5 ms·

In addition to that subtle memory issue, it has no reference at all to the space/time world we people model mentally to think with. So, basically, there is no

In addition to that subtle memory issue, it has no reference at all to the space/time world we people model mentally to think with. So, basically, there is no I in the GPT-3 AI, just A.

PaulHoule 5y ago
One can point to many necessary structural features that it is missing. Consider Ashby's law of requisite variety: https://www.edge.org/response-detail/27150 https://www.edge.org/response-detail/27150 Many GPT-3 cultists are educated in computer science so they should know better. GPT-3's "one pass" processing means that a fixed amount of resources are always used. Thus it can't sort a list of items unless the fixed time it uses is humongous. You might boil the oceans that way but you won't attain AGI. There are numerous arguments along the line of Turing's halting problem that restrict what that kind of thing can do. As it uses a finite amount of time it can't do anything that could require an unbounded time to complete or that could potentially not terminate. GPT-3 has no model for dealing with ambiguity or uncertainty. (Other than shooting in the dark.) Practically this requires some ability to backtrack either automatically or as a result of user feedback. The current obscurantism is that you need to have 20 PhD students work for 2 years to write a paper that makes the model "explainable" in some narrow domain. With this insight you can spend another $30 million training a new model that might get the answer right. A practical system needs to be told that "you did it wrong" and why and then be able to correct itself on the next pass if possible, otherwise in a few passes. Of course a system like that would be a real piece of engineering that people would become familiar with, not a outlet for their religious feelings that is kept on a pedestal.
- gibsonf1 5y ago
  The big issue is that it literally knows nothing - there is no reference to a model of the real world such as humans use when thinking about the real world. It is a very advanced pattern matching parrot, and in using words like a parrot, knows nothing about what those words mean.
  - PaulHoule 5y ago
    Exactly, with "language in language out" it can pass as a neurotypical (passing as a neurotypical doesn't mean you get the right answer, it means if you get a wrong answer it is a neurotypical-passing wrong answer.) Actual "understanding" means mapping language to something such as an action (I tell you to get me the plush bear and you get me the plush bear,) precise computer code, etc.
  - macrolocal 5y ago
    I'm inclined to agree, but positing that "the meaning of a word is its use in a language" is a perfectly respectable philosophical position. In this sense, GPT3 empirically bolsters Wittgenstein.
    - PaulHoule 5y ago
      Given that Chomksy's paradigm (and all the others) have failed so completely at automated language understanding I'm much more inclined to believe that the "language instinct" is a peripheral for an animal and if you want language understanding you need the animal or a simulation of the animal. Thus I reject Wittgenstein.
      - macrolocal 5y ago
        That's why it's so amazing that language models have gotten so far with just proximity statistics! Btw, this paper https://arxiv.org/abs/2106.07890 https://arxiv.org/abs/2106.07890 outlines a promising approach to mapping the semantics encoded in these models. Understanding their limits could make your rejection of Wittgenstein all the more precise.
- narrator 5y ago
  >There are numerous arguments along the line of Turing's halting problem that restrict what that kind of thing can do. As it uses a finite amount of time it can't do anything that could require an unbounded time to complete or that could potentially not terminate. I have used a similar argument to show that the simulation hypothesis is wrong. If any algorithm used to simulate the world takes longer than o(N) time, then the most efficient possible computer for that is the universe which computes everything in O(n) time where n is time. In other words, you never get "lag" in reality no matter how complex the scene you're looking at is. Worse than that, some simulation algorithms are exponential time complexity!
  - chowells 5y ago
    That doesn't prove or disprove anything. What we experience as time would be part of the simulation, were such a hypothesis true. As such, the way in which we experience it is fully independent from whatever costs it might have to compute.
    - narrator 5y ago
      So you're saying that an exponential time complexity algorithm with N of every atom in the universe will complete before the heat death of the other universe that the simulation is taking place in? Sorry, not plausible.
      - Bjartr 5y ago
        Why does the containing universe necessarily have comparable physical laws?
      - Jensson 5y ago
        Our laws of physics are space partitioned so the algorithm for simulating it isn't exponential. If the containing universe has like 21 dimensions and otherwise have similar tech computers as we do today then you should be able to simulate it on a datacenter just fine as computation ability grows exponentially with number of dimensions. 3 dimensions you have 2 dimensions of computation surface, 21 dimensions and you have 20 dimensions of computation surface, so our current computation to the power of 10. GPT3 used more than a petaflop real time compute during training, so 10 to the power of 15. Using the same hardware in our fictive universe would give us 10 to the power of 150 flops. We estimate atoms in the universe to be about 10 to the power of 80, with this computer we would have 10 to the power of 70 flops of compute per atom, that should be enough even if entanglement gets a bit messy. We have around that much memory per atom as well, so can compute a lot of small boxes and sum over all of it etc, to emulate particle waves. We wouldn't be able to detect computational anomalies on that small scale, so we can't say that there isn't such a computer emulating us.
- carrolldunham 5y ago
  GPT is a joke but Ashby's various "Laws" and "Theorems" are just theories. As in, think "Jung's theory of archetypes". Cyberneticists cribbed hard-sciences terminology for their psycho-speculation freely, knowing their humanities peers couldn't call them to account about technical meanings. Theorem and law just sound engineer-y

In addition to that subtle memory issue, it has no reference at all to the space/time world we people model mentally to think with. So, basically, there is no I in the GPT-3 AI, just A.

PaulHoule 5y ago

One can point to many necessary structural features that it is missing. Consider Ashby's law of requisite variety: https://www.edge.org/response-detail/27150 https://www.edge.org/response-detail/27150 Many GPT-3 cultists are educated in computer science so they should know better. GPT-3's "one pass" processing means that a fixed amount of resources are always used. Thus it can't sort a list of items unless the fixed time it uses is humongous. You might boil the oceans that way but you won't attain AGI. There are numerous arguments along the line of Turing's halting problem that restrict what that kind of thing can do. As it uses a finite amount of time it can't do anything that could require an unbounded time to complete or that could potentially not terminate. GPT-3 has no model for dealing with ambiguity or uncertainty. (Other than shooting in the dark.) Practically this requires some ability to backtrack either automatically or as a result of user feedback. The current obscurantism is that you need to have 20 PhD students work for 2 years to write a paper that makes the model "explainable" in some narrow domain. With this insight you can spend another $30 million training a new model that might get the answer right. A practical system needs to be told that "you did it wrong" and why and then be able to correct itself on the next pass if possible, otherwise in a few passes. Of course a system like that would be a real piece of engineering that people would become familiar with, not a outlet for their religious feelings that is kept on a pedestal.

gibsonf1 5y ago
The big issue is that it literally knows nothing - there is no reference to a model of the real world such as humans use when thinking about the real world. It is a very advanced pattern matching parrot, and in using words like a parrot, knows nothing about what those words mean.
- PaulHoule 5y ago
  Exactly, with "language in language out" it can pass as a neurotypical (passing as a neurotypical doesn't mean you get the right answer, it means if you get a wrong answer it is a neurotypical-passing wrong answer.) Actual "understanding" means mapping language to something such as an action (I tell you to get me the plush bear and you get me the plush bear,) precise computer code, etc.
- macrolocal 5y ago
  I'm inclined to agree, but positing that "the meaning of a word is its use in a language" is a perfectly respectable philosophical position. In this sense, GPT3 empirically bolsters Wittgenstein.
  - PaulHoule 5y ago
    Given that Chomksy's paradigm (and all the others) have failed so completely at automated language understanding I'm much more inclined to believe that the "language instinct" is a peripheral for an animal and if you want language understanding you need the animal or a simulation of the animal. Thus I reject Wittgenstein.
    - macrolocal 5y ago
      That's why it's so amazing that language models have gotten so far with just proximity statistics! Btw, this paper https://arxiv.org/abs/2106.07890 https://arxiv.org/abs/2106.07890 outlines a promising approach to mapping the semantics encoded in these models. Understanding their limits could make your rejection of Wittgenstein all the more precise.
narrator 5y ago
>There are numerous arguments along the line of Turing's halting problem that restrict what that kind of thing can do. As it uses a finite amount of time it can't do anything that could require an unbounded time to complete or that could potentially not terminate. I have used a similar argument to show that the simulation hypothesis is wrong. If any algorithm used to simulate the world takes longer than o(N) time, then the most efficient possible computer for that is the universe which computes everything in O(n) time where n is time. In other words, you never get "lag" in reality no matter how complex the scene you're looking at is. Worse than that, some simulation algorithms are exponential time complexity!
- chowells 5y ago
  That doesn't prove or disprove anything. What we experience as time would be part of the simulation, were such a hypothesis true. As such, the way in which we experience it is fully independent from whatever costs it might have to compute.
  - narrator 5y ago
    So you're saying that an exponential time complexity algorithm with N of every atom in the universe will complete before the heat death of the other universe that the simulation is taking place in? Sorry, not plausible.
    - Bjartr 5y ago
      Why does the containing universe necessarily have comparable physical laws?
    - Jensson 5y ago
      Our laws of physics are space partitioned so the algorithm for simulating it isn't exponential. If the containing universe has like 21 dimensions and otherwise have similar tech computers as we do today then you should be able to simulate it on a datacenter just fine as computation ability grows exponentially with number of dimensions. 3 dimensions you have 2 dimensions of computation surface, 21 dimensions and you have 20 dimensions of computation surface, so our current computation to the power of 10. GPT3 used more than a petaflop real time compute during training, so 10 to the power of 15. Using the same hardware in our fictive universe would give us 10 to the power of 150 flops. We estimate atoms in the universe to be about 10 to the power of 80, with this computer we would have 10 to the power of 70 flops of compute per atom, that should be enough even if entanglement gets a bit messy. We have around that much memory per atom as well, so can compute a lot of small boxes and sum over all of it etc, to emulate particle waves. We wouldn't be able to detect computational anomalies on that small scale, so we can't say that there isn't such a computer emulating us.
carrolldunham 5y ago
GPT is a joke but Ashby's various "Laws" and "Theorems" are just theories. As in, think "Jung's theory of archetypes". Cyberneticists cribbed hard-sciences terminology for their psycho-speculation freely, knowing their humanities peers couldn't call them to account about technical meanings. Theorem and law just sound engineer-y