@zeerak
Thinking about content moderation, equity, machine learning, and natural language processing. Now: Chancellor's Fellow (~Asst. Prof) @technomoralfutures.bsky.social @edinburgh-uni.bsky.social Past: MBZUAI, SFU, Uni of. {Sheffield, CPH}
WSJ reporting that the U.S. used Claude for the air strikes in Iran. Centcom has been using Claude "for intelligence assessments, target identification and simulating battle scenarios" www.wsj.com/livecoverage...
Gaza, south Lebanon, the West Bank, southwest Syria, this is the model that Israel wants across MENA: weak, permanently unstable, incapable of infrastructural reconstruction or administrative cohesion.
Somehow I don't recall Hitler sitting idly by just minding his own business and maintaining existing borders of Germany.
But go away. Read a little and maybe you can educate yourself out of the white man's burden.
The future of the any state has to be decided by its people, not belligerents.
🙄 What a lazy racist argument.
Me saying "Unprovoked attacks by a state on another state are illegal" somehow provokes "omg brown person must be on the side of oppressive regimes if they have a basic understanding of international law."
Legality of aggression doesn't change just because of where in the world the country is located.
The facts are pretty simple: there wasn't escalation from the Irani side, even by proxy, and the attack was thus unprovoked. Making it illegal.
The Irani regime being beyond awful doesn't change that
Ukraine doesn't fall into a different category and aggression is illegal.
For 1) there is a unilateral aggressor (Russia, US/Israel) targeting another state not because or some attacks or other diplomatic escalations (according to Oman, Iran was about to agree to not stockpile enriched uranium).
Iran must be bombing itself
For gods sake. It's meant to say young people from the global south.
One day I'll write about how ML-related academia and industry relies on a broad group of hopeful young people with interest/degrees in ML-related fields to help develop our multilingual benchmarks and then left once we're done exploiting their labour and resources.
🤷♂️We really can’t see the problem with that...
Honestly, this is the kind of organic messaging @prisonculture.bsky.social was talking about earlier. We don't have to wait for the media to take up these issues on their own. Make it all about these issues.
The headline does not convey how completely batshit this story is. The Archive Today (archive.ph etc) admin weaponized the site's captcha to attack a blogger who wrote about them and *altered archived screenshots* as part of the attack.
arstechnica.com/tech-policy/...
A black and white illustration with text reading "Substack is creating a walled garden you have to share with white supremacists". The illustration shows a round brick wall holding part of quote, some flowers, and a draped nazi flag. this is surrounded by barbed wire
I have watched more & more friends turn to Substack in the past few months. PLEASE DON'T DO THIS! #Substack has well-documented problems with platforming literal Nazis, as well as writers spreading Covid misinformation, anti-vaccine propaganda, & transphobia. Long post here:
bit.ly/substacksux
This post is going around again so I will add this excellent resource: leavesubstack.com
Really excellent reporting on how Palantir has repeatedly courted Swiss authorities and been rejected no matter how pathetically they acted.
www.republik.ch/2026/02/18/h...
🚀 Launching Every Eval Ever: Toward a Common Language for AI Eval Reporting 🚀
A shared schema + crowdsourced repository so we can finally compare evals across frameworks and stop rerunning everything from scratch 🔧
A tale of broken AI evals 🧵👇
evalevalai.com/projects/eve...
I've had to take bullshit from customers with a smile for MY ENTIRE SERVICE CAREER and was almost never allowed to draw boundaries or turn anyone away, but the AI gets to have boundaries???
Yesterday my friend told me about her PHILOSOPHY students all using ai for their assignments and like come the fuck on man you’re in Uni doing the subject thats Thinking about Thinking and u refuse to Think
AOIR compiled their risky research guide focused on researcher harms from distressing topics that might be useful: aoir.org/riskyresearc...
I am assembling resources for @aial.ie to mitigate/reduce risks (due to our research) from potential:
1)retaliation, defamation lawsuit etc for work on politically charged topics
2)emotional harm from dealing with sensitive issues (CSAM, hate, etc)
know of any helpful resources? pls share/repost
Jamie Webb, Shannon Vallor and Gina Helfrich sit together in the Edinburgh Futures Institute having a conversation. They are all smiling.
The CTMF brings together philosophers, technologists, policymakers & researchers to tackle one of the defining questions of our time:
How can we build technologies that truly serve human flourishing?
This video captures the heart of that mission & the voices of those shaping it ▶️ edin.ac/4ai5OiC
Can’t decide whether my favourite aspect of gender critical journalists’ commitment to free speech is their litigiousness, their abuse of power to blacklist trans writers and spike pieces by trans (any topic) and pro trans writers (anything inclusive) or their complete capture of literary editors
Picture of two pens. The first is an ordinary ballpoint pen, with the caption “how work feels doing it alone”. The second is a picture of one of those novelty pens with ten colours that you had when you were thirteen, with the caption, “how work feels worh an AI employee”. At their bottom it says “hire an AI employee for $0.97/day”.
yeah, just out of interest, how many people choose the pen on the right for real work or art? See a lot of them in professional workplaces, do you?
“Safety experts and tech critics have long condemned the Ring devices for security risks and privacy violations, not to mention their role in building the largest civilian surveillance network in US history.”
Missing from the coverage of redundancies at Edinburgh: this was done in such a haphazard, uncoordinated way, there are now core teams who have gone from 5 people to 1 with no change in workload or pressure and huge loss in knowledge of institutional process. www.heraldscotland.com/news/2576321...
Above anything else, we’re looking for curiosity, passion, interest, personality. Almost no one is a good writer when they join. I can teach you how to write well but can’t teach you to care enough about what you’re doing to actually do it.
I have some bad news. One, if you need to use ChatGPT for your graduate application letter, you are not ready for graduate school.
Two, no, not "everyone is doing it" and those who don't do it aren't "losing out." ChatGPT is not an advantage.
This is a pretty common mistake I see in prospective students reaching out. The assumption is that polished prose is good prose and that unpolished prose is bad. Both are wrong.
When you write an application / cold email to faculty, we care more about you and what you say than polish.
