Feels like Vegyn has taken the Headache project up many levels artistically with this second album. Beats more nuanced and lyricism behind the synthetic speech delivery also quite different from the first album. Recommend. music.apple.com/nz/album/tha...

Apple really missed a trick with Apple Intelligence and the Shortcuts app. Imagine the power and utility of an LLM-driven system to generate working shortcuts sequences. Explainable traceable no code apps run on-device for everyday tasks. Instead I’m using home-made MCP and Claude code to workaround

What do people mean when they talk about “Sovereign AI”? Since starting as a skeptic, I’ve found it useful to think about these things - and it’s timely to share this paper in response to the NZ Government’s AI Strategy. Let me know what you think. docref.org/brainbox/sov...

Such an interesting read! Such an important topic! What will you do if you lose internet in an emergency??

Budget 2026 for bulk download as structured data!! Launching my campaign now and thinking of everyone right clicking on these PDFs.

Syncopate's DocRef system has been deployed by the New Zealand public sector for the API Guidelines review. We've also shared some AI-generated annotation sets to provide a curated view for users and launched the DocRef blog. This is a significant milestone! www.syncopatelab.com/post/docref-...

Just a diff on today's latest changes to the Residential Tenancy Act 1986 if you want a curated view: docref.org/nz/rta/0160/...

Have you ever seen the first web-page ever? info.cern.ch/hypertext/WW...

I’m self hosting an old iPod music collection from about 2015 pre-streaming. Feels so fresh and clean! Setting it up was a real journey which wound up being supported a LOT by LLMs (particularly Claude) - planning to write something up on my new site soon.

Last work task of the year is a lovely one: sharing that my 2025 plans include leading an exciting new research project on internet infrastructure resilience and climate change www.brainbox.institute/post/interne...

“Where should we put the criteria for categorising artificial intelligence systems that could have a significant impact on society as a whole?” “Put it in Annex XIII.” docref.org/eu/ai/32024r...

“Here’s our regulation on the most advanced technology in human history. We thought we’d use Roman numerals to number some of the key document elements.”

Researcher access to platform data: @rebekahtromble.bsky.social, with support from the Coalition for Independent Technology Research, has kindly shared a DocRef annotation set to support responses to the delegated regulation under the DSA. Consultation closes 10 December: docref.org/eu/ares(2024...

Machine readability is a key requirement for a lot of transparency materials under the DSA, but when it comes to risk assessment and audit reports (the most complex documents) they may as well be finger paintings. Is there any scope to add machine readability requirements to current/future reports?

I’ve added PDFs to Google Pinpoint, a prototype for collaborative highlighting and linking. It’s heavily oriented toward Gemini… I think DocRef.org is better! H/t to @alhohlfeld.bsky.social and @olivermmarsh.bsky.social for tracker and PDFs journaliststudio.google.com/pinpoint/sea...

I want to convert all the DSA risk assessments and audit reports to formats that can be annotated, tagged, shared and given pinpoint links to specific paragraphs and blocks. Like or repost if you’d find that useful. Reach out if you want to collaborate on seeking project funding.

Collecting the #DSA risk and audit reports here: docs.google.com/spreadsheets... Still many lines to fill, curious to see which ones come next

Great response to our webinar on rules as code and digital regulatory infrastructure! Webinar recording and slides available here: www.syncopatelab.com/post/webinar...

Starting in 30m!

#RulesAsCode starter pack, reply if you're doing this and not on it! bsky.app/starter-pack...

Really good interest so far!

🤖 Draft Code of Practice on General-Purpose AI Systems (EU): I'm thinking about publishing this on DocRef with additional links back to the EU AI Act. Let me know if that'd be useful! Example links follow (EU COP on Disinformation, and EU AI Act).

The draft code of practice on general purpose AI mentions a feedback form available via Futurium. I can’t track it down, please share a link with me if you have one 🙏

We’re running a webinar next week to talk about the future we see for digital regulatory infrastructure, incl our tools and approaches for #rulesascode. We’ll also be touching on recent publications from the NZ Ministry for Regulation and PCO on publishing + stewardship. Join us! lu.ma/65rdychb

Using DocRef, we’ve published the OECD Council on Information Integrity’s draft recommendation as well as the EU Code of Practice on Disinformation. Consultations closing this month, including on DSA researcher access regs. Access these instruments and more: www.syncopatelab.com/post/three-n...

Welcome new followers! ICYMI, we’re digitising legal instruments to make them easier to work with. See thread below for the draft DSA regulation on researcher access to data. We’ll be sharing more resources soon!

Periodic reminder that I made a #trustandsafety starter pack. Let me know about new arrivals who should be added to this list!

We’ve digitised the draft regulation on researcher access to platform data to help you make a submission more effectively. To access it, and for more information, use the links below!

Me and @tombcgh.bsky.social have a paper published on the home page of law.mit.edu - it's about governance of #RulesAsCode and AI systems and if that's of interest to you then *hat tip* why hello fellow nerd XD

The EU disinfo code document is formatted like hostile architecture

Celebratory dumplings for dinner! News is out I’ve joined @brainboxinstitute.bsky.social to live my best tech policy think tank life with the team there 🤓💻 🎉🥟 www.brainbox.institute/post/the-bra...

I’m repeating a presentation I gave last week at Victoria University’s AI & Society seminar series on what regulating AI in New Zealand could look like. Do join us! The session will be recorded, but the recording won’t be widely distributed.

From work! We're actually doing cool stuff with good partners. Here's how to Do Transparency Right!

Such a privilege to discuss systematic approaches to meaningful tech transparency with our amazing panellists. You can now access the recording below. Thanks to colleagues at @thegni.bsky.social Integrity Institute @ac-transparency.bsky.social @brainboxinstitute.bsky.social for their support!

Looking forward to facilitating this fantastic panel! Hope you can join us.

It’s fantastic to formally launch the Transparency Initiatives Portal from @ac-transparency.bsky.social - see @techpolicypress.bsky.social to hear why systematic collaborative infrastructure will be important for making globally proliferating transparency frameworks meaningful

If we’re going to maximise the benefits and minimise the harms of AI in New Zealand, we’re going to need to consider a collaborative mixture of different approaches. Follow us here, on LinkedIn, or on our mailing list for updates.

Introducing the NZ AI Policy Tracker, a new tool that centralises New Zealand’s disparate AI regulatory environment and provides a one-stop spot for engaging with relevant materials.  Check it out + let us know what you think: www.brainbox.institute/post/the-bra...

Could Taylor Swift be the catalyst for AI regulation? Brainbox Institute's AI Lead Allyn Robins explores this important question on RNZ Nights with Emile Donovan. Listen here: www.rnz.co.nz/national/pro...

🧵 Disinformation is a notable topic in the recently released report from the New Zealand Independent Electoral Review. In our submission, we called for caution, and we’re encouraged to see the panel has taken a similar approach.

Check out our summary paper of the panel discussion we co-hosted w/ the EU Delegation to NZ + Vic Uni about online platform accountability, ft insights on EU & NZ approaches from Gerard de Graaf, Anjum Rahman, Paul Ash, @tombcgh.bsky.social + Prof Ali Knott: www.brainbox.institute/post/online-...

We co hosted a seminar with the EU delegation to New Zealand on accountability for online platforms while Gerard de Graaf was in town. It was a complex discussion but we’ve summarised it and offered some recommendations in a policy brief, now available here: www.linkedin.com/posts/tom-ba...

It was a privilege to attend and participate in the Asia NZ Foundation ASEAN Australia New Zealand Roundtable in October this year. I've written a piece sharing some opinions and reflections on broader themes, and how New Zealand should respond to them. See here: www.asianz.org.nz/track-2/opin...

The information environment around the war between Israel and Hamas is the first public road test of the now-enforced Digital Services Act during an international crisis. The EU’s actions will inevitably set standards for this landmark law’s impact and future approaches.

I’m at #IGF2023 - let’s talk!

We think it’s time to start shaping up AI regulation for New Zealand.