If you've been feeling tired recently, it's because Earth is spinning faster, making some days ~1 millisecond shorter than 24 hours so you're getting less sleep. 😴
23.07.2025 10:29 — 👍 189 🔁 29 💬 11 📌 6@rayo.dev.bsky.social
PhD Designing Responsible NLP at the University of Edinburgh, on AI literacy through co-design with young people Trustee of Prewired ex-CTO of Voxsio he/they 🇳🇱🏴🏳️🌈 https://rayo.dev
If you've been feeling tired recently, it's because Earth is spinning faster, making some days ~1 millisecond shorter than 24 hours so you're getting less sleep. 😴
23.07.2025 10:29 — 👍 189 🔁 29 💬 11 📌 6i genuinely think that part of the problem is people don’t have the mental model for how the “chat bubbles” map to autocomplete
it doesn’t help that the chat interfaces actively hide what’s going on and make the illusion opaque
Rayo standing in front of a projector screen with a slide that says “Children as Creators of Real-World AI-Powered Tools for Education”
Yay I’m at @acm-idc.bsky.social in Reykjavík this week! I spoke about my PhD plans at the Doctoral Consortium and am so excited to get to know everyone this week :)
23.06.2025 13:37 — 👍 3 🔁 0 💬 0 📌 0It was absolutely wonderful to work with the Digital Leaders at St Francis Xavier’s Primary School. Thanks again for having me!
17.06.2025 18:30 — 👍 1 🔁 0 💬 0 📌 0drawing from a child with bold, colourful text that ready 'AI should respect our privacy'
📣 “𝗖𝗵𝗶𝗹𝗱𝗿𝗲𝗻 𝗻𝗲𝗲𝗱 𝘁𝗼 𝗸𝗻𝗼𝘄.”
That’s the message we heard loud and clear from Members of the Children’s Parliament.
𝗠𝗮𝗻𝘆 𝗰𝗵𝗶𝗹𝗱𝗿𝗲𝗻 𝗵𝗮𝘃𝗲 𝘃𝗼𝗶𝗰𝗲𝗱 𝗰𝗼𝗻𝗰𝗲𝗿𝗻𝘀 𝗮𝗯𝗼𝘂𝘁 𝗵𝗼𝘄 𝗔𝗜 𝗶𝘀 𝗮𝗳𝗳𝗲𝗰𝘁𝗶𝗻𝗴 𝘁𝗵𝗲𝗶𝗿 𝘀𝗮𝗳𝗲𝘁𝘆, 𝗽𝗿𝗶𝘃𝗮𝗰𝘆, 𝗮𝗻𝗱 𝗿𝗶𝗴𝗵𝘁𝘀. One key worry? They don’t know what’s happening with their data and they want answers.
The petition to ban conversion therapy practices in the EU is lacking enough votes from:
- #germany (49.9%)
- #sweden (49.96%)
- #Netherlands (94%)
- #austria (25.5%)
- #Slovenia (83.14%)
- #portugal (24.46%)
If you're a citizen there you can help advance a ban on conversion therapy by voting:
Thankfully people have realised that wasn’t enough and actual programming is part of the curriculum again. Would be great if we could not make the same mistakes again when it comes to AI!
12.04.2025 09:51 — 👍 4 🔁 0 💬 0 📌 0There was a similar push in the 2000s where computing lessons in schools shifted from teaching basic programming to teaching MS Office as that was deemed to be more important. And instead of teaching something fundamental about how computers work it became about prepping kids for office jobs
12.04.2025 09:51 — 👍 3 🔁 0 💬 2 📌 0Co-opting the term literacy for corporate AI lessons is particularly devious because it’s hard to argue against. Everyone thinks literacy is important! But in the chaos of schools trying to adapt it’s easy for the actual meanings of these words to get confused
12.04.2025 09:51 — 👍 3 🔁 0 💬 1 📌 0If literacy is the ability to effectively understand and interpret text, to me, AI literacy is understanding the limits of AI output, knowing why LLMs can hallucinate, being aware of the risks of model biases, etc etc. Not “how to use AI to offload your thinking”
12.04.2025 09:51 — 👍 5 🔁 0 💬 1 📌 0I get so frustrated with the dilution of the term ‘literacy’ - in no world should ‘AI literacy’ mean ‘prompt engineering classes’ but apparently here we are
12.04.2025 09:51 — 👍 21 🔁 2 💬 2 📌 0Join us at AI for All during Edinburgh Science Festival! Discover how we make #NLP shape our daily lives in a fair, accountable and transparent way. Explore exhibits, activities, and meet the next generation of researchers.
📅 Fri 11 Apr
⏰ 10am – 4pm|Free / Drop-In
📍@edfuturesinstitute.bsky.social
Okay enough ranting for tonight. What? You thought I was about to offer that solution? Haha very funny vier april kikker in je bil
04.04.2025 22:54 — 👍 0 🔁 0 💬 0 📌 0Look this is hardly the biggest problem in the world but every day I see so many people write about the dangers of genAI in education and like I hear you and probably agree with you 100% but at a certain point you gotta offer an actual alternative solution
04.04.2025 22:54 — 👍 0 🔁 0 💬 1 📌 0For example, CodeCademy makes it super difficult nowadays to get to their free courses. And on the other side the new RPF code editor doesn’t let you add external links (anchor tags) to HTML files because “safety”
04.04.2025 22:54 — 👍 0 🔁 0 💬 0 📌 0They are genuinely brilliant, got everything they could out of Scratch and are hungry to learn more complex stuff. But intro to coding resources are either aimed at kids and super restrictive/linear or aimed at adult learners and, well, also restrictive plus you need a credit card
04.04.2025 22:54 — 👍 0 🔁 0 💬 1 📌 0This is so cool. Have been wanting to get into PoE properly for so long now and I think this might actually do it
26.03.2025 21:56 — 👍 1 🔁 0 💬 0 📌 0A really fun reflection, I thought, not just on AI but on the purpose and origins of language in general
24.03.2025 20:49 — 👍 0 🔁 0 💬 0 📌 0“We are not in fact in danger of the machines overtaking us in what we do best. What we do best is to mess with mom, and to whisper sweet nothings to our lovers. But we are in grave danger, at present, of misidentifying what we do best”
24.03.2025 20:49 — 👍 2 🔁 0 💬 1 📌 0There are *good*, positive use of GenAI. But forcing it on people -- particularly as a massive technology corporation with its fingers in many aspects of billions of peoples' lives -- creates more risk than potential benefit. Let people opt-in as they see fit, don't force it down their throat.
27.01.2025 20:16 — 👍 97 🔁 6 💬 4 📌 2The image is from a "Transparency Center" document and lists guidelines regarding acceptable and prohibited content for insults. It mentions: 1. Insults about: Character, such as cowardice, dishonesty, criminality, and sexual promiscuity or immorality. Mental characteristics, including but not limited to accusations of stupidity, intellectual capacity, and mental illness, as well as unsupported comparisons among politically correct (PC) groups based on inherent intellectual traits. 2. Highlighted section: The document allows allegations of mental illness or abnormality when tied to gender or sexual orientation, referencing political and religious discourse about transgenderism and homosexuality. It also acknowledges the non-serious use of terms like "weird."
Meta literally created a LGBTQ exception for calling someone mentally ill as an insult. You can't do it for any other group except LGBTQ people.
08.01.2025 01:51 — 👍 15535 🔁 6400 💬 662 📌 1625They must've calculated the optimal way to space out these words. Like, at the exact spots people are most likely to stop using the app they drop one of these to lull users into a false sense of "oh, I guess I *can* learn this language!"
07.01.2025 08:59 — 👍 1 🔁 0 💬 1 📌 0It’s cool that humans have been eating bread so long that when we touch a loaf for a moment we can sense ah this is going to be awful bread. Like badgers sensing vibrations underground through their claws.
07.01.2025 04:07 — 👍 10567 🔁 1096 💬 172 📌 43Beat BG3 in Honour Mode last night and honestly this might be where 2025 peaks for me
02.01.2025 11:15 — 👍 1 🔁 0 💬 0 📌 0👀
18.12.2024 16:42 — 👍 1 🔁 0 💬 0 📌 0OpenAI and tech bros, take notes
(this slide is worth the intercontinental trip to #NeurIPS24)
Hence "down the line"... Am curious to see if/how it evolves
01.12.2024 21:08 — 👍 0 🔁 0 💬 0 📌 0Just updated my handle to @rayo.dev and this bring-your-own-domain thing might be my favourite part of Bluesky. Feels like an elegant (if not entirely complete) way to do verification as well as enable more decentralisation down the line without the up-front complexity of Mastodon
01.12.2024 19:14 — 👍 3 🔁 0 💬 1 📌 0💙💙
27.11.2024 13:14 — 👍 1 🔁 0 💬 0 📌 0Hi just found this, that's amazing! Would you be able to add me? 😊
27.11.2024 13:06 — 👍 1 🔁 0 💬 1 📌 0