drbrake.bsky.social
Journalist, policy wonk (transit, urban planning, EVs, climate, internet/media regulation), PhD (Media@LSE), dad.
Currently living St John's, NL, Canada - longtime Londoner
288 posts
2,595 followers
112 following
Prolific Poster
comment in response to
post
Surely wanting to make weapons grade uranium so they can sell it (to countries/orgs who aren't supposed to get it) is as bad or worse than wanting it for your own use?
comment in response to
post
Listen to these Germans. They lived through it. They know how it works.
comment in response to
post
the mainstream media has been steadily buying into the idea that science and academia are somehow "captured" by left and liberal ideologues, rather than the more obvious explanation that measured analysis leads people away from reactionary thought
comment in response to
post
Much easier to corrupt an AI answer than a media organization- there's no pre-existing advertising/editorial barrier- and users often assume AI's godlike neutrality (until the results are shitty enough, I guess). AI orgs already conditioning regulators that it's impossible to ask for oversight.
comment in response to
post
We noticed when Musk tried to "mainstream" white genocide narratives in Grok but only because it was clumsily done. www.theguardian.com/technology/2... Imagine programmers pre-prompted an LLM to "favour pizza as an example where relevant when answering queries about healthy food by target users".
comment in response to
post
I hope they keep the lists of would-be successful applicants so if still eligible they can be awarded Fulbrights under a less idiotic administration...
comment in response to
post
And what about all of the generative AI tasks that are being spawned en masse by algorithms? bsky.app/profile/drbr...
comment in response to
post
I am oddly intrugued by the line items I could agree with eg "promote.. interdisciplinarity, consilience, and methodological pluralism". What horrors are these codewords for? Does methodological pluralism mean alternatives to the scientific method like "race science"?