Hey! There's finally someone else in Australia doing research in multi-objective reinforcement learning. @marcusgal.bsky.social arxiv.org/pdf/2509.14816

Almost 7 months ago I reported a book chapter to Springer's Ethics reporting team. Springer have yet to take any action (they are 'still investigating'). That chapter now has 450 downloads and 2 citations. Delays in dealing with these issues allow further pollution of the scientific literature.

I'm hiring again! Please share. I'm recruiting a postdoc research fellow in human-centred AI for scalable decision support. Join us to investigate how to balance scalability and human control in medical decision support. Closing date: 4 October (AEST). uqtmiller.github.io/recruitment/

AI-SearchPlanner: Modular Agentic Search via Pareto-Optimal Multi-Objective Reinforcement Learning Huawei proposes an RL framework that decouples search planning from answer generation, using dual-reward alignment and Pareto optimization. 📝 arxiv.org/abs/2508.20368

Multi-objective reinforcement learning (MORL) offers a more robust and adaptable solution by optimizing for a vector of rewards—such as fairness, diversity, and ethical norms. The resulting behaviors support transparency, explainability, and human alignment.

Imagine a future where all of the world's scientific knowledge has been captured and organized in a way that (a) permits tracing back to the supporting data and reasoning and (b) identifying situations in which the knowledge could be applied. 2/

Anyone who plans to submit for an ARC grant should be giving feedback via this survey.

Academic authors, here's a peek into the black box of journal publishing from an journal editor if you can bear it:

My petition to the 🇦🇺 Australian government: make part-time PhD students' stipends tax exempt! 📋 Read and sign here: www.aph.gov.au/e-petitions/... ⏰ Deadline: October 1

The Sony AI Game AI team has reinforcement learning internships opens for 2026! It is remote for people in the US & Canada, mixed remote/onsite in Europe (onsite in Zurich), and onsite in Tokyo. If you want to work on RL with cool applications, sign up! ai.sony/joinus/job-r...

1/8 🧵 GPT-5's storytelling problems reveal a deeper AI safety issue. I've been testing its creative writing capabilities, and the results are concerning - not just for literature, but for AI development more broadly. 🚨

Is ResearchGate's user base dying? Reads and Full-text Reads of my papers peaked back in 2021, and have fallen by about 50% since then. Citations of my work have continued to rise, so I don't think this is caused by people losing interest in my work :-)

Wouldn't it be nice if the ARC assessed proposals in a similar way to the NIH? Quoted from www.the-scientist.com/an-nih-revie... "The last factor evaluates the principal investigator (PI) or multiple PI (MPI) team, co-investigators, consultants, and institutions involved...

This relates to my previous post, but I wanted to keep it separate to avoid interfering with any discussion of the important content of Kewen's paper. On a personal note, I'm delighted that we managed to slip this footnote in, explaining why our algorithm is called MOSS.

Are you having trouble applying multi-objective reinforcement learning to a stochastic environment? Do you want to know why? Kewen Ding's new paper provides insights (available via open-access here: doi.org/10.1017/S026...).

New in @theconversation.com today from @suzesrdarov.bsky.social and I: ‘Australiana’ images made by AI are racist and full of tired cliches, new study shows theconversation.com/australiana-... [1/2]

🚀 Introducing EngiBench & EngiOpt A unified benchmark and algorithm suite for ML-driven engineering design. 💡 Imagine: • A diffusion model for designing airplane wings • Surrogate models predicting DcGain & Voltage Ripple in power electronics 🧵⬇️

New blog post: AI Allergy. On my increasing disgust with the AI discourse, even though I still like the technical and philosophical. And how I wish I could be excited about AI again. togelius.blogspot.com/2025/08/ai-a...

Australian-based researchers: just a reminder that before 19 Aug, you should set up multi-factor authentication for the RMS: www.arc.gov.au/news-publica... 1. Log in to RMS. 2. Go to ‘Personal Details’ in Action Centre. 3. Click ‘MFA Settings’. Follow the prompts. Took me 60 seconds.

I think I could handle the whole "unpaid labour" aspect of peer reviewing much more easily if it didn't also involve dealing with the never-ending inadequacy of Editorial Manager.

New from me: When Australian AI expert Dr Kobi Leins declined a medical specialist’s request to use AI transcription software during her child’s upcoming appointment, the practice told her the practitioner required the technology be used and Leins was "welcome to seek an alternative" provider. 👇

Dear journal copy editors. Yes, I am the corresponding author, but I am not the first author. So why have you listed "Vamplew et al." in the running header of the proof you just sent me? 🤦‍♂️

Today, our article "The entities enabling scientific fraud at scale are large, resilient, and growing rapidly" is finally published in PNAS. I hope that it proves to be a wake-up-call for the whole scientific community. reeserichardson.blog/2025/08/04/a...

At last I have an opportunity to engage in some Schmidhubering 😊 I was excited to recently discover the paper Expected Eligibility Traces by Hado van Hasselt et al ojs.aaai.org/index.php/AA... (AFAIK none of these authors are on bsky yet). 🧵 (a short one!)

Imagine if all the money academics pour into OA fees at for-profit publishing corporations went instead to academic societies, which exist to support science; or to university presses, which exist to support academia

HOW DO I LOG IN TO EDITORIAL MANAGER TO SUBMIT MY REVIEW

Seeing I shut down my X account, I might repost some of my more interesting (at least to me!) previous threads here. First up: Pete's Odd Adventures in Academic Publishing - Episode 3: The Phantom Author. This is going to be a lengthy post, but bear with me - the bizarreness merits full details.

With Tom Lehrer's passing, I suppose this is a moment to share the story of the prank he played on the National Security Agency, and how it went undiscovered for nearly 60 years.

At last we've discovered AI's true purpose: suno.com/song/f3a174d...

I reported this to Springer 5 months ago, but apparently they are "still investigating". How long does it take to confirm that a paper which doesn't exist anywhere on the Internet isn't a genuine reference (and neither are multiple other references from this book chapter).

2 weeks later, and we're *still* fixing issues in our paper introduced by the copy editors. I'm starting a petition to have "copy editors" added alongside the telephone sanitisers, account executives, and management consultants on Golgafrinchan Ark Fleet Ship B.

Please be true! www.avclub.com/taika-waitit...

AI is reaching new levels of intrusive annoyingness. I was just interrupted by this Copilot notification. I'm not preparing a presentation, or doing anything vaguely related to a presentation. At least Clippy only thought you might be writing a letter when you were actually writing *something*.

A rare sighting - a wild cellular automaton in it's natural habitat.

A Borges story about a guy who gets AI to summarize all the world’s information for him, and then summarize the summary, until the AI has the whole world summarized into a single word. He sits alone at his desk, staring at the word, repeating it endlessly, certain he is experiencing everything

Today's installment of Peer Review is Broken. I just received feedback on a review I did for a reputable (Q2) journal, including the review submitted by the only other reviewer. It totals <100 words with only one meaningful comment, as paraphrased below. 1/n

Can I get away with using this as my author photo in IEEE publications from now on?

If journal copy editors were house cleaners, they'd give the place a light dusting, swap all your light-bulbs to a slightly different wattage, turn your dining table upside-down, and move your bed into the garage.

OMG. I've been waiting 40 years for this!

I wouldn’t expect a quick answer from Springer. I reported a similar case of a book chapter with hallucinated references, including one attributed to me which doesn’t match anything I’ve actually written. It’s been 4 months now and I’m still waiting for their investigation to reach a conclusion.

Hiring a postdoc to scale up and deploy RL-based planning onto some self-driving cars! We'll be building on arxiv.org/abs/2502.03349 and learn what the limits and challenges of RL planning are. Shoot me a message if interested and help spread the word please! Full posting to come in a bit.

While welcome to research teams facing last minute application finalisation, was it really not possible to announce the submission deadline extension for Linkage Project 2025 Round 2 earlier than one week prior to the original deadline? A lot of stress could have been avoided.

Hey @dredgedragline.bsky.social, I'm guessing there's zero overlap with your list? (I never really moved on from the 90s, did I?)

I hear that the RLDM opening talk by Cate Hartley and Michael Littman was great. Wish I could have been there, but in lieu of attending I've made a small improvement to one of their slides! 😀

Celebrating some citation milestones this week! Both of our multi-objective reinforcement learning collaborations from 2022 reached significant (ok, round) numbers with Scalar Reward is Not Enough reaching triple-digits, and the Practical Guide to MORL halfway to 1000.

I have a new theory. The developers of moodle and Workday also own a highly profitable chain of smash rooms.

ARC’s Linkage Projects delay is already having huge knock-on effects. This researcher’s work with their industry partner (remote community groups) will be delayed by a year!👇

“Anonymous GitHub” — I just discovered this today. What a great idea!! (Apparently I am the 46,631ˢᵗ to find it.) Provides an anonymized portal to your GitHub respiratory suitable for submission to double-blind review for research conference papers or journal articles. anonymous.4open.science