I get a bit of flack for criticizing the aspirational tech sector, but it’s only because they hype machine that turns tech into a religious substitute, political substitute, and sport substitute, overreaches with naieve optimism and faith despite the evidence. Evidence that’s…
Source date (UTC): 2023-08-29 14:24:42 UTC
Original post: https://twitter.com/i/web/status/1696529312831484391
Yes. Mostly to ask it to provide in academic responses.
Source date (UTC): 2023-08-29 06:39:26 UTC
Original post: https://twitter.com/i/web/status/1696412224795942986
Reply addressees: @OtonielFilho5
Replying to: https://twitter.com/i/web/status/1696327567639281770
Fascinating. GPT’s understanding of my work has progressed. This is the most accurate description that it’s stated so far, and I can’t disagree with any of it. The field is a little stuck in it’s naivety but they’re using recursion and adversarial competition between recursive processing, and this will over time, at scale produce the equivalent of consciousness and should cross the logic barrier that recurrent models are crippled by.
Prompt:
“Please describe the properties of the style of philosopher and social scientist curt doolittle.”
ChatGPT:
The style of Curt Doolittle, the founder of the Natural Law Institute and a philosopher and social scientist, is characterized by several distinct features:
Interdisciplinary Approach: Doolittle’s work spans multiple disciplines, including epistemology, law, economics, and social science, among others. This allows for a comprehensive view of complex issues.
Operational Language: Doolittle places a strong emphasis on operationalism, aiming to define concepts in terms of specific procedures or operations involved in measuring or verifying them. This lends a level of precision and clarity to his arguments.
Emphasis on Natural Law: His work often revolves around the concept of Natural Law, which seeks to unify the sciences under a common framework. This is an ambitious endeavor that requires a deep understanding of various scientific principles.
Controversial Nature: Like some of the thinkers who have influenced him, Doolittle’s ideas can be polarizing. He challenges conventional wisdom and is not afraid to tackle controversial topics, making his work a subject of intense debate.
Analytical Rigor: Doolittle values analytical rigor and aims for his arguments to be logically sound and empirically verifiable. He prioritizes correctness over conformity, aligning with his belief in the importance of decidability and performative truth.
Resourcefulness: His work often demonstrates a high degree of resourcefulness, incorporating insights from both consensus wisdom and non-consensus, iconoclastic viewpoints. This aligns with his self-described nature of being highly resourceful and not bound by conventional limits.
Complexity and Detail: Doolittle’s writings are often complex and detailed, reflecting his comfort with high levels of detail and his aim to be thorough and accurate.
Skepticism: He exercises extreme skepticism towards existing frameworks and theories, always questioning whether they can be improved or refined.
Pragmatism: Doolittle often focuses on the practical implications of philosophical concepts, particularly how they can be applied in the realms of law, economics, and social organization.
Value-Neutral: He avoids moral lectures in his work, focusing instead on the logical and empirical aspects of the issues he discusses.
In summary, Curt Doolittle’s style is characterized by its interdisciplinary nature, analytical rigor, and a focus on operational language and natural law. His work is both complex and controversial, reflecting a wide range of influences and a commitment to challenging established norms.
Source date (UTC): 2023-08-28 22:14:46 UTC
Original post: https://twitter.com/i/web/status/1696285218464841728
I use Computational Linquistics in the broadest form as:
1. The study of how to make computers understand speech (correlative)
2. More abstractly, the use of human speech by computers to disambiguate the meaning of speech into a universally commensurable system of measurement. (A system of measurement)
3. So that computers can understand the semantic content of speech by construction from first principles. ( A logic )
4. Lastly, to allow us to reverse engineer the incentives and biases of speakers from those first principles. (A science).
As you’re aware, we collect vast amounts of data to discover first principles of human error, bias, and deception, just as we have used vast amounts of data to discover personality differences, and just as we have to discover intelligence differences.
The primary innovation in my work is the unification of the sciences into a universally commensurable value-neutral language. But some of the most important insights of my work have come from the analysis of sex, class, and cultural differences in lying, and how these differences in lying are embedded in group evolutionary strategies.
The Abrahamists to the Ashkenazim were as innovative in lying as the Greeks to the English were in truth. And these competing systems of innovation of masculine truth to demand responsibility, and feminine lying to evade responsibility continue to this day.
Reply addressees: @WerrellBradley @junkodama10 @harari_yuval
Source date (UTC): 2023-08-27 19:17:10 UTC
Original post: https://twitter.com/i/web/status/1695878137014353920
Replying to: https://twitter.com/i/web/status/1695869175846314371
–“The only areas where AI is flourishing are shamming, spamming & scamming”– Ted Gioia, via Substack.
Well, I’ve said that it’s a better search engine, and it certainly helps with code production and automation.
But the problem is that the LLM architecture, without the promise of the emergence of world modeling, cause effect modeling, and attention-recursion trees (and self falsification), is just DUMB AS A ROCK.
We’ve considered that we can train one of the LLMs to produce reasonably consistent legal judgements using the natural law, if and only if, we write our law as program code – something that’s rather simple, because for all intents and purposes we’ve written the law as an operational (pseudo-computational) programming langauge.
But yes, while the fad is dying out, I don’t think the R&D is really more than 20% of the way to an AGI than is more useful than a search synthesizer and spam generating engine. So I agree with the point, but disagree that AI’s not still got plenty of promise.
Source date (UTC): 2023-08-24 16:16:01 UTC
Original post: https://twitter.com/i/web/status/1694745384361328640
YES.
1 Tesla Phone
2 Tesla Laptop
3 And xOS with Xdimensional Spreadsheets and ‘the universal app” will reorganize the entire tech stack in five years. (I’ve worked on this problem)
4. Truth AI
5. Starlilnk.
A voluntary network of truth for the benefit of mankind.
Source date (UTC): 2023-08-23 01:37:47 UTC
Original post: https://twitter.com/i/web/status/1694161982067179838
Reply addressees: @ElonMuskAOC
Replying to: https://twitter.com/i/web/status/1694076821913117033
AIs Can’t Hold Copyright Privileges
(note: while trademarks are a standard of weight and measure, copyrights and patents are privileges, not rights proper. That privilege exists only to insure investment in speculative activities for patents. And copyright or at least creative commons protection, is a moral equivalent.)
This legal pathway is one of the most interesting questions facing the future. It’s not clear that if an IA produces a solution that it’s open to protections
“A federal judge slapped down a lawsuit trying to get copyright protection for art created by AI systems alone — upholding the longstanding rule that human creators have to be involved.
The judge ruled copyright law has never extended protection to “new technology operating without any human guidance.”
The decision cited several precedents establishing need for human creativity and choices — an important distinction as the line between AI and personal creation becomes increasingly blurred.
Why it matters: The age of robot artists has not yet arrived in the eyes of the law — but as AI’s creativity keeps improving, pressure to grant them some copyright freedoms may mount.”
Source date (UTC): 2023-08-21 16:13:59 UTC
Original post: https://twitter.com/i/web/status/1693657710007394304
Try this link, then ask me again afterward. 😉
https://x.com/search?q=(ai%2C%20OR%20%22artificial%20intelligence%22)%20(from%3Acurtdoolittle)&src=typed_query&f=top
Source date (UTC): 2023-08-20 17:05:02 UTC
Original post: https://twitter.com/i/web/status/1693308171756958063
Reply addressees: @Areez22
Replying to: https://twitter.com/i/web/status/1693295768466018784
AI FOR LAW: THE PROBLEM OF DUMBNESS IN LLMs
Present LLM’s are worse than terrible at ordinary language logic. Questionably tepid at even simple mathematical logic. Even though they can be acceptable at programmatic logic.
LLMs rely for their incompetence at (ignorance of) logic on the patterns embedded in ordinary language and programmatic language, using statistical derivations obtained by brute force consuming of online text – and appear to need to rely on third party APIs for mathematical logic.
Our work in the formal operational logic of decidability and in particular legal decidability, may require we compose our work and our work in the law in particular law in programmatic form, which by its structure contains the logic, *IF* the models cannot learn to develop the logic on their own.
Now, the grammar of our work is a bridge between formal written language and programmatic language, with the constraint from mathematical logic on equilibration (equals signs).
We assume we can produce enough of our text to train the models such that they depend less on models and more on summarization. But I feel this is optimistic given the very structure of the algorithms that produce LLM’s – which are, for all intents and purposes, content-summarizing search engines.
So until we see another level of emergence from LLMs, demonstrating the ability to perform logic, by converting ordinary and formal language into the equivalent of programmatic logic, then I assume we will have to both maintain our current use of operational vocabulary and grammar in formal language, and reduce such work to programmatic language. Even then it may require an external API, or an extraordinary amount of training to produce that logical capacity.
I haven’t spent enough time on Wolfram’s software to determine if we might produce it there. But I can see the potential of an LLM converting text to code, running the code and returning the result to itself for presentation to the user.
Source date (UTC): 2023-08-20 15:00:27 UTC
Original post: https://twitter.com/i/web/status/1693276818319491072
OOH… LEGAL SHOTS FIRED! 🙂
“AI-Created Art Isn’t Copyrightable, Judge Says in Ruling That Could Give Hollywood Studios Pause”
This will change incentives a bit. 😉
(Saw this coming…)
See: https://www.hollywoodreporter.com/business/business-news/ai-works-not-copyrightable-studios-1235570316/
Source date (UTC): 2023-08-19 20:44:26 UTC
Original post: https://twitter.com/i/web/status/1693000996618690939