Theme: AI

  • I understand that he’s taking the platform into the universal app to give the we

    I understand that he’s taking the platform into the universal app to give the west the same utility as asia has with their Wechat. He is integrating the phone, the universal app, a ‘truth ai’, and satellite internet.


    Source date (UTC): 2023-07-24 13:15:53 UTC

    Original post: https://twitter.com/i/web/status/1683466030385905668

    Reply addressees: @BigImpactHumans

    Replying to: https://twitter.com/i/web/status/1683182703418437633

  • RT @elonmusk: We should stop making our appliances so “smart”

    RT @elonmusk: We should stop making our appliances so “smart”


    Source date (UTC): 2023-07-23 17:34:24 UTC

    Original post: https://twitter.com/i/web/status/1683168698025095169

  • Sorry. Musk’s vision is genius. Did you forget Zuckerberg’s shift from facebook

    Sorry. Musk’s vision is genius.
    Did you forget Zuckerberg’s shift from facebook to meta and the great meta failure? I mean, meta’s a total failure. Musk’s vision of the http://x.app (everything app) is genius.


    Source date (UTC): 2023-07-23 11:51:18 UTC

    Original post: https://twitter.com/i/web/status/1683082355613769736

    Reply addressees: @lisa_liberal

    Replying to: https://twitter.com/i/web/status/1683052674738974720

  • The military paid for chips and technlology. Spreadsheets paid for computer chip

    The military paid for chips and technlology.
    Spreadsheets paid for computer chips.
    Games paid for graphic processing.
    Pornography paid for the internet.
    Search engine advertising paid for content
    Social media and dating sites paid for destruction of the family 😉
    The academy paid…


    Source date (UTC): 2023-07-22 17:57:38 UTC

    Original post: https://twitter.com/i/web/status/1682812159607070720

  • its not ram. they maintain state but cant predict. we traded state for predictio

    its not ram. they maintain state but cant predict. we traded state for prediction. can explain if necessary.


    Source date (UTC): 2023-07-19 08:34:13 UTC

    Original post: https://twitter.com/i/web/status/1681583208364646400

    Reply addressees: @BillAdelberger @erasmuse @charlesmurray

    Replying to: https://twitter.com/i/web/status/1681581296995250178

  • I hope working with GPT4 demonstrates that what’s necessary is sufficient disamb

    I hope working with GPT4 demonstrates that what’s necessary is sufficient disambiguation of the criteria necessary for decidability such that an AI can render a decision by satisfying or failing to satisfy the criteria. In other words if an AI can do it we can be certain that a court can do it. Even such when the AI can’t do it what it provides us with is a set of criteria that state why it can’t decide. IMO this will make courtrooms more honest as well.

    Reply addressees: @ToddNQuick1 @PepeFisher


    Source date (UTC): 2023-07-18 18:01:54 UTC

    Original post: https://twitter.com/i/web/status/1681363681202798603

    Replying to: https://twitter.com/i/web/status/1681363130021187585

  • GPT4 NATURAL LAW TESTS (FUN) (this thing is amazing – at least until it runs out

    GPT4 NATURAL LAW TESTS (FUN)
    (this thing is amazing – at least until it runs out of context memory)

    CurtD
    Did we discuss the criteria for truth earlier?

    ChatGPT
    Yes, we did discuss the criteria for truth in our earlier conversation. As we defined it, truth is a performative action, which can be testified to across several dimensions. These dimensions include:

    Realism: The statement or claim aligns with the observable and measurable reality.

    Categorical Identity (Unambiguity): The statement or claim is clear, precise, and unambiguous.

    Internal Consistency (Logic): The statement or claim is logically consistent and does not contradict itself.

    Operational Possibility (Constructability): The statement or claim can be operationalized or constructed in reality.

    External Correspondence (Empirical Correspondence): The statement or claim corresponds with empirical evidence or observations.

    Rational Choice (Bounded Rational Choice): The statement or claim aligns with rational choice theory, considering the limits of human rationality.

    Reciprocity (Reciprocally Bounded Rational Choice): The statement or claim aligns with the principle of reciprocity, considering the limits of human rationality.

    Full Accounting: The statement or claim accounts for all relevant factors and does not omit important information.

    Restitutability: The statement or claim allows for restitution or correction if proven false.

    These criteria provide a comprehensive framework for evaluating the truthfulness of a statement or claim. They emphasize the importance of evidence, logic, clarity, and accountability in the pursuit of truth.

    CurtD
    So can I claim that i saw a unicorn and that I”m speaking truthfully?

    ChatGPT
    Based on the criteria for truth we discussed earlier, your claim of seeing a unicorn would be evaluated across several dimensions:

    Realism: Unicorns, as traditionally depicted (horse-like creatures with a single horn), are not known to exist in observable and measurable reality. This would challenge the realism of your claim.

    Categorical Identity (Unambiguity): The term “unicorn” is clear and unambiguous, so this criterion is met.

    Internal Consistency (Logic): There’s no inherent logical contradiction in claiming to have seen something, so this criterion is met.

    Operational Possibility (Constructability): Seeing a creature is an operationally possible action, so this criterion is met.

    External Correspondence (Empirical Correspondence): Given that unicorns are not empirically observed or documented in our world, this criterion is not met.

    Rational Choice (Bounded Rational Choice): Claiming to see a unicorn doesn’t necessarily align with rational choice theory, as it doesn’t provide a clear benefit or advantage given the lack of empirical evidence for unicorns’ existence.

    Reciprocity (Reciprocally Bounded Rational Choice): This criterion is more applicable to interactions between individuals, so it may not be directly relevant to your claim.

    Full Accounting: Unless you provide more information about the circumstances of your sighting (e.g., where and when it happened, what the unicorn looked like, etc.), this criterion may not be fully met.

    Restitutability: If your claim is proven false, there would need to be a way to correct the misinformation or make restitution.

    So, based on these criteria, your claim of seeing a unicorn would likely be challenged due to the lack of realism and empirical correspondence, and possibly other factors depending on the specifics of your claim.


    Source date (UTC): 2023-07-18 17:58:24 UTC

    Original post: https://twitter.com/i/web/status/1681362799857893385

  • Problem is that Nvidia can’t produce hardware fast enough to serve the market

    Problem is that Nvidia can’t produce hardware fast enough to serve the market.


    Source date (UTC): 2023-07-18 16:58:40 UTC

    Original post: https://twitter.com/i/web/status/1681347769699647493

    Reply addressees: @PepeFisher

    Replying to: https://twitter.com/i/web/status/1681325168621154304

  • Today I’m going to try to teach GPT4 ‘The Method’, or what we sometimes internal

    Today I’m going to try to teach GPT4 ‘The Method’, or what we sometimes internally refer to as the NL-Method, or P-Method. For those who don’t necessarily follow me (us), that means ‘Producing an ordered system of measurement between limits using definitions of terms (dimensions) that are reducible to analogies to subjective experience using operational language and grammar, by the process of disambiguation by enumeration, operationalization, and adversarial serialization, and narrowly defining those terms providing unambiguity.

    (Had a bad night with allergies and such last night.)


    Source date (UTC): 2023-07-18 15:26:54 UTC

    Original post: https://twitter.com/i/web/status/1681324675417227264

  • THERE IS NO AI RACE. ONLY OPENAI, TESLA, AND WANNA-BE’S. I don’t know what the r

    THERE IS NO AI RACE. ONLY OPENAI, TESLA, AND WANNA-BE’S.

    I don’t know what the rest of the world thinks but there is really no comparison between ChatGPT4 and any of the other AI’s. For all intents and purposes, it’s the only general AI capable of any meaningful discussion whatsoever. And the more I use ChatGPT4 the more emergent capability it has that surprises me. If they get the context memory up a bit this thing is going to be unbelievable.

    I mean, I can use GPT4 for research, discussing research, and producing content that’s the result of the research, as well as I can use an exceptionally talented pair of grad students. And given that I work in the MOST cross-disciplinary field in the world, that’s saying something substantive. So far my favorite experience is conducting socratic dialogs with GPT4 that do a far better job of explaining the rather absurd complexity of my work than even working with others.

    So, at least OpenAI for general knowledge, and Tesla for World Modeling, that’s all we have. The convergence will occur between these two, top down (llm) and bottom up (world model), technologies.

    Seriously. Every other tech platform is a total waste of time by comparison.

    Maybe Google will eventually catch up, and maybe X AI will surpass OpenAI because Musk’s ambition is to be even more truthful. But so far the evidence is what it is: google sat on the tech because it was a threat to their search revenue stream. And in doing so they might very well have ceded the field.

    #OpenAI


    Source date (UTC): 2023-07-18 15:22:00 UTC

    Original post: https://twitter.com/i/web/status/1681323441461264384