Sly Gryphon @sgryphon@qoto.org
Lead Consultant @Telstra, doing Internet-of-Things (#IoT), #dotnet, #blockchain, #DevSecOps. Certified Azure IoT Developer, MCSD: ALM, #PRINCE2, Scrum. Tabletop gamer.
Joined Oct 2020
The #ATO Australian Tax Return for Companies is a #poordesign form. It is a multi-page form, with numbered item sections, and then letter labels for individual fields. However the combination of page + item number + label letter is not unique, so you have to refer to items by their full name. e.g. page 6, item 8 - label K, could refer to item K - Commercial Debt forgiveness, or K - Unfranked dividends paid.
Just found out my #tesla has a Bioweapon Defense Mode !
#tesla #fsd now launched in #Australia. First media reports now coming out. https://www.youtube.com/watch?v=h406fHRsI_o&t=35s
#TP-Link #Tapo P110M latest firmware 1.3 now supports #Matter device 'Electrical Sensor', with clusters for Power Measurement (0x90) and #Energy Measurement (0x91), joining with #Eve Energy and #Meross to allow native power monitoring via Matter. A few years later than expected, but at least it is here now. https://community.tp-link.com/us/smart-home/forum/topic/835084
AGENTS.md is emerging as the standard instructions file for agentic coding.
The timeline is amazingly fast:
https://sgryphon.gamertheory.net/2025/07/agents-md-standardisation-for-agentic-coding-systems/
First autonomous home delivery of a vehicle as a #tesla drives from the factory and delivers itself to its new owner. https://www.youtube.com/watch?v=GU16hXSSGKs
@JoeGrowling @tante and the human brain neurons "just fire their electrical charge when their input charge is above a threshold".
It is not the underlying simple mechanism that is relevant, but the emergent complexity. LLMs are so good at modelling thought patterns (to determine the next word) that they are thought patterns.
@sgryphon@mastodon.au what about mixed replies @sly.gryphon@threads.net ?
@sgryphon@mastodon.au can see other Fediverse
@IcarusGames after a bit of a hiatus, my wife is finishing off the Dragonlance campaign she has been running for a few years, with only a few sessions left starting Sunday.
@juergen_hubert Thunderbird, even when on Windows for my personal mail (work is Outlook, but I only use the web version generally).
Sure, #AI #LLMs are "next word predictors", but to do so optimally you need to model the entire process that produces those words. This involves not only language, but also human reasoning, physics, history, economics, politics, etc. In essence predicting the next word optimally, based on a data set generated by human brains, is the same as simulating the human #brain.
Taken from: https://www.reddit.com/r/MachineLearning/comments/12qe5hm/d_microsoft_research_paper_sparks_of_artificial/
@Legit_Spaghetti have a look at the old game Dream Park, where you play a person playing a LARP i.e. both the player and the LARP character
Lead Consultant @Telstra, doing Internet-of-Things (#IoT), #dotnet, #blockchain, #DevSecOps. Certified Azure IoT Developer, MCSD: ALM, #PRINCE2, Scrum. Tabletop gamer.
Joined Oct 2020
