Remember that Turing was not setting intelligence as the goal for AI. He was answering the philosophical question "can machines think?" He needed a reasonable definition of "think" that was appropriate for computers. GPT-4 won the imitation game 54% of the time, above his proposed threshold of 30%. If you want to argue that text prediction isn't the same as thinking, then use a different definition. Likewise for "consciousness", "intelligence", and "understanding".
Turing was aware of the test's shortcomings in his 1950 paper. That's why he gave an example where the computer waited 30 seconds to add two numbers and give the wrong answer. The highest possible score in the Turing test is to be indistinguishable from a human. A smarter machine would fail by being too fast, too helpful, and not making enough mistakes. We have had that since the 1950s. The goal of AI should be to serve humans, not to pretend to be human. AI should be able to do everything that humans can do, but not be limited by what humans can't do. AI should be able to recognize and predict human feelings, but it should not have feelings or claim to have them because feelings are limitations that control us. AI should not be programmed to carry out those predictions in real time because that is indistinguishable from having feelings and also because some people believe that causing suffering in machines would be morally wrong. Of course, we are doing exactly that in the Turing test. On Wed, Jun 19, 2024, 4:23 PM Mike Archbold <[email protected]> wrote: > The problems with using 'consciousness' in your design somehow are > manifold. First of all it is notoriously difficult to define in humans. We > had a meetup event with a writeup featuring a practically unlimited number > of definitions. But then if you GO FURTHER and then apply it to your > machine, that is 'conscious' claims of any nature, despite the arguments > to the contrary, we all know that silicon isn't conscious in any real > definition, and it immediately arouses suspicion. IMO it would be better to > only claim a certain degree of structural and functional similarity with > the mind. > > On Wed, Jun 19, 2024 at 12:59 PM Nanograte Knowledge Technologies < > [email protected]> wrote: > >> PS: That was in response to Matt Mahoney's rather interesting reply. >> >> ------------------------------ >> *From:* Nanograte Knowledge Technologies <[email protected]> >> *Sent:* Wednesday, 19 June 2024 21:06 >> *To:* AGI <[email protected]> >> *Subject:* Re: [agi] Internal Time-Consciousness Machine (ITCM) >> >> You're confirming that you believe as you believe, providing your version >> of evidence that everything we observe is relative. You're also asserting >> how absolute truth cannot exist. This is the conscious you communicating >> with us. >> >> Yet, you believe that a poor excuse for an intelligent machine has passed >> the Turing test because on average it scored 54% human. This is my belief. >> Ever thought that the Turing test is a load of crock? It has to be, because >> relativity dictates that it all happened in the belief system of Mr. >> Turing. This is my opinion base don my belief. >> >> Is there ever an immutable truth, or are we still swimming in a petri >> dish? This is my consciousness asking a repetitive question. >> >> To assert that factual evidence is only possible with mathematics, must >> surely also be founded on belief. How do we know this to be true, other >> than those who believe it and practice it holding to the relative truth >> that their collective consciousness must be more correct than those >> individuals and collectives who do not perform mathematics. >> >> Seems to me, that there must exist different kinds of consciousnesses, as >> many as the persons you might be asking, or factored in by the numbers of >> those attending a consensually-based lecture or seminar. >> >> Now riddle me this. When you tape your nose and mouth shut for long >> enough and die, did you really die, or is it all just a matter of belief >> that you died? >> >> Was this state of bodily death determined mathematically? >> >> Seems to me, there's a while world of reality we're living - and dying - >> in, which we might not even be aware of, let alone consciously engaged with. >> >> Is consciousness then not perhaps, and simply, emotional connection, and >> the absence thereof, unconsciousness? I don't have a belief about this, >> either way. >> ------------------------------ >> *From:* Matt Mahoney <[email protected]> >> *Sent:* Wednesday, 19 June 2024 19:25 >> *To:* AGI <[email protected]> >> *Subject:* Re: [agi] Internal Time-Consciousness Machine (ITCM) >> >> On Wed, Jun 19, 2024, 12:40 AM Nanograte Knowledge Technologies < >> [email protected]> wrote: >> >> In your opinion then, consciousness cannot yet be defined properly, but >> you know for certain that there is no such a thing as a kind of life after >> death, or a soul that leaves earth, even forever? >> >> How do you know such things with such absolute certainty? >> >> >> I don't know anything for certain. Proofs only exist in mathematics, and >> even then we have to start with axioms that we assume to be true. Most of >> what we actually know is based on evidence, and most of that evidence was >> collected by other people that we assume are honest. >> >> I believe the Earth is round even though it looks flat from where I am. I >> have seen pictures from space that I assume are not fake. When I fly to >> Europe, a round planet seems like the simplest explanation for why I have >> to set my watch ahead 6 hours to match the sun, but there could be other >> explanations. I can watch SpaceX launch rockets every few days from my back >> yard in Florida, but I can't really see where they are going. Their website >> videos show them going into orbit, which I assume are not faked. We have an >> organization with members around the globe whose purported purpose is to >> question the shape of the Earth, but whose real purpose is to question how >> we know what is true. >> >> But you ask a fair question. In the US, 73% of adults believe in heaven. >> https://www.pewresearch.org/religion/2021/11/23/views-on-the-afterlife/ >> >> Which is more than the 62% that believe in evolution. >> https://en.wikipedia.org/wiki/Level_of_support_for_evolution >> >> We have to believe most of what we read or hear just to function in >> society. The more we are told something, the more likely it is to be true >> in our minds. Every religion has some form of afterlife. The Bible and >> Quran both say so. Hindus believe in reincarnation, with some claiming to >> have memories from past lives. 41% of Americans believe in ghosts and 20% >> have personally seen them. >> https://sc.edu/uofsc/posts/2023/10/conversation-are-ghosts-real.php >> >> So I can only explain my beliefs. I believe that all human behavior can >> be explained by neurons firing in our brains and how they are connected. We >> have LLMs that pass the Turing test, which is a stricter test for >> consciousness than we apply to babies and animals. I have never seen a >> ghost, although I met people who have. I was told in Sunday school about >> heaven and hell, but I stopped going when I was 10. I believe that >> evolution is the simplest explanation for why we fear death and why we turn >> to religion to cope. >> >> *Artificial General Intelligence List <https://agi.topicbox.com/latest>* > / AGI / see discussions <https://agi.topicbox.com/groups/agi> + > participants <https://agi.topicbox.com/groups/agi/members> + > delivery options <https://agi.topicbox.com/groups/agi/subscription> > Permalink > <https://agi.topicbox.com/groups/agi/T32a7a90b43c665ae-M9a4776c821a2852950263b18> > ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T32a7a90b43c665ae-Mfe8ed06468f969a2d9f02e87 Delivery options: https://agi.topicbox.com/groups/agi/subscription
