iiinova — where your digital self keeps evolving

Raphaël Merx @rapha.dev · 14/03/2026

—> it’s paradigm replacement, not task automation, that causes widespread job displacement from substack.com/home/post/p-...

substack.com

Why ATMs didn’t kill bank teller jobs, but the iPhone did

There's a lot more to replacing labor than just automating tasks

000

Raphaël Merx @rapha.dev · 14/03/2026

cool analogy for AI & jobs: - when ATMs came, the number of bank tellers rose, bc ATMs lowered the cost of running bank branches, so more branches opened - but in the 2010s, the number of bank tellers plummetted, bc mobile banking made branches unnecessary

100

Raphaël Merx @rapha.dev · 14/03/2026

a *great* chart to teach confounding variables

000

Raphaël Merx @rapha.dev · 30/10/2025

This is some legit really impressive work!!

031

Reposted by Raphaël Merx

Multilingual Representation Workshop @ EMNLP 2026 @mrl-workshop.bsky.social · 29/10/2025

Introducing Global PIQA, a new multilingual benchmark for 100+ languages. This benchmark is the outcome of this year’s MRL shared task, in collaboration with 300+ researchers from 65 countries. This dataset evaluates physical commonsense reasoning in culturally relevant contexts.

12210

Raphaël Merx @rapha.dev · 18/10/2025

the paper www2.statmt.org/wmt25/pdf/20...

www2.statmt.org

000

Raphaël Merx @rapha.dev · 18/10/2025

They say it's because (1) test sets have become more challenging, (2) include more lang pairs, (3) are longer, and (4) used ESA instead of MQM. But we need an ablation study!

100

Raphaël Merx @rapha.dev · 18/10/2025

Whoa the #WMT25 results on MT Evaluation are wild! ChrF outperforms pretty much all neural metrics 🙀

100

Raphaël Merx @rapha.dev · 06/10/2025

kudos to whoever came up with that paper name 👌

010

Raphaël Merx @rapha.dev · 27/07/2025

in Vienna for ACL, presenting Tulun, a system for low-resource in-domain translation, using LLMs Tuesday @ 4pm Working w 2 real use cases: medical translation into Tetun 🇹🇱 & disaster relief speech translation in Bislama 🇻🇺

131

Raphaël Merx @rapha.dev · 08/06/2025

Cool paper, at the intersection of grammar and LLM interpretability. I like that they use linguistic datasets for their experiments, then get results that can contribute to linguistics as a field too! (on structural priming vs L1/L2)

010

Raphaël Merx @rapha.dev · 26/05/2025

Thanks a lot! I didn't make it to Albuquerque unfortunately, but I hope to be in Vienna for ACL. Might see you there?

100

Raphaël Merx @rapha.dev · 25/05/2025

(3) The vast majority of usage is on mobile (over 90% of users / over 80k devices) Takeaway: publishing MT model in mobile apps is probably more impactful than setting up a website / HuggingFace space.

120

Raphaël Merx @rapha.dev · 25/05/2025

(2) Translation into Tetun is in higher demand (by >2x) than translation from Tetun Takeaway for us MT folks: focus on translation into low-res langs, harder but more impactful

100

Raphaël Merx @rapha.dev · 25/05/2025

We find that (1) a LOT of usage is for educational purposes (>50% of translated text) --> contrasts sharply with Tetun corpora (e.g. MADLAD), dominated by news & religion. Takeaway: don't evaluate MT on overrepresented domains (e.g. religion)! You risk misrepresenting end-user exp.

100

Raphaël Merx @rapha.dev · 25/05/2025

Our paper on who uses tetun.org, and what for, got published at the LoResMT 2025 workshop! An emotional paper for me, going back to the project that got me into a machine learning PhD in the first place.

230

Raphaël Merx @rapha.dev · 13/05/2025

Very interesting findings, particularly the benefit (or lack thereof) of test-time scaling across domains

000

Raphaël Merx @rapha.dev · 08/05/2025

My favourite ICLR paper so far. Methodology, findings and their implications are all very cool. In particular Fig. 2 + this discussion point:

041

Raphaël Merx @rapha.dev · 02/05/2025

Incredible paper, finding that large companies can game the LMArena through statistical noise (via many model submissions), over-sampling of their models, and overfitting to Arena-style prompts (without real gains on model reasoning) The experiments they run to show this are pretty cool too!

040

Raphaël Merx @rapha.dev · 23/04/2025

Cool summary of issues with multilingual LLM eval, and potential solutions! If you're doubtful of all these non-reproducible evals on translated multiple choice questions, this paper is for you

120

Reposted by Raphaël Merx

PyCon AU @pyconau.bsky.social · 30/03/2025

👋 Hey Bluesky! We’ve just touched down and we’re excited to be here 🌤️🐍 This is the official PyCon AU account, your go-to space for updates, announcements, and all things Python in Australia✨ Hit that follow button and stay tuned because we’ve got some awesome things coming your way! #PyConAU

PyConAU We are on BlueSky! Follow us and stay tuned! @pyconau.bsky.social

067

Raphaël Merx @rapha.dev · 31/03/2025

AI dev tools. In particular agents: are they hype or useful or both?

000

Raphaël Merx @rapha.dev · 17/03/2025

The right thing to do, thanks for this *SEM

020

Raphaël Merx @rapha.dev · 20/02/2025

Super impactful, thank you for this! A natural sequel of Gatitos. I'm esp. fond of your "researcher in the loop" method to ensure wide vocab coverage.

010

Reposted by Raphaël Merx

iseeaswell.bsky.social @iseeaswell.bsky.social · 19/02/2025

😼SMOL DATA ALERT! 😼Anouncing SMOL, a professionally-translated dataset for 115 very low-resource languages! Paper: arxiv.org/pdf/2502.12301 Huggingface: huggingface.co/datasets/goo...

2148

Reposted by Raphaël Merx

Kris 🎃 Lorischild @direkris.itch.io · 15/01/2025

Been hearing a lot about recency bias lately. Must be pretty important

011927

Raphaël Merx @rapha.dev · 17/02/2025

Such a well put together video! Gherkins in the background got a supporting role

110

Raphaël Merx @rapha.dev · 24/01/2025

Congrats! I'm just getting started but really liked your papers. Cool, impactful and well-written

010

Raphaël Merx @rapha.dev · 05/12/2024

Our paper on generating bilingual example sentences with LLMs got best paper award @ ALTA in Canberra! arxiv.org/abs/2410.03182 We work with French / Indonesian / Tetun, find that annotators don't agree about what's a "good example", but that LLMs can align with a specific annotator.

010

Reposted by Raphaël Merx

Irene Chen @irenetrampoline.bsky.social · 15/11/2024

What do it mean to be a “low resourced” language? I’ve seen definitions for less training data to low number of speakers. Great to see this important clarifying work at #EMNLP2024 from @hellinanigatu.bsky.social et al aclanthology.org/2024.emnlp-m...

0243

Raphaël Merx @rapha.dev · 20/11/2024

this guy lives rent free in my hippocampus

000

Raphaël Merx @rapha.dev · 20/11/2024

Is productionisation (and move to gimmicks like CoT in o1-preview) at OpenAI and Anthropic a sign that scaling laws are slowing? And if so, where are we headed in LLMs? Slightly pretentious but enjoyable read: www.generalist.com/briefing/the...

generalist.com

The Bitter Religion: AI’s Holy War Over Scaling Laws | The Generalist

The AI community is locked in a doctrinal battle about its future and whether sufficient scale will create God.

030