Talk:DABUS

May 2023 contributions by Periksson28

User:Periksson28 made some contributions in May 2023 which, while helpful in some respects, relied mostly on publications of Stephen Thaler, who is basically the subject of this article. User:Edcolins correctly applied Template:Third-party for this reason.

Periksson28's [https://en.wikipedia.org/w/index.php?title=Special:Contributions/Periksson28 edit history] seems to consist entirely of Thaler-related edits. In 2013, Periksson28 was warned for edit warring. There are three relevant discussions on Talk:Computational creativity where other users complain about Periksson28 adding Thaler material to the article which is not supported by reliable sources.

I deleted the introductory paragraph of this article, authored by Periksson28, which presents Thaler's scientific claims about DABUS as fact. DABUS is notable as the subject of widely-reported litigation initiated by Thaler, but his scientific claims are nonsense. Periksson28 reverted my deletion and referred to "neutral and authoritative peer reviewers" – presumably these citations I deleted:

• Boltuc, Piotr (2021). "Philosophy and Computing Conference at IS4SI 2021". MDPI 2022. 8(1): 149. doi:10.3390/proceedings2022081149. Retrieved 22 May 2023.{{cite journal}}: CS1 maint: unflagged free DOI (link)

• Mikki, Said (2023). "Artificial General Intelligence and Noncomputability: A Dynamical Framework". Journal of Artificial Intelligence and Consciousness. 10: 71–101. doi:10.1142/S2705078522500163. Retrieved 22 May 2023.

These articles do not satisfy WP:SCHOLARSHIP and are not reliable sources for Thaler's scientific claims. See MDPI#Evaluation_and_controversies in relation to the first link and, as for the second, well, just read it. This isn't science.

… we support the claim that infinitary noncomputability might constitute a viable path toward future AGI implementations, especially if coupled with nonlocality and a nonclassical probabilistic structure such as those in the quantum world. A theoretical mathematical framework for realizing AGI through non-Markovian stochastic dynamic systems is then presented and illustrated by describing multi-agent AGI assemblages comprised of interconnected dynamic agents. We envision that such networked dynamical assemblages might be powered by noncomputable physics or arranged in an infinitary structure …

I'll invite Periksson28 to respond to this and leave it to others to revert the reversion, if they agree with me. splintax ^(talk) 01:40, 25 May 2023 (UTC)

In Response:

Respectfully, Splintax, the Mikki paper adeptly explores the ongoing challenges surrounding the development of Artificial General Intelligence (AGI) and draws upon lexicon and concepts widely accepted within the AI community. Published by World Scientific, it introduces a theoretical mathematical framework that offers a promising path towards achieving AGI through non-Markovian stochastic dynamic systems. These systems are comprised of multi-agent AGI assemblages, consisting of interconnected dynamic agents. Pertinently, in the case of DABUS, these dynamic agents are artificial neural modules that dynamically bind together, forming non-Markovian chains for encoding complex ideas and their ramifications. While some readers may not be immediately familiar with these concepts, it is important to recognize that they are firmly rooted in connectionist principles and are far from being regarded as "unscientific."

In the context of DABUS, the term "infinitary" elucidates the countless potential trajectories taken by patterns as they circulate through these non-Markov chains. The nonclassical structure and noncomputability arise from the extensive ensemble averaging of the numerous chaotic processes transpiring within these chains.

I hold the Mikki paper in high regard, acknowledging it as a valid secondary source. Nevertheless, it would be immensely valuable if you could provide a more detailed critique of the paper, allowing for a more nuanced and comprehensive discussion.Periksson28 (talk) 18:19, 25 May 2023 (UTC)

Also, consider these contributors to Journal of Artificial Intelligence and Consciousness: Pentti O. A. Haikonen, Bernard J. Baaars, Stan Franklin, Antonio Chella, Wendell Wallach, Igor Aleksander, Pei Wang, and Ben Goertzel. These are not unscientific lightweights and most can be found referenced on Wikipedia itself. So, I would have to say that the scholarship requirements have been met for this secondary reference by Mikki.Periksson28 (talk) 21:09, 25 May 2023 (UTC)

Does the Mikki's paper support the claims made about "DABUS", or is it a general paper about "artificial general intelligence (AGI)"? This paper is behind a pay wall and its abstract does not mention "DABUS". --Edcolins (talk) 19:53, 26 May 2023 (UTC)

Thanks, Edcolins. Mikki presents a theoretical mathematical framework for realizing AGI through non-Markovian stochastic dynamic systems achieved through multi-agent AGI assemblages consisting of interconnected dynamic agents. DABUS is largely that, therefore Mikki cites it relying upon two primary sources:

Thaler, S. L. [2016] Cycles of insanity and creativity within contemplative neural systems, Med. Hyp. 94, 138–147.

Thaler, S. L. [2021] Vast topological learning and sentient AGI, J. Artif. Intell. Conscious. 8(1), 81–111.

So, Mikki broadly supports DABUS as a path toward AGI, at least theoretically.Periksson28 (talk) 20:17, 28 May 2023 (UTC)

Thaler and IP copyright

Thaler is also pursuing cases about copyright for IP created by an AI. The image A Recent Entrance to Paradise

A Recent Entrance to Paradise

has failed to gain copyright protection because it is created by an AI. Philh-591 (talk) 21:01, 21 August 2023 (UTC)

References