Hmm, well you could code your own custom pattern matching / substituting functions in MeTTa for sure, but there will also be some "hard coded" functions for static pattern matching in the MeTTa interpreter (as implemented in Rust), which are tuned for efficient interoperation with Distributed Atomspace ...
On Fri, Oct 15, 2021 at 2:05 AM Ivan V. <[email protected]> wrote: > > Hello, > > I skimmed over explanatory docs, but didn't find some functionalities of my > interest. Specifically, I'm interested in manual controlling of (1) finding > expressions by stating patterns and (2) substituting found expressions; in > both (1) forward and (2) backward chaining directions. It seems that these > are somehow automated now, but I'd like to have full control over these at > the meta level. As the current AtomSpace has atoms representing these > functionalities, I wonder will there be some equivalents in MeTTa? > > Otherwise, I welcome the idea of MeTTa being a human-use-intended language > and introduction of gradual types. I especially like opting out for minimal > core features (`:`, `->`, `=`) of which all other functionalities would be > composed. > > All best, > Ivan > > > čet, 14. lis 2021. u 21:19 Ben Goertzel <[email protected]> napisao je: >> >> Here is a somewhat concise, not-maximally-didactic description of the >> new language (MeTTa) we have designed for OpenCog Hyperon: >> >> https://wiki.opencog.org/w/File:MeTTa_Specification.pdf >> >> This is intended as a language for human developer use, and also as a >> language for structuring the internals of the Hyperon Atomspace. >> >> In Hyperon rather than doing like in OpenCog Classic and writing >> Scheme / python / Haskell /C++ scripts for manipulating Atomspace, one >> would write MeTTa scripts. >> >> Reasoning, program learning and other Atomspace-based AI methods would >> be conducted on MeTTA programs as well... >> >> Note that the textual-for-humans language syntax is not yet >> determined; the above document uses vaguely LISP-like syntax but we >> might opt for Haskell or Idris like instead. Current implementation >> work in Rust is working w the LISP-ish syntax, but we may add a whole >> different syntactic-sugar layer. >> >> Some requirements behind the MeTTa design are given here: >> >> https://wiki.opencog.org/w/File:Basic_Atomese_Features_required.pdf >> >> And some earlier related design thoughts are here: >> >> https://wiki.opencog.org/w/Hyperon:Atomese >> >> How to store and access MeTTa programs/expressions is a different >> question not addressed in these documents. There have been some new >> developments in the design and prototyping of a distributed Atomspace >> on top of a combination of existing no-SQL databases which will be >> covered in other documents. Complementing a distributed Atomspace >> store there will need to be a local in-RAM Atomspace cache on each >> machine running a MeTTa interpreter, and the design of this is still >> in progress; but it's worth noting that here some serious deviations >> from the current OpenCog Classic Atomspace are being considered; for >> instance, it may be that the "atomic" element stored in this cache is >> actually the "MeTTa expression" rather than the individual Atom >> (though individual Atoms would still exist conceptually within MeTTa >> expressions). >> >> Alexey Potapov will give a talk on MeTTa in the Oct 15 workshop on >> "Scaling up Neural-Symbolic and Integrative AGI Architectures" at >> AGI-21, which will be mixed virtual/F2F and live-streamed on >> SingularityNET YOuTube. >> >> https://wiki.opencog.org/w/Workshop_on_Scaling_up_Neural-Symbolic_and_Integrative_AGI_Architectures >> >> The name MeTTa is an acronym for Meta Type Talk (because the language >> is gradually typed, allowing interoperation of multiple type systems >> w/ their own type checkers, and is intended to foster runtime human >> and AI creation of type systems including funky ones like >> probabilistic dependent type systems etc.). Metta is also a Pali >> word for lovingkindness, which has some AGI ethics resonance. >> >> >> -- Ben >> >> -- >> Ben Goertzel, PhD >> [email protected] >> >> "My humanity is a constant self-overcoming" -- Friedrich Nietzsche >> >> -- >> You received this message because you are subscribed to the Google Groups >> "opencog" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected]. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/opencog/CACYTDBcPcUt7UR4KTgkB7asFqOEu4VG7Hkot0tTb4O2%3DuZ_USA%40mail.gmail.com. > > -- > You received this message because you are subscribed to the Google Groups > "opencog" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/opencog/CAB5%3Dj6W4s%3Dhvi%3Dcr%3DTN2Te4W92fsGzQ2A7gdDs%2B8DvNQ1M_J%3Dw%40mail.gmail.com. -- Ben Goertzel, PhD [email protected] "My humanity is a constant self-overcoming" -- Friedrich Nietzsche -- You received this message because you are subscribed to the Google Groups "opencog" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/opencog/CACYTDBeLYZ%3D_mxvC357u%3DOcBi8sbCcTKB9sVedYbwosXWGbkSQ%40mail.gmail.com.
