4 ms·

> How do you define conscious? 'Feeling' just like you do? I think many modern AI's could be said to experience a different, simple form of consciousness as the

> How do you define conscious? 'Feeling' just like you do? I think many modern AI's could be said to experience a different, simple form of consciousness as they process inputs. What an odd comment. So because we performed some clever math to make some basic neural nets, our AI is now experiencing a form of consciousness? This is just better algorithms that are more capable of fooling you, still running on the same microprocessor as before. Consciousness is described as a state of constant awareness, not of performing the act of perception itself. > I think it's safe to say that a single neuron is not as complex and mysterious as you may think. Not sure if trolling or not? You should look up protein folding and protein machines at some point, basically there is a LOT going on at the nano scale that we still don't fully understand. Additionally tell any neuroscientist that the brains neural networks are the same as our computer version of NNs and you'll be laughed out the room. And perception is a function of the brain yes, how is that an opinion? Your arguments really make no sense.

lend000 10y ago
> What an odd comment. So because we performed some clever math to make some basic neural nets, our AI is now experiencing a form of consciousness? Yes, you keep re-hashing this point which makes it clear that you still (and probably always will) believe that consciousness is something 'more' -- something mystical and unexplainable -- because it's so difficult and non-intuitive to wrap your head around (as it is for most people). > Not sure if trolling or not? You should look up protein folding and protein machines at some point, basically there is a LOT going on at the nano scale that we still don't fully understand. Additionally tell any neuroscientist that the brains neural networks are the same as our computer version of NNs and you'll be laughed out the room. Who said our models are the same? You're missing the point. Re-read the paragraph you're replying to, and look what we've accomplished with just dozens of simplified neurons. The point is that the details beneath our high level understanding of neurons (basically just spiking and activation) are likely not important to consciousness. Just like a high school physics student can usefully understand the mechanics of rubber balls without fully grasping the chemistry at the atomic level. > And perception is a function of the brain yes, how is that an opinion? Your arguments really make no sense. Here's the context, to remind you: > Amazing that every sentence you wrote is objectively wrong. >> 7. This internal stream of perceptions and judgements is consciousness. > No, this is just a function of part of the brain, which we can be aware of through consciousness. Just want to feel like you won an argument today? No need to turn this into a troll battle -- I'm actually trying to explain my viewpoint to you, if you're interested.
- hacker_9 10y ago
  You are arguing that by performing the algorithms for perception, one is conscious. I am arguing that algorithms alone aren't enough, and that yes conscious is still some unexplained physics that takes the results of those perception algorithms in our brains and generates our awareness and experience from them. I am not trying to win any argument, but it does annoy me when people argue with incomplete information and try to trivialize consciousness, because of what you know about some basic neural nets in computer science. Look at this video released last month, where scientists show a model for just 3 neurons [1], which they think is super important to consciousness. Only 3 neurons out of billions. There is just so much we still don't know. [1] https://www.youtube.com/watch?v=5s5I-rUyDTA https://www.youtube.com/watch?v=5s5I-rUyDTA
  - lend000 10y ago
    > You are arguing that by performing the algorithms for perception, one is conscious. Pretty close -- my argument is that for an entity to be intelligent, self-aware, and perceptive, it must also believe itself to be conscious, in the sense that it believes it has "experiences." Furthermore, and most importantly, if an entity believes it is conscious, it is conscious. > I am arguing that algorithms alone aren't enough, and that yes conscious is still some unexplained physics that takes the results of those perception algorithms in our brains and generates our awareness and experience from them. There's so much undiscovered truth about physics that it's always possible there are important, still completely unknown properties of physics that play specific roles in consciousness. However, that's mostly speculation (like the article) and there's no real evidence for it, whereas it's becoming increasingly intuitive to some AI researchers and myself that such a 'Holy Grail of Physics' yet-to-be-discovered property does not seem to be necessary. Many AI experts presume that by scaling and adding complexity to models that are pretty similar to today's AI models, we can achieve general intelligence. I think that this logically requires the intelligent agent to believe itself to be conscious, and therefore be just as conscious as any human, albeit in a different way. The combination of intelligence and self-awareness are logically incompatible with a lack of consciousness.

lend000 10y ago

> What an odd comment. So because we performed some clever math to make some basic neural nets, our AI is now experiencing a form of consciousness? Yes, you keep re-hashing this point which makes it clear that you still (and probably always will) believe that consciousness is something 'more' -- something mystical and unexplainable -- because it's so difficult and non-intuitive to wrap your head around (as it is for most people). > Not sure if trolling or not? You should look up protein folding and protein machines at some point, basically there is a LOT going on at the nano scale that we still don't fully understand. Additionally tell any neuroscientist that the brains neural networks are the same as our computer version of NNs and you'll be laughed out the room. Who said our models are the same? You're missing the point. Re-read the paragraph you're replying to, and look what we've accomplished with just dozens of simplified neurons. The point is that the details beneath our high level understanding of neurons (basically just spiking and activation) are likely not important to consciousness. Just like a high school physics student can usefully understand the mechanics of rubber balls without fully grasping the chemistry at the atomic level. > And perception is a function of the brain yes, how is that an opinion? Your arguments really make no sense. Here's the context, to remind you: > Amazing that every sentence you wrote is objectively wrong. >> 7. This internal stream of perceptions and judgements is consciousness. > No, this is just a function of part of the brain, which we can be aware of through consciousness. Just want to feel like you won an argument today? No need to turn this into a troll battle -- I'm actually trying to explain my viewpoint to you, if you're interested.

hacker_9 10y ago
You are arguing that by performing the algorithms for perception, one is conscious. I am arguing that algorithms alone aren't enough, and that yes conscious is still some unexplained physics that takes the results of those perception algorithms in our brains and generates our awareness and experience from them. I am not trying to win any argument, but it does annoy me when people argue with incomplete information and try to trivialize consciousness, because of what you know about some basic neural nets in computer science. Look at this video released last month, where scientists show a model for just 3 neurons [1], which they think is super important to consciousness. Only 3 neurons out of billions. There is just so much we still don't know. [1] https://www.youtube.com/watch?v=5s5I-rUyDTA https://www.youtube.com/watch?v=5s5I-rUyDTA
- lend000 10y ago
  > You are arguing that by performing the algorithms for perception, one is conscious. Pretty close -- my argument is that for an entity to be intelligent, self-aware, and perceptive, it must also believe itself to be conscious, in the sense that it believes it has "experiences." Furthermore, and most importantly, if an entity believes it is conscious, it is conscious. > I am arguing that algorithms alone aren't enough, and that yes conscious is still some unexplained physics that takes the results of those perception algorithms in our brains and generates our awareness and experience from them. There's so much undiscovered truth about physics that it's always possible there are important, still completely unknown properties of physics that play specific roles in consciousness. However, that's mostly speculation (like the article) and there's no real evidence for it, whereas it's becoming increasingly intuitive to some AI researchers and myself that such a 'Holy Grail of Physics' yet-to-be-discovered property does not seem to be necessary. Many AI experts presume that by scaling and adding complexity to models that are pretty similar to today's AI models, we can achieve general intelligence. I think that this logically requires the intelligent agent to believe itself to be conscious, and therefore be just as conscious as any human, albeit in a different way. The combination of intelligence and self-awareness are logically incompatible with a lack of consciousness.