Well there ARE worse places to drive (Georgia, Nepal, China) but NZ is certainly not among them ๐
Well there ARE worse places to drive (Georgia, Nepal, China) but NZ is certainly not among them ๐
Oh no, thatโs wayyy too exciting for NZ. No, the headline was โPolice: Teach tourists road rulesโ, and we were there for illustration. Got interviewed while trying to sell the car we used for our 2.5 months trip
My wife and I were on the front page of a newspaperโฆ in New Zealand
ืืคืืืช ืฉืืืช ืืืชื ืืฉืื ืืืื ืืจื ื?
ืื ืืืจื ืืืืืื ืฉืื ืืงืจืืชื ืืช ืื
@marcdebus.bsky.social and I (2024) echo a claim that building coalitions has become increasingly difficult due to party system fragmentation across western democracies. I wanted to visualise that. Again, Codex to the rescue:
Link: doi.org/10.1016/j.el...
@fortunato.bsky.social has a paper (2019) that greatly influenced my thinking on political compromise. I wanted to show his operationalisation of compromise graphically. Claude gave me this in ~5 minutes:
Link: doi.org/10.1017/S000...
@swegmann.bsky.social has a cool dataset on opposition powers around the world (2022). I wanted a world map to show it. ~10-30 minutes with Codex and done!
Link to the paper: www.tandfonline.com/doi/full/10....
I'm late to the party (wake? depending on one's perspective) but been playing around with AI: Claude, ChatGPT Codex etc. and I love it for my teaching. Specifically - creating graphs to present stuff from articles in a better(-looking) way. No shade at the authors, just different mediums! examples:
ืืืจืณื ืืืจืฅ, ืืื ืืชืื. ืืืืืื ืืืื ืื ืืฉืื ืืืืื ืฉืืขืืืจ ืื ืืขืืื
ืื ื ืืืื ืื ืืจืืกื. ืื ื ืื ื ืืื ืืขืฉืืช ืืืื ืืคื ื ืฉืื ื ืืคืจืกื ืืืจืื.
I imagine some of the โdrainโ will actually be the YOUNG talent that - under normal circumstances - would have come from EU etc but now will avoid the US.
ืื ืืื ืดืจืฆื ืฆืื ื ืืืคืจื ืืืื ืฉื ืืืื ืืืืกืจโฆ ืืื ืืืืืชื ืืืชืจ, 7/10 ืืกืืื ืคืืืืืด
ืืืโฆ ืืืืจ ืืช ืื ืืืืืื?
ืชืงืืคื ืืฉืจืืืืช ืืคืจืก ืืงืจืืช ืคืืจืื ืืฉื ืืืืฆืข ืื ืืจืคืจืจ ืืืืืืช ืืกืชืจ? ืคืกืคืืก.
ืืคืืืช ืืืืื ืืืฆื ืื ๐คก
ืื ืืื. ืืกืชืื ืื ืืืขืช ืืคืืจืืืช ืคืคืจืืื ืื, ืืขืื ื (ืื ื ืืฉืจืืื ืืืื ืื ืื ืืืืจ ืืชื ืื) ืืืืจืื ืื ืชืืชืื ืืืืื ืืโฆ
ืื ื ืืืื ืืืืชื ืืืื ืืื. ืืื ืืืื ืื ืืืื ืฉืื
Exactly. No conference-finals team *chooses* to suck the year afterwards. And if it would, wouldnโt it tank the semi-finals to avoid this restriction? If anythingโs worse than the current tanking, it would be playoff tanking
I'm sick but not too bad. Which means I'll probably be good for teaching next week. Which means I need to prepare my classes. Which means I need to concentrate. But I can't because I'm (a bit) sick ๐
Wine Cooler is a cool AF spy name!
"Chase a championship"? is he announcing he's getting traded already?
Finally, I know Claude Code is all the rage now. But I'm not (yet) ready to pay both openAI *and* Anthropic, so used Codex instead.
Of course, I take no responsibility for data accuracy or app behaviour. I mean, it won't do any harm I imagine, but links could not work or direct to wrong Wikipedia pages. And table data may be wrong. But I hope it can do what it was meant to - give a quick look into ๐ช๐บ cabinets. At least for me.
It took me 0.5 hour to make the shift in underlying datasets. Then another 3.5 hours to implement a new feature - wikipedia links to the relevant cabinet. 4 hours, and I haven't looked at the code it self, not once!
Here it is: tuttnauer.shinyapps.io/CabinetLookup/
A few years ago, as parlgov.org discontinued its dynamic website, I created my own minimalistic Parlgov Lookup to quickly show which cabinet governed when.
Since parlgov itself is now kinda-discontinued, I wanted to move my web-app to rely on www.repdem.org. How? ChatGPT Codex to the rescue!
Not a US expert, and Iโm very sympathetic to the cause, but US voters had already two chances on electing a women Democrat, and instead elected the most corrupt, inept, old white/orange clown possible. Stuff needs to radically change before AOC is elected president, Iโm afraid.
Thatโs me. I taught myself R a few years back so I can read what my PhD student and assistants do and thereโs data manipulation stuff I do on R myself, but then went back to analysis on Stata cuz it was easier. I guess Iโll move to Claude Code the next time I start a new project?โฆ
Yes. Not Claude code (yet). Same I think for Mannheim, AFAIK they have ChatGPT (through some gateway). But I admit, Iโm not the most up-to-date on these matters
I guess part of the issue is how long it will take for the unis to start paying for Claude rather than Stata. Right now, Iโm using Stata but itโs not me thatโs paying them
ืฉืืข, ืืื ืกืคืง ืฉืื ื *ืื* ืขืืงื ืืืจืื ืืืื ืืจืืฆืืช, ืืื ืืื ืื ืฉืื ืืฉืง ืืจืืฅ ืืขืฆืื, ืืื ืืืจืื ืืืืชืจืืฉืืช ืฉื ืกืืคืืจื ืืจืืฆื ืฉืื ืื ืกืืื ืขืืืืช (ืงืฆืช ืืฆืืืง ืืื ืืืืื ืื ืืฆืื ืืืช) ๐
ืืืฆืืื!