward to the day we discover AGI.
I'm just curious that where is IBM in the AGI development? I think some companies you know who already passed the stage where IBM may can't see that yet.
Exacty. For example, Open AI has secret models that seem to be breakthroughs in extended red team mode. AGI is not far off, according to my understanding and knowledge of current models as well as models to come.
Is any thing under the sun, AGI will not able to do? The premise of use case for AGI is flawed as usecse for humans.
Main advantage of the AGI is to act as life coach and mentor to each person. The main disadvantage is that once you hand your life over to the machine you don't get it back.
number 1 wont really help much since people want to talk with people, unless AGI can be held responsible which cannot.
and any customer services that can be solved by the AGI should be able to be solved by classic GUI based App.
Can you give a video on work that AGI or SI can't do without human intraction and the future of human work
how about the military?
I absolutely agree about the eight AGI-benefits: Machines will better simulate empathy than old-school call-center agent, write code only machines will be able to maintain, cover-up car accidents, find reasons to raise your health insurance, substitute these awful 'hugging' teachers, kick the blue-collar asses, widen the social gap via SE, and lower the already low income of scientists.
The autonomous transportation topic. I think it is cool technology and should be a life saving innovation as an _assistant_ to humans for defensive driving or seeing something that we miss like a child running after a ball in the road if driver is distracted. It would improve life, safety, peace of mind and insurance rates. Just all and all a very good technology advancement.
The elon complete human driverless auto mobile with no one to be there to control the car, in case of some sort of emergency I find to be a bad idea for a list of reasons and frankly a “stupid idea”. It attempts to solve a problem that doesn’t exist. There will be no robotaxi.
right now everyone uses 'scans' in the stock market to find the runners or the under valued-- it would be nice if AI did some/all of this work-- but I dont have the time or skills to learn AI. so I have to do the work by hand.
I feel very sad because even I as a good software engineer, I don't find a job in France. I am just stuck in this country where skills doesn't matter...
Will I be able to work as an engineer? Or have I done all the years of effort for nothing?
But I love the topic.
Tell me more, what are you skills? How good are you with AI? Do you have experience? If not, what are you good at?
If tech companies want to convince people to use ai, they should use ai to code most of their programs first.
1. pass the butter
What is my purpose?
So its not too far away 😊
❤
Not according to Google ex Ceo. It will be out in 2025
Il est temps de passer à la propriété intellectuelle et éthique mettre fin à certaines pratiques malsaine.
Stop using "use case" this way. You know this isn't what it means. "Use case" is a term from UML referring to which parts of your program should have access to other parts., not a synonym for "way to use".
Not anymore grandad