I aim to confuse
27.11.2025 17:31 β π 3 π 0 π¬ 0 π 0@maosbot.bsky.social
Parent, spouse, Australian, Professor of Machine Learning in Oxford. Long Covid, trans rights, music, reggae on Fridays, AI must be good for humans, https://www.robots.ox.ac.uk/~mosb
I aim to confuse
27.11.2025 17:31 β π 3 π 0 π¬ 0 π 0It is funny
27.11.2025 13:12 β π 3 π 0 π¬ 0 π 09YO had not heard of Cash and Nico (but is watching Minecraft videos on YouTube as I write): we will check out with thanks for the tip!
25.11.2025 17:30 β π 1 π 0 π¬ 0 π 0FNF is also big in our household (and is also stupidly hard)!
23.11.2025 05:26 β π 1 π 0 π¬ 0 π 0Haha our kids are clearly well-aligned! It was indeed EVW 9YO was telling me about, and he sings the gangster rap non-stop (and plays the level). Very cool that your son is creating himself, I will share with 9YO :)
23.11.2025 05:24 β π 1 π 0 π¬ 1 π 0With my ageing fine-motor skills and slowing reflexes
22.11.2025 17:17 β π 1 π 0 π¬ 2 π 0I have played it, and am AWFUL
22.11.2025 17:17 β π 1 π 0 π¬ 1 π 09YO was just telling me about a Geometry Dash player who βis really good, even though heβs 39 years oldβ. I am 44. I will never recover from this devastating blow
22.11.2025 16:25 β π 46 π 0 π¬ 7 π 0Epilogue: I got another email like this today, and then I looked up the paper and discovered that it was real, I had co-authored it and then forgotten about it
18.11.2025 15:18 β π 16 π 1 π¬ 0 π 0Pretty bold to include a hallucinated citation putatively written by the very person to whom you are applying
12.11.2025 14:28 β π 21 π 0 π¬ 3 π 0Your work on Gaussian processes for environmental modeling (βActive Learning of Environmental Models,β Osborne et al., 2012) and multi-objective optimization for sustainability trade-offs (βThe Automation of Science,β King et al., 2009, with your contributions) directly informs my methodological approach. I propose: 1. Bayesian optimization to identify high-impact monitoring
My work is so influential that prospective PhD students are now being influenced by papers that an LLM thinks I wrote but actually did not
12.11.2025 14:26 β π 24 π 2 π¬ 1 π 1Youβd think so, but to my knowledge this question has only been addressed in lab settings e.g. Lindsley et al. (2021),
pmc.ncbi.nlm.nih.gov/articles/PMC..., Patel et al. (2016)
pmc.ncbi.nlm.nih.gov/articles/PMC....
Source control is more effective than receiver control (intuitively)
Thanks David! One-way masking in a FFP3 seems to be working for me, but I have other layers of protection in addition (regular vaccination, clean air)
04.11.2025 15:21 β π 7 π 1 π¬ 1 π 0Came here to say this! Chris Foss is a legend chrisfossart.com
04.11.2025 15:18 β π 3 π 0 π¬ 1 π 0Havenβt been posting much, but I would like to confirm that I am still masking and still havenβt been reinfected with covid (funny how that works)
04.11.2025 15:16 β π 172 π 8 π¬ 6 π 3None of us play, but we would like to buy a piano keyboard for our kids: shouldnβt be top-of-the-line, but shouldnβt be junk, either. Any tips appreciated!
26.10.2025 12:34 β π 7 π 0 π¬ 7 π 0a lovely coastal scene with a red and white striped lifeguard station on a sandy beach. The foreground shows marram grass in the dunes, whilst offshore wind turbines are visible on the horizon. The dramatic cloudy sky adds atmosphere to the composition.
Copenhagen, 2024
07.10.2025 11:23 β π 30 π 1 π¬ 1 π 0π¨New AI Safety Course
@aims_oxford
!
Iβm thrilled to launch a new called AI Safety & Alignment (AISAA) course on the foundations & frontier research of making advanced AI systems safe and aligned at
@UniofOxford
what to expect π
robots.ox.ac.uk/~fazl/aisaa/
Thank you! I wasnβt sure about your post and appreciate the explicit confirmation!
06.10.2025 11:47 β π 2 π 0 π¬ 0 π 0I donβt love the description of increasing autism diagnoses as an βepidemicβ
05.10.2025 17:41 β π 0 π 0 π¬ 1 π 0As evidence for your theory, staff in shops often initiate conversations with me in English (unlike other customers), and the only reason i can imagine for why they do so is my mask
04.10.2025 17:16 β π 6 π 0 π¬ 1 π 0I am happy to defer to your judgement hereβin addition to my visible privileges (gender, height), Iβve only lived in Aarhus, and only for a few months, which is unlikely to be representative
04.10.2025 13:58 β π 2 π 0 π¬ 1 π 0I am pretty sure that @brebis.bsky.social also lives in Denmark. My own 2p: no one else masks here, but I havenβt gotten any grief over my mask, either
04.10.2025 10:46 β π 8 π 0 π¬ 0 π 0Apes smart enough to figure out the various disciplines of complex science explaining how deeply they are screwing up the thin eggshell of biosphere that is their only plausible home, yet too dumb to do anything about it
04.10.2025 10:22 β π 54 π 6 π¬ 0 π 0I just got my Covid booster, and you should, too!
24.09.2025 19:00 β π 60 π 1 π¬ 4 π 0Multilayer pony
My little perceptron
Covid is still out there, and masks still work
20.09.2025 17:14 β π 217 π 38 π¬ 6 π 121575 (crazy!)
20.09.2025 05:53 β π 1 π 0 π¬ 1 π 05,290 papers will be published at NeurIPS this year. If you could still get the proceedings in a book, it would be like 50,000 pages long. Thatβs 14.5 papers a day to read each and every day until the next NeurIPS
18.09.2025 19:06 β π 38 π 9 π¬ 2 π 4Danes keep assuming I am foreign because I am wearing a mask and TBF theyβre correct
12.09.2025 16:57 β π 28 π 0 π¬ 2 π 0