- Me: "write a small function that does one thing"
- LLM: would you like me to add a calculation for the answer to the ultimate question to life, the universe, and everything?
- Me: [deletes, includes print("42")].
@svanbaal.bsky.social
Lecturer in Decision Making at University of Leeds. Main interests: time preferences, impulsivity, self-control. Incl applications like gambling, consumption/saving, difficult decisions. Check out the behavioural science feed I made, and tag #BehSci.
- Me: "write a small function that does one thing"
- LLM: would you like me to add a calculation for the answer to the ultimate question to life, the universe, and everything?
- Me: [deletes, includes print("42")].
Self-control in daily life is more than willpower β but how well do commonly used trait scales capture this recent, broader view on self-control?
New preprint with @kaihorstmann.bsky.social & @mhennecke.bsky.social: osf.io/preprints/ps...
Short summary below.
Great to see this work! Also I love this angle on how impulsivity and self-control come apart (and are not opposite ends of a spectrum).
28.07.2025 15:57 β π 1 π 0 π¬ 0 π 0So true. Good to avoid unnecessary suffering!
25.07.2025 16:05 β π 3 π 1 π¬ 0 π 0I'm happy to share our latest manuscript "Impediments to countering racist pseudoscience", coauthored with @gillianrbrown1.bsky.social @kztwyman.bsky.social & Marcus Feldman.
25.07.2025 09:17 β π 70 π 30 π¬ 1 π 2@madselk.bsky.social, @benansell.bsky.social, Laure Bokobza, @aslicansunar.bsky.social, @mhaslberger.bsky.social, and @jacobnyrup.bsky.social. 2025. βWhy Is It so Hard to Counteract Wealth Inequality? Evidence from the United Kingdomβ. World Politics. 77(3): 515β61. muse.jhu.edu/article/964464
14.07.2025 08:51 β π 28 π 4 π¬ 1 π 1100% agreed. So frustrating. Sometimes Iβm unsure whether itβs just rhetoric or politicians truly donβt understand the ramifications.
14.07.2025 09:47 β π 1 π 0 π¬ 0 π 0We need this badly. Unfortunately, there's also too little attention to cold-related deaths. Way more people die of the cold in Europe, and that nearly never makes the news. Though that balance will probably flip in a few decades.
14.07.2025 09:08 β π 3 π 0 π¬ 1 π 0Shows death rates in US and eu cities by temperature, air con being the decisive factor
Sitting here bleary after another poor night's sleep, disinclined to engage in a day of cognitive effort, this chart from @jburnmurdoch.ft.com really strikes home on.ft.com/4eFZG4l
11.07.2025 06:09 β π 369 π 102 π¬ 30 π 26Great PSA! To echo: please don't use the chat window (your research decisions are being made for you). Send me a message and I will happily provide you with code to prompt OpenAI's model endpoints and store the conversational context throughout.
08.07.2025 08:58 β π 2 π 0 π¬ 0 π 0WIRED loves Fairphone and everything it stands forβbut people just arenβt buying its devices, and the few who have donβt need to upgrade. www.wired.com/story/fairph...
29.06.2025 11:04 β π 163 π 30 π¬ 19 π 11Great article linked below. I was unaware journals could negotiate and inflate the metric by excluding article types. I suppose that is part of the reason why there are always a bunch of different article types. Yet another example of Goodhart's Law!
journals.plos.org/plosmedicine...
Registration is open for the Summer School on Computational Decision Sciences, 8β12 Sept
Join us in London for a week of cutting-edge research on decision-making, AI, cognitive science and more!
(free, register by 11Aug)
center-decision-sciences.com/cds-summer-s...
#PsychSciSky π§ #neuroskyenceπ§ͺ
Chatbots β LLMs β do not know facts and are not designed to be able to accurately answer factual questions. They are designed to find and mimic patterns of words, probabilistically. When theyβre βrightβ itβs because correct things are often written down, so those patterns are frequent. Thatβs all.
19.06.2025 11:21 β π 29332 π 8578 π¬ 576 π 700@lukaszwalasek.bsky.social
17.06.2025 14:29 β π 0 π 0 π¬ 0 π 0Love to see it. Thoughtful piece too :)
17.06.2025 13:23 β π 1 π 0 π¬ 0 π 0I love a good traffic violation (especially if you don't operate heavy machinery for your travel), so I will grab the popcorn for this conversation. Imo if many pedestrians are crossing where/when they "shouldn't", it's a design error - they're deprioritised for automobile traffic flow.
17.06.2025 10:55 β π 2 π 0 π¬ 1 π 0Added it to the list! Re your point: not to mention the history of the term jaywalking!
17.06.2025 10:47 β π 1 π 0 π¬ 0 π 0I can hear a faint echo of: βRotgΓ€nger, totgΓ€nger!!!β
17.06.2025 10:40 β π 1 π 0 π¬ 1 π 0First, thatβs so valid. Second, sorry but I have to say thatβs the most German complaint you could make π Iβm so here for it!!
17.06.2025 10:39 β π 2 π 0 π¬ 1 π 0Thatβs definitely part of it. Another point is that we get a lot of negative press about gun violence, lack of public transportation, expensive healthcare etc. From the outside, itβs difficult to see how localised some of these things are.
17.06.2025 10:26 β π 1 π 0 π¬ 2 π 0Oh absolutely, arguably even more so in the US! Especially in the North East and the West Coast. But the US market has a finite supply for faculty positionsβ¦ Also probation is usually less intense than tenure, so that could contribute!
17.06.2025 10:23 β π 1 π 0 π¬ 1 π 0Great point!
17.06.2025 09:48 β π 1 π 0 π¬ 1 π 0Though US & CA pay much better, the pay in the UK is commensurate with many EU countries, so itβs a straightforward option if you wish to stay within reach of family and friends.
17.06.2025 09:22 β π 1 π 0 π¬ 1 π 0Also the higher education market in the UK has always benefited from a prestige privilege, inflating funding for faculty positions via intl student fees. This creates network effects, making it more attractive to work here due to big departments w more colleagues in oneβs own field
17.06.2025 09:19 β π 2 π 0 π¬ 2 π 0Thanks!!
17.06.2025 09:07 β π 0 π 0 π¬ 0 π 0#CDRLeeds 's Simon Baal @svanbaal.bsky.social and colleagues have received an AFSG Minor Exploratory Grant (afsg.org) to improve use of the Problem Gambling Severity Index. They'll attempt identifying different problematic gambling profiles using the score and natural language processing tools.
16.06.2025 17:22 β π 5 π 1 π¬ 1 π 0New work by Sudeep Bhatia @sdpbht.bsky.social, Simon van Baal @svanbaal.bsky.social, Feiyi Wang, and @lukaszwalasek.bsky.social Lukasz Walasek -- now out in Proceedings of the National Academy of Sciences!
www.pnas.org/doi/10.1073/...
Would you like to extract and exploit textual data from the internet? No matter whether you want to analyse product reviews, or political opinions, #CDRLeeds 4-day workshop Webtext2Insight will teach you how to do it. 9-12 June 2025 in Leeds. More info and registration: shorturl.at/JCiKT
20.05.2025 17:43 β π 0 π 2 π¬ 0 π 0bsky.app/profile/svan...
07.05.2025 11:45 β π 0 π 0 π¬ 0 π 0