clairelwilmot.bsky.social
journalist @tbij.bsky.social | visiting postdoc fellow, London School of Economics | writer, journalist, academic | critical epistemology, tech, criminal legal systems, gender, (in)security, (in)equality
www.clairewilmot.com
38 posts
699 followers
260 following
Regular Contributor
Active Commenter
comment in response to
post
“Why is it that we want to be able to predict these things in the first place? Instead of investing in predictive technologies, why not change the social structure?”
comment in response to
post
“This, in turn, invites and incentivizes the research that continues—despite the potential ethical or social harms—to toe the line of racial science.”
comment in response to
post
“Funders are increasingly prioritizing funding for sciences that utilize genetic, neuroscientific, & now AI-based logics as the best tools to understand and potentially solve our social problems.”
comment in response to
post
“So there’s both the legal question: Are we actually making the legal system more efficient toward injustice by scientistically overlooking these structural factors? And the ethical question: What are you asking society to do with this notion of risk?”
comment in response to
post
logicmag.io/out-of-place...
comment in response to
post
good principles in this moment:
"strengthening our collective tools & institutions.. shared commitments to inclusive democracy & human rights, equity & labour rights, environmental responsibility, reconciliation with Indigenous Peoples, good global citizenship."
actionnetwork.org/petitions/pl...
comment in response to
post
Thanks so much Yirga 🙏
comment in response to
post
Special thanks to @pulitzercenter.bsky.social, whose funding & support made my second reporting trip in October possible.
comment in response to
post
I'm so grateful to the dozens of survivors & witnesses who shared their stories with me, particularly the Gebre family, who lose 5 of their sons & grandsons in the massacre.
comment in response to
post
The community is trying to make sense of this unprecedented violence through their religious frameworks, but the sense that St Garima has wandered far away from these lands, along the with the Gospels, continues to haunt survivors.
comment in response to
post
Members of the Garima community believe the Garima Gospels are relics of Saint Garima himself -- their patron saint and historic protector, who has kept the community safe for centuries.
"If the gospels are gone, it means we are gone," one of the monks told me.
comment in response to
post
While the war has been over for over 2 years, the gospels are yet to return to the monastery -- the peace is fragile
Their absence has compounded a sense of spiritual abandonment at Garima.
Over 100 civilians were killed by Eritrean soldiers just weeks before a peace deal was signed, in 2022
comment in response to
post
The Garima Gospels are over 1500 years old. For centuries they have been kept safe inside the Garima monastery
But in 2020, at the start of the Tigray war, Garima's monks hid these irreplaceable texts to protect them from Ethiopian & Eritrean soldiers, who fought against Tigray's regional forces
comment in response to
post
Calls for proposals for these kinds of tools are on the rise, but many aren’t being scrutinized.
Read our overview here:
www.thebureauinvestigates.com/stories/2024...
comment in response to
post
Many of these products (and they are products – most are also used for marketing purposes) tap into a desire for simple, seemingly “scientific” solutions to problems of power, politics, & global inequities.
comment in response to
post
As @timnitgebru.bsky.social explained: “Even if you could accurately detect markers of emotion, that doesn't translate into being able to detect someone’s internal emotional state. And even if these models could do that, they would be extremely unethical.”
comment in response to
post
This means these models assume what people say is evidence of what they think or feel.
But human languages are ambiguous, relational, & embedded in diverse cultural contexts.
comment in response to
post
But these so-called “AI” tools are different versions of machine learning & large language models (LLMs) that emerged from the field of natural language processing (NLP).
comment in response to
post
Some companies even claim they can replicate individuals and societies in order to predict how they will respond to different interventions.
comment in response to
post
Tools that predict or recommend are far from objective, “scientific” technologies — they invariably reflect the worldviews of their creators, prioritizing some social sciences theories over others, & are fed on biased or inadequate training data.
comment in response to
post
Tools trained on past or failed peace agreements tend to recommend approaches that have been tried & failed, & provide only shallow explanations about the political barriers to peace.
comment in response to
post
That risk comes from something called “automation bias” (studies show that people tend to trust automated tools more than they should, & will make consequential decisions based on them) as well as the limitations of these AI models themselves.
comment in response to
post
Conflict negotiators told us that narrow, built-for-purpose tools to enhance access to info or include more voices in poltical dialogues could be useful in some contexts. But they said that tools that “recommend” or “predict” were more risky.
comment in response to
post
These technologies range from narrow, purpose-built tools to more general, predictive ones.