as part of @RBAInfo @DigiFinanceCRC Project Acacia, here's a primer on pubilc versus permissioned DLTs (distributed ledger technology), and on the back of the @RedbellyNetwork Insights podcast on right now (see xitter.cf/RedbellyNetwork/status…), here's some explainers First up Public vs Private DLT Private Permissioned DLT: - Only selected, known participants (such as specific banks or financial institutions) are allowed to join and validate transactions. - Access is tightly controlled and not open to the general public; every participant is vetted and given explicit permission to participate. - Used when confidentiality and control are top priorities, often for sensitive financial transactions. Public Permissioned DLT (think more eligibility criteria which depends on market as @RedbellyAlan mentions in the show): - The underlying platform is open and potentially accessible to anyone for certain functions (like reading the blockchain or using apps), but only approved participants can validate transactions or issue assets. - A wider group can interact with the platform, but core functions (like settling tokenised assets or issuing CBDC) are restricted to those given explicit permission. - Balances transparency and openness with the need for regulatory control and permissioned access. In summary: So in summary - Private permissioned = access and participation are tightly restricted to a closed group. - Public permissioned = the network is broadly accessible, but only whitelisted participants can perform critical functions.

OpenAI Dev Day came and went and apart from Dots (anyone setup theirs already?) the next most interesting feature was Plugins. Imagine being able to have the app you've built sitting inside of ChatGPT. Here's @ETFtrackerAU doing just that. One interesting thing is that you can get data to come across easily (thanks to building out an MCP for ETFtracker) and whilst it can easily use ChatGPT's native charts (image 1). A tip though for those building... getting it to use your apps charting libraries is a bit hit and miss as is it fails about half the time when loading it (despite looking better design wise per image 2). Hopefully won't be too long till this can just be publicly available and if you're keen to learn how to do this, hit me up! @foundrylabsau @ausblockai @DECAustralia

If this independent investigation into the Medicare hack is close to the truth then this New Yorker style cartoon still stands (the article: lnkd.in/gtR5U6yH). TL;DR: The suggestion is that the AI didnt exploit a vulnerability or break through authentication. It hit a blocked front end, found the underlying reporting service and used guest access that the system itself provided. The sites own code may even have exposed the paths needed to reach reports that werent supposed to be public. If that is what happened, the real security failure is that the security boundary wasn't where we thought it was. I had a higher bar for what a "hack" is, and definitions are important. But let's see how this one plays out. If you flip the script here, this is actually a major opportunity to use AI to help highlight flaws in existing systems, starting with the most important ones, like Medicare. @foundrylabsau

new insights from ETFtracker on ASX ETFs

Let's go!

Experts in the Loop (LIVE) - 3rd July 2026 xitter.cf/i/broadcasts/1rxmqqvzZ…

I'm glad Belgium had their 12th player working with them for this game hahaha - Senegal were robbed

Live with Restream, June 19 xitter.cf/i/broadcasts/1vJpPPjDw…

check out the latest Experts in the Loop podcast where @digisincas and I chat with @ArmsRosenberg on investing and AI youtu.be/nLbRG3dSz44?si=uYO_…