Author: Good Data (https://xitter.cf/t.co/l70CeA6ZzI). Researcher: tech & politics @BennettInst. Advisor @StateUpHQ. Bestyrelsesmedlem @DataFGood. Previously @manypets_uk
Copenhagen, Denmark
Joined October 2011
- Tweets4.4K
- Following1.9K
- Followers1.6K
- Likes3.6K
Pinned Tweet
Good Data is published today!
Inside you will find:
🎯 A critique of surveillance capitalism theory
🔓 A case for data openness
🦸♀️ A new theory of digital power
☕️ Some hilarious startup anecdotes
🐶 Unexpected pugs
It's widely available (eg: amzn.to/2O2nzLD)
ALT The author Sam Gilbert holding a copy of the hardback edition of his book Good Data: An Optimist's Guide to Our Digital Future
A clear, engaging and very welcome explanation & critique of the argument that AI poses an existential risk to humanity
Delighted to publish this excellent conversation with @danwilliamsphil, on the existential risk from AI. For those of you want to understand the arguments from both sides of this debate, start here. ianleslie.substack.com/p/new…
Sam Gilbert retweeted
If you want to understand why it’s so hard to pause the AI race, the best place to start is Prozac
My latest for @skynews:
Pretty wild that the Foreword to the PISA report was clearly AI-assisted
ALT Screenshot from page 7 of OECD, PISA 2025 Results (Volume I), with the text "AI won’t just automate our tasks – it could quietly erode some of the foundations of our civic life." highlighted
Media and public discussion of the economy usually takes for granted that growth is both well-understood and benign. But why should it?
bennettschool.cam.ac.uk/blog…
Economists who engage with degrowthers tend to poke fun at what they perceive as their naivety, utopianism, and lack of intellectual rigour.
noahpinion.blog/p/degrowth-w…
But perhaps it’s time for those who believe growth is a national necessity to meet people where they are:
bennettschool.cam.ac.uk/blog…
Sam Gilbert retweeted
"Since I use the internet only for specific purposes and not very intensively, I have no experience of social networks like Facebook and cannot speak to the solidarising effect of electronic communication, if there is any" - Habermas in 2010
Sam Gilbert retweeted
"There is little sense in discouraging citizens from using ChatGPT and other generative AI tools for environmental reasons" @samgib writes, but "transparency requirements may be needed to force AI companies [so their] carbon footprint can be quantified" bennettinstitute.cam.ac.uk/b…
I suspect this finding will surprise everyone who has never run a Facebook Ads campaign and no one who has.
A feature of the format is constant feedback from users - advertisers are incentivized to run ads people like (or at least don't object to).
These are ads for the AI companion app Replika, run from burner Facebook pages.
A current FTC complaint claims Replika harms users by encouraging emotional dependence on its companions.
I've made an archive of the ads for online safety researchers & campaigners [link in reply]
Replika Facebook ads archive & commentary: docs.google.com/document/d/1…
Sam Gilbert retweeted
I've been waiting for someone to compile what we know + don't know about the environmental impacts of "AI." Sam does a fine job. While there are real impacts, it's also clear people have been significantly overplaying them, especially in comparison to lots of ordinary activities.
Should we be worried about the climate impact of our generative AI use? 🤔
It's complicated. Drawing on work by @Laurenebridges @STS_News @SashaMTL @AndyMasley + data from Ofgem, Water UK & the IEA, I tried to bottom it out 🙏
Warning: may distress roast potato lovers 🥔
Should we be worried about the climate impact of our generative AI use? 🤔
It's complicated. Drawing on work by @Laurenebridges @STS_News @SashaMTL @AndyMasley + data from Ofgem, Water UK & the IEA, I tried to bottom it out 🙏
Warning: may distress roast potato lovers 🥔
Is ChatGPT bad for the environment?
Rather than focusing on individual generative AI use, more transparency is needed from AI companies on the climate impact of developing their models, writes @samgilb
bennettinstitute.cam.ac.uk/b…
The Department of Government Efficiency sure sounds like a case of crypto populism
bennettinstitute.cam.ac.uk/b…
Always nice to get a boost from @syllabus_tweets 🙏
🔗the-syllabus.com/cabinet/syl…
In a new paper in the Journal of Moral Philosophy,@AmandaRGreene and I argue that both the sovereign and the structural models of power are inadequate when it comes to tech companies like Meta and OpenAI.
doi.org/10.1163/17455243-202…
1/🧵
Sam Gilbert retweeted
In the 2nd part of this blog, @samgilb discusses the implications of policies which seek to restrict or ban business models based on targeted digital advertising. He says such policies can be expected to have regressive effects when applied to tech giants.
bennettinstitute.cam.ac.uk/b…
Is targeted digital advertising based on individuals' online activity morally wrong?
In a 2-part blog, @samgilb suggests that targeted advertising can actually help the Global South & promote fairness.
bennettinstitute.cam.ac.uk/b…
Sam Gilbert retweeted
In the first of a 2-part blog, @samgilb challenges the widely-held belief that targeted digital advertising is morally wrong.
He argues that business models based on targeted advertising could benefit the Global South & even enable distributive justice.
bennettinstitute.cam.ac.uk/b…
