@jackhardinges
Data optimist. Interested in: data infrastructures, governance, platforms, AI, the commons, etc. Working with: Foresight (1m eye scans), Icebreaker One (data for net zero), Creative Commons (preference signals) + OpenMined (federated data networks).
"a legal directive [will] shortly allow long-standing clinical studies to include secure access to GP records for.. participants in UK Biobank, Our Future Health and Genomics England who have already consented to share their data for research" www.hdruk.ac.uk/news/linking...
Smart Data (regulated data portability for UK consumers and businesses) looks like it might get quite expensive to deliver... assets.publishing.service.gov.uk/media/697b89...
that's it :)
fair point, given the answer to your question is 'quite a lot of it'. it depends on whether the $$$ keeps flowing to support LLM product + diffusal, or whether it follows other branches of research like Gary's. maybe it goes away to find a new hype train altogether and the field 'freezes' again
the importance and relevance of this statement is a function of how AGI-pilled one is. AGI is not my objective, nor (imo) ultimately achievable. despite their inherent incapabilities, LLMs are not a dead end to building things that people want to use - is that not a reasonable objective function?
nice. recent and related (if you haven't seen it already): www.thecut.com/article/broo...
added this to my moodboard on technology's ongoing erosion of leeway and a closing of all the gaps, workarounds, loopholes that previous generations benefitted from. 'Bureaucracies often have something that computers do not: logical escape valves' www.publicstrategist.com/2015/05/leew...
"Any organization building state-of-the-art LLMs needs to be a data curation and generation lab first, obsessively identifying and filling the gaps" huggingface.co/spaces/lvwer...
"Of course, we’re also hoping to eventually offset risk by diversifying into galoots, simpletons, and outright morons. But the important thing now is to get many, many more idiots confused enough to believe they have any chance in hell of making money off this grift”
just watched a brilliant talk by Paul Keller @openfuture.bsky.social. it includes a case for *conditional openness*, building on OF's longstanding argument that the open movement must better account for concentrations of power www.youtube.com/watch?v=ROig...
a job posting for 'Public Sector Data Markets Lead' for the National Data Library. responsible for "interventions to test the potential commercial market for public sector data. They will identify promising opportunities for data commercialisation" www.civilservicejobs.service.gov.uk/csr/jobs.cgi...
"Today, OpenAI is co-founding the Agentic AI Foundation under the Linux Foundation, alongside Anthropic and Block, and with the support of Google, Microsoft, AWS, Bloomberg, and Cloudflare" openai.com/index/agenti...
I've written up some thoughts on large AI models’ evolving relationship with data (with my new @openmined.bsky.social hat on): openmined.org/blog/ai-hasn...
"Data spaces as meta-organisations" www.tandfonline.com/doi/full/10....
“This commoditization will start to bifurcate the market into a spectrum with extremes at either end: a smaller premium market at one end, and a larger commodity market at the other. A classic barbell effect where the middle disappears" radicallyinformed.substack.com/p/beyond-the...
"Conditional openness means that datasets are freely accessible for non-commercial, public-interest uses, while commercial access may be subject to additional conditions— such as registration, purpose declaration, and payment where appropriate" openfuture.eu/wp-content/u...
I've written up a briefing note for @creativecommons.bsky.social on pay-to-crawl: new systems being used by websites to automate compensation for when their content is accessed by AI bots and other machines
ah, I see. I didn't realise they hadn't published those specs
what do you mean? info about the users of the non-paid APIs, or something different?
'The Most Boring Dataset in the World' foresight.org/resource/gre...
Wikimedia's paid-for APIs now generating £8m+ revenues, presumably fuelled by AI users diff.wikimedia.org/2025/11/24/w...
"It’s incoherent to claim that data is extremely valuable and that it should somehow also be free. This doesn’t mean there aren’t reasons to subsidize data access to make it available at no cost, but it means we need to acknowledge the costs" radiant.earth/blog/2025/11...
"The question isn’t whether we can add integrity to AI but whether the architecture permits integrity at all. Today’s AI agents observe the Internet, orient via statistics, decide probabilistically, and act without verification" (via @peterkwells.com) www.schneier.com/blog/archive...
'How AGI became the most consequential conspiracy theory of our time' www.technologyreview.com/2025/10/30/1...
@joshdaddario.bsky.social isn't there a simple ODI guide to publishing good open data? iirc there was a github one?
the Smart Data Group has published a proposal for what the 'Future Entity' for Open Banking + all Smart Data Schemes should look like. Special Purpose Vehicles for each Scheme.. www.linkedin.com/posts/smart-...
VLOM
'Welcome to the Post-Naive Internet Era' www.mozillafoundation.org/en/nothing-p...
"The open movement must likewise evolve from the libertarian ethos of the early internet toward a civic ethos fit for the age of AI. We must design infrastructures.. that preserve the spirit of accessibility while protecting against its weaponisation" sverhulst.medium.com/the-weaponis...
