Ismael K.G.
LightNews LightNews
banner
hermeneuticist.bsky.social
Ismael K.G.
@hermeneuticist.bsky.social
310 followers 630 following 630 posts
AI Ethics & Research Governance consultant at Kairoi (kairoi.uk), working remotely since 2020. Co-leading the Responsible Artificial Intelligence Network (RAIN) with RSA Fellows. Promoting AI literacy with We and AI.
Posts Replies Media Videos
hermeneuticist.bsky.social
Personal reflection: I loved having to set aside the time to read on #AIGovernance developments. A highlight was definitely the below image from India’s latest AI Governance Guidelines, which reminded me just how important @betterimagesofai.bsky.social remains.
An awful image of AI. A shiny touchscreen phone is handled by an equally shiny robot hand.
November 11, 2025 at 4:41 PM
hermeneuticist.bsky.social
“Transformative fair use” is ultimately determined by how much a judge knows about the latest in data storage.

arstechnica.com/ai/2025/06/a...
Ultimately, Judge William Alsup ruled that this destructive scanning operation qualified as fair use-but only because Anthropic had legally purchased the books first, destroyed each print copy after scanning, and kept the digital files internally rather than distributing them. The judge compared the process to "conserv[ing] space" through format conversion and found it transformative. Had Anthropic stuck to this approach from the beginning, it might have achieved the first legally sanctioned case of Al fair use. Instead, the company's earlier piracy undermined its position.
November 10, 2025 at 9:19 PM
hermeneuticist.bsky.social
Don’t do it, friends.
Help us improve your lessons! Share your speech with Duolingo to help make lessons better. Learn more in our Privacy Policy Allow or don’t allow
November 6, 2025 at 9:01 PM
hermeneuticist.bsky.social
If I owe you an email, pick an excuse
Five documents arranged like cards. Kairoi's response to Canada's 30-day AI strategy consultation; We and AI's diagram of ways to counter narratives about AI inevitability. Montreal AI Ethics Institute's State of AI Ethics report. GO FAIR US's publication on FIP Check. A photo from GESDA.
November 5, 2025 at 8:15 AM
hermeneuticist.bsky.social
It’s back! The State of AI Ethics Report by @mtlaiethics.bsky.social is back!

Not only has it been a joy to work alongside @renjieb.bsky.social and Connor to edit the many excellent articles, but also an important exploration of who Abhishek Gupta was, and his legacy.

doi.org/10.5281/zeno...
State of AI Ethics Report 7 Al at the Crossroads: A Practitioner's Guide to Community-Centered Solutions Editors: Renjie Butalid, Connor Wright and Ismael Kherroubi García DOI: 10.5281/zenodo. 17328882
November 4, 2025 at 10:40 AM
hermeneuticist.bsky.social
Feeling hopeful, grateful and somewhat giddy.

Thank you @taniaduarte.bsky.social and @ftollon.bsky.social for the wonderful event last night. Thank you @renjieb.bsky.social for the continued collaborations, and thank you @amysampleward.bsky.social for your relentless leadership.
Ismael happy next to a portrait of Karl Marx
October 10, 2025 at 9:11 AM
hermeneuticist.bsky.social
Thank you, Canada, for the most detailed style guide, but also a case study on how software may shape how language standards evolve (suspiciously in French).
Recommendation Legislative counsel should use dashes (in moderation) when doing so would clarify matters for readers. Analysis The Canadian Style sums up the purpose of dashes rather well: In most of its uses, the em dash is simply a substitute for a colon, semicolon or comma … Dashes give greater emphasis to parenthetic material than do commas or parentheses. If the parenthetic material contains internal punctuation or forms a complete sentence, the commas that might have been used to enclose it should be replaced by dashes or parentheses, depending on the degree of emphasis desired or the strength of the relationship to the rest of the sentence. Parentheses are generally used to enclose material more remote from the main thrust of the sentence, dashes for material more closely related[1]. Although in English the accepted typographic practice is to leave no space between the dash and the words next to it, our current software limitations make it advisable to separate the dash from its neighbours with a space — as, in fact, is the custom in French. In standard English, mid-sentence punctuation is not restricted to commas. Dashes and parentheses both are useful in structuring sentences visually for the reader.
October 4, 2025 at 2:59 PM
hermeneuticist.bsky.social
Business card math
How many cards do you want? 50 cost £37. 100 cost £71. 150 cost £108. 200 cost £103.
September 25, 2025 at 11:17 AM
hermeneuticist.bsky.social
You can find the report in question here: zenodo.org/records/1519...

I will be joining in my capacity as the founder of the Responsible Artificial Intelligence Network (RAIN), whose theory of change calls for an innovation ecosystem inclusive of diverse voices.

drive.google.com/file/d/1yVuw...
What is responsible AI? An AI lifecycle inclusive of diverse stakeholders who, striving for social justice, inform AI narratives and governance.
September 19, 2025 at 8:06 PM
hermeneuticist.bsky.social
Thrilled to be joining @ftollon.bsky.social at RSA House and online, 9th October at 6pm (UTC+1), with @taniaduarte.bsky.social to discuss Tollon’s recent work with @shannonvallor.bsky.social on the #ResponsibleAI landscape.

www.eventbrite.co.uk/e/the-respon...
Poster for the responsible AI landscape event, taking place at RSA House and online on 9th October at 6pm (UTC+1).
September 19, 2025 at 8:06 PM
hermeneuticist.bsky.social
Thinking about the need to disentangle "avoidability" and "inevitability" (easy), but also picture what is deemed "inevitable" when it comes to #AI (hard)

#AGI #ResponsibleAI #FutureofAI #AIHype

open.substack.com/pub/ismaelkh...
Al may be unavoidable but it's not inevitable It's just a question of time
September 15, 2025 at 4:56 PM
hermeneuticist.bsky.social
My company email inbox receives a lot of spam, with many strange subject lines that rarely speak to my company’s values but this one may take the prize.
Email heading: Limited time Tesla deal for Kairoi Ltd!
September 12, 2025 at 10:35 AM
hermeneuticist.bsky.social
“Continuing to intensify, Ismael was upgraded into a hurricane late on August 10 and subsequently developed an eye.”

What the legends say are true.

en.m.wikipedia.org/wiki/Hurrica...
Ismael at peak intensity on August 11
September 8, 2025 at 4:05 PM
hermeneuticist.bsky.social
“Hey Ismael,” they say to me: “hey Ismael, why do you prefer living in the middle of nowhere and running events online over…”
A list of articles and videos on the ongoing tube strike in London
September 8, 2025 at 10:50 AM
hermeneuticist.bsky.social
Do you know anybody suffering from AI fatigue? What symptoms do they show and what treatments do they find work best?

open.substack.com/pub/ismaelkh...
Do you suffer from Al Fatigue? 6 in 10 people are struggling to care for Al and the number is growing, made-up survey finds
September 1, 2025 at 9:07 PM
hermeneuticist.bsky.social
I made a "No AI Guarantee" label to show that no AI was used
Red stamp that says in all-caps: "no nai guarante"
September 1, 2025 at 2:39 PM
hermeneuticist.bsky.social
Preparing for the event I am speaking at in a couple of weeks on #AI and the #future of #work, I am beginning to see a potential link between what #BigTech did for the #GigEconomy, and what is to come...

open.substack.com/pub/ismaelkh...
"This is the world the proliferation of Al tools will bring about: quicker services of lower quality due to regulatory blindspots, and increasingly precarious work to survive in an increasingly expensive world."
August 26, 2025 at 4:03 PM
hermeneuticist.bsky.social
New #event with the Responsible Artificial Intelligence Network (RAIN)! 8th September, 18:30 (UTC+1). We’ll be talking about AI’s impacts on the modern workforce.

#AI #London #Online
Event Announcement: Navigating Al & Modern Career Disruption The first of "The Adaptation Series," with support from the Responsible Artificial Intelligence Network (RAIN)
August 15, 2025 at 12:27 AM
hermeneuticist.bsky.social
Somewhat tangential but I am fascinated by this page (so many maps!) and the fact that Nova Scotia has no area in the "easily accessible" category, which is only found in B.C., Alberta, Manitoba and Ontario, from what I can tell. hillnotes.ca/2025/07/02/m...
Map of Nova Scotia depicting the remoteness index of its subdivisions.
August 10, 2025 at 6:40 PM
hermeneuticist.bsky.social
I wrote about it. I do not miss working in British workplaces... open.substack.com/pub/ismaelkh...
"How am I meant to find it in me to respect you going forwards?" I asked. "What am I meant to do now?"
August 9, 2025 at 9:48 PM
hermeneuticist.bsky.social
If someone helps explain why sharing lived experiences enables empathy, I’d really appreciate it. It feels like it’s just an intuition at this point. ismaelkherroubigarcia.substack.com/p/disentangl... 💡 #Philsky #philosophy
Disentangling Lived Experience, Empathy and Expertise This one took way too long to write...
August 3, 2025 at 11:38 AM
hermeneuticist.bsky.social
Had my age verified for Bluesky the other day. I went down the photo route. The alternative was sharing my credit card for some reason. Anyway, it’s age estimation and not verification, which I find quite funny.
How age estimation works Just take a photo of your face using your phone or other device. We use artificial intelligence to inspect your image and accurately estimate your age. © We won't tell you an exact estimate - we just confirm whether you meet a required age. How accurate is it? As it stands, age estimation guesses the age of 13 to 24 year olds within a 2-year margin of error. This is proven to be much more accurate than human estimates. How many attempts do I get? If your estimated age doesn't meet the required age, you can try age estimation again. Alternatively, if you're given a choice of ways to confirm your age, feel free to use other methods. This technology is powered by identity verification providers, Yoti. Learn more at yoti.com Age estimation is fast, safe and secure Confirm your age in an instant No-one sees, stores or shares your image Your photo is protected with secure encryption You'll be asked to enable camera access
July 30, 2025 at 11:50 AM
hermeneuticist.bsky.social
I think I'm becoming more radical in my views about the social impacts of AI. Tomorrow's post will be more level-headed, I promise.

ismaelkherroubigarcia.substack.com/p/mapping-th...
"working class factions create new battles amongst themselves; some are too excited about Al, some are too afraid of it, some are too focused on the current impacts of Al, some are too focused on speculations about AI..."
July 26, 2025 at 11:12 PM
hermeneuticist.bsky.social
My latest on #AI, the #US and that #EO.

open.substack.com/pub/ismaelkh...
The latest in the US' fight against Responsible Al: The spirit of the Al Moratorium lingers on
July 24, 2025 at 11:57 PM
hermeneuticist.bsky.social
open.substack.com/pub/ismaelkh...
I do not appreciate people using AI chatbot outputs as evidence of anything other than the fact that AI chatbots produce outputs.
July 21, 2025 at 1:59 AM