Attempt of an allegory of our digital present. It is set in a world, in which algorithms and neuronal networks are not abstract ideas but physical entities. Hopefully, this can better explain the technology, their relationship to society and their historical context. The goal is that the underlying mechanics of the world function like a consistent framework of the digital, in which digital entities can be build; a literary sandbox like Minecraft or LEGO but for the digital.

What if in 1989, Tim Berners Lee invented the semantic web instead of the world wide web?

Prequel and Sequel to "Harry Potter and the Methods of Rationality", bridging first into a medieval setting in which the magic system works like incremental programming and then into the present day, in which it works like logical programming. Philosophically, the medieval setting is post-modernism and the present day is analytic philosophy, while HPATMOR is enlightenment rationalism. It's a bit like a digital Alice in wonderland.

Why do official clerks not have the equivalent of a calculator like engineers do but for inference? The fundamental idea of this paper is for ChatGPT-like apps to lose natural language for less energy consumption and more determinism in their answers based on controlled natural languages like ACE; and to capture this new paradigm in a new type of browser that has natural language as its primary interface, here called a semantic web-first browser. The idea is proposed in several design steps, beginning with a simple to use calculator-like program to do inference with natural language (pocket-inferer), for which, when a programmer-mode is turned on, transforms into an IDE-like ACE-editor. The idea is then further developed into a semantic web browser, which can also reference data and queries from the semantic web and later, it is philosphised how a web-paradigm agnostic SemanticWebBrowser could be realized.

This poses a fundamental anthesis to ChatGPT-like apps and LLM-centered visions for the WWW with the biggest merit being to tradeoff natural language for more precision and less energy-consumption through controlled natural language. The five main points this paper makes are:

• 1. AI browsers with their high energy consumption are not suitable for daily usage (not in the near future, if ever).
• 2. There has not yet been found a sufficient interface for the semantic web to be appealing to end-users and reach wider adoption
• 3. Controlled natural language like ACE could serve well as a main interface for semantic data, because they manage to capture the potential of semantic web data better than any visualization ever could
• 4. The best application for this approach would be a new kind of browser, which realizes “language as an interface” for the semantic web
• 5. Derived from language as the main interface, the browser needs to center around the interaction with language and therefore look like a text editor or IDE.
• 6. While showing the merits of the semantic web, the browser should also be “backwards compatible” with the traditional world wide web.

It is largely based on the following work: Kaarel Kaljurand. “Attempto Controlled English as a Semantic Web Language” (2007).

cross-posted from: https://lemmy.world/post/34532775

This idea combines a neuro-symbolic AI system (take a LLM and use it to generate logical code; then make inferences from it, see Neural | Symbolic Type) with Attempto Controlled English, which is a controlled natural language that looks like English but is formally defined and as powerful as first order logic.

The main benefit is that the result of the transformation from document/natural language to the logical language would be readable by not IT experts, as well as editable. They could check the result, add their own rules and facts, as well as queries.

I created a small prototype to show in which direction it would be going (heavily work in progress though). What do you think of this? Would love to here your opinions :)

This idea combines a neuro-symbolic AI system (take a LLM and use it to generate logical code; then make inferences from it, see Neural | Symbolic Type) with Attempto Controlled English, which is a controlled natural language that looks like English but is formally defined and as powerful as first order logic.

The main benefit is that the result of the transformation from document/natural language to the logical language would be readable by not IT experts, as well as editable. They could check the result, add their own rules and facts, as well as queries.

I created a small prototype to show in which direction it would be going (heavily work in progress though). What do you think of this? Would love to here your opinions :)

cross-posted from: https://lemmy.world/post/34416839

The fundamental idea of this paper is for ChatGPT-like apps to lose natural language for less energy consumption and more determinism in their answers based on controlled natural languages like ACE; for the user to be able to modify this trade-off-ratio at will based on LLMs (which is not possible when starting from a ChatGPT-like app); and to capture this new paradigm in a new type of browser that has natural language as its primary interface, here called a semantic web-first browser.

The fundamental idea of this paper is for ChatGPT-like apps to lose natural language for less energy consumption and more determinism in their answers based on controlled natural languages like ACE; for the user to be able to modify this trade-off-ratio at will based on LLMs (which is not possible when starting from a ChatGPT-like app); and to capture this new paradigm in a new type of browser that has natural language as its primary interface, here called a semantic web-first browser.

In its finished state, the software is supposed to work like this:

• Users start Anthem, which act as a peer in the network
• Once they have found other peers, they connect to a master node, which coordinates the training
• All peers then train a model with their uploaded songs
• Then they generate songs from their local models
• Each user can now listen to and generate their own AI songs

Would love to hear any feedback from you guys :) Does this make the current AI situation better or worse? I could imagine it acting as a protest against AI companies (similar to what Napster did), but I'm not sure which effect it will have on smaller artists (after all, Napster basically lead to music labels coming under the hood of Apple's ecosystem, which in the end, wasn't so good for artists but on the other hand also brought the Fediverse into being ... I'm not sure).

I did a prototype implementation of a “network of ML networks” - an internet-like protocol for federated learning where nodes can discover, join, and migrate between different learning groups based on performance metrics (Repo: https://github.com/bluebbberry/MyceliumNetServer). It's build on Flower AI.

Want do you think of this? It could be used to build a Napster/BitTorrent-like app on this to collaboratively train and share arbitrary machine learning models with other people while keeping data private and only sharing gradients instead of whole models to save bandwidth. Would this be a good counter-weight for big AI companies or actually make things worse?

Would love to hear your opinion ;)