mastodon 4.8.0-alpha.3digital sociologist / strategist / activist. Open Future (EU), Communia (EU), Creative Commons (World), Centrum Cyfrowe (PL)
New piece by Mariana Mazzucato argues that there's no "collective intelligence" without the "common good". Probably obvious to most of people here. She makes a good argument for the not-yet convinced.
And "collective intelligence" is indeed becoming an overly-romantic buzzword and narrative.
Same risk, by the way, appllies to narratives about communities and communal governance.
I just read a July Euractiv piece about Meta's AI models in Europe. The company is signalling that it will no longer release models in the EU, due to what they perceive "unpredictable regulatory environment"
(Just as a reminder, this environment stopped meta from training it's AI on our personal data)
This is a well known move in Meta's playbook - reacting to regulation by moving out of a market (or threatening to do so).
The novelty of the case comes from the fact that we are talking about open models.
So it will be interesting to see how Meta would go about introducing geographical limits on openly released models - I find this hard to imagine, but who knows?
Dan Cohen and Dave Hansen wrote recently a really good piece on books, libraries and AI training (the piece refers to the paper on Books Data Commons that I co-authored).
They start with a well-known argument about levelling the field: without offering public access to training resources, AI monopolies will benefit from information asymmetries. Google already has access to 40 million scanned books.
They add to this a key point about libraries' public interest stance - and suggest that libraries could actively govern / gatekeep access to books.
This reminds me of the recent paper by Melanie Dulong de Rosnay and Yaniv Benhamou, which for me is groundbreaking - it proposes that license-based approaches to sharing are combined with trusted institutions that offer more fine-grained access governance.
So it's good to see that this line of thinking is getting traction.
https://www.authorsalliance.org/2024/05/13/books-are-big-ais-achilles-heel/
Interesting data from a new edition of the Foundation Model Transaprency Index - collected six months after the initial index was released.
Overall, there's big improvement, with average score jumping from 37 to 58 point (out of a 100). That's a lot!
The interesting fact is that researchers contacted developers and solicited data - interactions count.
More importantly, there is little improvement, and little overall transparency in a category that researchers describe as "upstream": on data, labour and compute that goes into training. And "data access" gets the lowest score of all the parameters.
More at Tech Policy Press: https://www.techpolicy.press/the-foundation-model-transparency-index-what-changed-in-6-months/
. henryfarrell@mastodon.social wrote a great essay, outlining a political economy of #generativeAI.
My thinking aligns with his in a lot of ways, and I especially like:
✦ how he takes the "Shoggoth" metaphor, often used to incite moral panic about AGI, and shows that corporations are the real Shoggoths that we should be worried about
✦ how he deploys the "map and territory" metaphor to describe political stakes related to genAI - the struggle is for control of technologies, with the help of which there are increasing attempts to replace maps for real territories
✦ how he notes a reconfiguration of political positions of activists and organizations like Open Future Foundation - and signals the need for a new advocacy agenda based on a good understanding of emergent ways of creating synthetic knowledge and culture, and focused on supporting and protecting human knowledge.
https://www.programmablemutter.com/p/the-political-economy-of-ai
I saw a link to a LinkedIn profile called "Relevance Models". I assumed it's about frontiers of LLM development. Turned out to be a page of a modelling agency. To my surprise!
A sign of times.
Meta released a set of new research #LLM, based on a new, multi-token prediction model. These are licensed ... differently than the Llama models! Under what seems to be a new flavour of a *non-commercial* open source license.
The obvious, but not very interesting by now, comment is that this is not #opensource AI.
A more interesting observation is that Meta is showing a lot of experimentation with licensing of their AI materials.
Which is symptomatic of a general state of licensing experimentation in this space, we're living through an open-ish licensing summer!
@luis_in_brief@social.coop what's your take on this?
"Archival Images of AI", great new project from Sound & Vision.
It starts with a simple premise: you can usage heritage collections to make images of AI that are better than the ever-present illustrations of humanoid robots.
And turns out to be an opportunity to ask serious questions about heritage, the commons, and AI.
https://aixdesign.co/posts/archival-images-of-ai
#aicommons #ai #DigitalHeritage
"I’ve decided to stop fighting it and just accept that a Liza Minnelli updates twitter account will be how i get my news from now on."
@taylorlorenz@mastodon.social wrote a good piece on the changing online news media landscape, analysing the case of Biden's announcement.
("Changing" might be too weak of a word to describe the craziness that Taylor analyses).
Basically: less trusted media sources, more of everything else: a decentralized, distributed soup of messaging that you'd rarely call news or information in the past.
As it often happens, my first thought was: McLuhan sort of predicted this.
And then I thought of Benedict Andersson and his telling of how newspapers built our modern societies.
If this is a post-news media world, then we need to look at -pre-newspaper times to understand it. And we might not be modern anymore.
Also, where does #Wikimedia fit in all of this, and how does it adapt?