full flow (can be repeated 1 on 1 for experiments)

we won't pivot to anything, no MPC, TEE, coprocessor, or other bullshit

check out the benchmarks (they're public, and the code is open sourced, wouldn't be surprised if a round trip to MPC takes longer than homomorphic operations on hfhe)

octra will rewrite the actor scheme on Elixir, math and consensus will be on OCaml

carefully starting to update the main net

all LLM and AI tokens are a scam, no project that works with ML algorithms needs a token, it's only for infrastructure (potentially and in rare cases, for quick adoption)

none the existing projects with bots/agents and other crap are of any use, you should completely ignore each of them and not invest a single cent

synthetic systems (combining biological and digital systems) have potential value but are at a very early stage and cannot have any tokens

bottomline: 100% ai-ml-bio tokens are a complete scam

today, the mainnet is being updated (node ​​by node), it may work unstably

at the end of this process, webcli will support the mainnet and all other features

mini-alpha for devs: there are several working methods.

-development in a native language (Applied Meta Language or octra-assembly)

-WASM (write in whatever you want)

-circle environment (you can deploy a container with any code in the microkernel format with embedded private memory) *

* there will even be support for NIX
https://nixos.org/

** built-in support for FHE + TAPE aka "ZK" (generation of proofs for private apps) directly at the state and kernel level

*** multi storage data commit with HFHE/PVAC encryption support (built-in IPFS on steroids)

no compromises: no TEE, no MPC, no coprocessors, pure math, no trusted environments, all within a single state

write in anything, compile to everything