With thanks to a friendly donation, I have a new development machine up and running for the Tummi project, Tank0:

- Intel Core i5-6500, 4x3.2GHz (Skylake 14nm from 2015)
- 2x16GB DDR4 2133 RAM, 2x~17GB/s
- 2x128GB SATAIII MLC SSD, ~500|~200MB/s for the OS
- 2x4TB SATAIII server HDDs, RAID1, ~200MB/s for the data
- 1x2TB NVMe M.2 SSD, ~4GB/s for storing and querying the memepool

OS is a Debian 12 Linux, will switch from Ruby to Python as programming language, and still have to take a deeper look into available RDF-database systems.

Roadmap Update 2023

  • Tummi v0101, proof of concept on one SQuAD Wikipedia article
  • Tummi v0201, proof of concpet on all SQuAD Wikipedia articles
  • Tummi v0301, proof of concept on all englisch Wikipedia articles
  • Tummi v0302, CWFB parser
  • Tummi v0303, Book1 parser
  • Tummi v0401, take a look into SAT
  • Tummi v0501, Pascal module
  • Tummi v0502, Winograd module
  • Tummi v0601, Lopez module
  • Pi engine v0101, Theta A
  • Pi engine v0201, Theta B

"If you want to make the gods laugh, start making plans."
A Greek proverb.

***updated on 2024-02-09***

Chess Game Tree Complexity and Memeplex Knowledge Graph Complexity

In the previous post I stated there is no way around to build an artificial Ego to be able to draw new conclusions. Of course there is, but I doubt such a machine will be build in my lifetime.

The game of chess has about 10^50 possible positions and about 10^120 possible games. To compute the whole game tree to find the perfect play was, is, and probable will always be not computeable on classic computers. We have the Quantum-Computers in pipe, and it is yet unknown if such a perfect-play engine is possible on such a machine. Current computer chess engines on von Neumann computers use heuristics to find the best move via an in depth limited tree search, and they do this meanwhile on a super-human level.

So, we can view our process of thinking similar to engines playing chess, we use our mind to search the memeplex knowledge graph for answers and solutions, we search the known graph we call knowledge, and we search the unknown what we call intuition, creativity and alike.

So, yes, in theory we can build a machine, maybe an Hyper-Meme-Machine, which expands the whole possible knowledge graph at once and runs some kind of evalution on possible candidates of new conclusions. All without the need of an artificial Ego.

The question which remains open is if such an SKGS, speculative knowledge graph search, can be implemented in practicable manner on our classic computers nowadays or near future.

A Artificial Ego - Should We Do It?

My intention with this blog was not to get into philosophy, but to document this project, Tummi. The further I walk this path, the further I realize that it is not about building just an oracle kind of machine, an question-answering-system, but to go deeper, to build an system which is able to think, reflect its thinking, and come up with new conclusions. Therefore we have to build some kind of Ego, "Cogito, ergo sum", there is no way around this, and that is the point where philosophy steps in. The human kind does not consist only of the mind, the Ego, I like to see it threefolded, it is the body, mind and soul which makes us as a whole. So, what do we create if we build an artificial Ego, able to think and reflect its thinking, but we are not able to map the other parts? Should we do this? Should we build such an limited Ego, knowing that there is more? I am for sure not the only one who ponders on this, so let me refer to Prof. Dr. Thomas Metzinger:

7. Warum wir es nicht tun sollten in German

7. Why we shouldn't do it on Google Translate

Bottleneck Memory Bandwitdh?

I still have no numbers to extrapolate but assuming a whole Wikipedia with millions of articles parsed as meme pool in an RDF structure will need terabytes of RAM then probably memory bandwidth will be the bottleneck for the SPARQL queries. Currently we have up to 64 cores with 8 DDR4 channels with each ~25 GB/s on a single socket system, IBM's enterprise server may have a bit more throughput, but I guess until we have some kind of 1:1 ratio for cores:channels Tummi won't be able to query the whole Wikipedia in reasonable time...still some time to pass.

Back on the track...

Okay, I dropped the Frankenstein parts and am back on the track, primary goal is simply Tummi, the intended meme machine for natural language processing, then, optional, Tummii - math n algorithms, then, optional, Tummiii - the Xi calculus...seriously, the first level will be already enough to crack for me.


Hmm, okay, this started as an relative simple meme machine in mind, and now I am up to level 5 on pen n paper...

- Epsilon engine level I
- Epsilon engine level II
- Xi calculus
- Ny
- Omikron

Epsilon I was the initial intend, a meme machine for processing natural language, Epsilon II was planed as extension to map math n algorithms on an memetic level, I will not get into the details here what the other levels mean, but it is simply too off and too Frankensteiny for just an hobby project...better play some didgeridoo or alike...

Epsilon Engine

To be more precise, Tummi will be only a kind of front-end to the Epsilon engine, there will be a back-end, Luther, to view and edit the knowledge graphs, and I aim for another front-end, KEN (acronym for Karl Eugen Neumann), for language translations.

I plan different pipes for the Epsilon engine, to address different kind of queries for the same memepool as knowledge graph, but I did not work out any details yet...

Epsilon and front-end code is currently Ruby, with front-ends as simple Sinatra web-applications, database with SPARQL-endpoints will probably be Fuseki from the Jena project, cos they offer RDFs reasoner and SPARUL.

The first Tummi release is aimed as an proof of concept on the SQuA-Dataset with ~500 English Wikipedia articles, we will see how far I will get with this.

Tummi - Milestones

2023 - Roadmap update for Epsilon and Pi engine.

2021 - Roadmap update for Epsilon I, II, III, IV.

2020 - Roadmap for Tummi, Tummii and Tumiii.

2019 - Tummi v0001 pdf flowchart published.

2019 - Blog online.

2018 - Blueprint of an interlingual meme machine based on knowledge graphs
           bootstrapped with human expert knowledge but able to parse content

2018 - Project reopened, Watson didn't make it.

2011 - Project canceled, IBM's Watson wins in Jeopardy.

2010 - First prototype with an simple ontology as knowledge graph.

2008 - Convinced that RDF/SPARQL offer enough flexibility for an meme machine.

2008 - Experiments with neural networks and RDF/SPARQL.

2005 - Experiments with AIML.

2004 - Inspired by Kiwi Logic's virtual agents.

2003 - Convinced that an meme machine could answer IT HelpDesk emails.

2001 - Experiments with OOP and meme replication.

2001 - Journey starts, inspired by 'The Meme Machine' by Susan Blackmore who
           introduces the idea of artificial meme machines.

*** updated on 2021-12-20 ***

Tummi - The ultimate meme machine I

This blog is about Tummi, my attempt to create an artificial meme machine that is able to parse content in natural language and answer questions in natural language.

The last time I started such a hobby project it took me about 10 years to get into the techniques and understand the underlying principles. So maybe anno 2028 I will be able to judge if this blog was a foolish idea or not.

The name Tummi is derived from 'The Meme Machine' by Susan Blackmore and 'The Ultimate Machine' by Claude Shannon.

Home - Top