Profile avatar
robredotruthe.bsky.social
1,414 posts 124 followers 859 following
Regular Contributor
Active Commenter
comment in response to post
it's crazy how many people went in to nov thinking trump was some kinda dedicated anti-nterventionist peacenik. concerning!
comment in response to post
lol they definitely don't have a nuke. absolutely nobody (even the israelis themselves) are claiming iran have a nuke wth
comment in response to post
lol pmf are already in baghdad, definitely not lookin dehydrated πŸ‘€
comment in response to post
in a 'who gonna overplay their hand?'-off, it definitely looks like trump's way more @ risk of overplaying than newsom is....even trump supporters dnt appear to support sending marines to police kids holding placards
comment in response to post
not sure i follow....the paper indicates that even when presented with the necessary algorithm to solve the problem, LLMs/LRMs with sufficient compute tokens available couldnt use the algorithm; this a known limitation of AR models to not see the global picture therefore invalidating the paper?
comment in response to post
probably a fair point, the problem with your use cases proposal is: 1) most users have zero idea what the model is trained on...making it impossible forthem tounderstand which kinda problems it can solve 2) does sufficient specific training undermine other capabilities?
comment in response to post
i keep seeing this 'LLMs aren't great @ spatial reasoning' point about this paper....im not sure i really understood that as a valid critique; but surely the LLMs being given the required algorithm + then unnable to deploy it kinda undermines this critique?
comment in response to post
i'm curious what u think is going on here? from where i'm sitting it looks like its been trained on some 'human must take the animal' problems, and those are assumed in the model until explicitly ruled out....?
comment in response to post
i'm not sure which part of the paper you think this undermines and how it undermines it?
comment in response to post
yeah and that's fair enough right- blair was a better communicator in the kinda blockbuster 'people's princess' sense...and brown definitely the more robust/less fairydust of the 2... but the idea he was a bad communicator because he wasn't as good as blair is nuts....
comment in response to post
i'm curious if you've ever seen brown communicate? he was probably in the top 3 political communicators of his generation. i'm amazed how history has been rewritten because comms were bad during his time as PM.
comment in response to post
lol when was brown known as less effective communicator? and by who? and compared to who?! i feel like 15 years latern everybodyks got their own personal version of new labour history
comment in response to post
weird bruh
comment in response to post
khodorkovsky wasnt a govt advisor....whas ur point...
comment in response to post
khodorkovsky? this is a full prigozhin meltdown
comment in response to post
Boss move for Ramaphosa would be to offer Musk political asylum.
comment in response to post
cmon what about 'serious' stevey miller. πŸ˜‚πŸ˜‚
comment in response to post
the thing that's impressive about u more than anything else is your willingness to read the drivel these loons churn out, so the rest of us don't have to. major appreciation
comment in response to post
if we keep goin on current trajectories cap, just working with people from all voer the world will get u kicked onto team 'non purebred indigenous white'...and we'll be happy to have you. u probably have a mean right hook as well.
comment in response to post
well so then, in a whites-vs-non-whites scrap (inevitable apparently) the non-whites will be closer to victory. my issue with this is, if u take demographics, all those 70 year old whites in 2040 won't be that useful in a scrap anyway, and will provide ample opportunities for hostage-led bargaining
comment in response to post
i'm curious what you mean by capitalizing on the success of the spider's web op in washington? could you describe your thinking here?
comment in response to post
The courts won’t let the RN sink boats full of civilians. Imagine that.
comment in response to post
This is like the Michael Gove article about Brexit where he complained that every time he wanted to do something the civil service said they'd love to but the EU wouldn't let them, and he concluded leaving the EU would mean civil servants would do everything he wanted.
comment in response to post
sure, and some completely sober people are too much oif a driving hazard to safely drive around....that doesnt make drink driving safe though. never thought i'd have to point that out πŸ˜‚
comment in response to post
lol its fair to say 'if heroin addicts had a cheap steady supply they wouldnt die from a dodgy batch, or rob cars to pay for it' ....not sure how we've gone from that to '...and therefore heroin has no impact on a dude's capacity to function'
comment in response to post
lol not sure its stigmatizing drugs to say: 'i'd rather not have a ketamine-fiend organizing my company accounts' or 'can we find someone who isn't an alcoholic to manage my daughter's school's security guards' weird that i'd have to point this out, but here we are πŸ˜‚
comment in response to post
surely the most unexpected part of the phenomenon has to be that whilst abandoning badly translated aramaic/greek/latin names, society hasn't collapsed under the abandonment of 'judeo-christian values' being replaced with uncodified pagan moral codes
comment in response to post
'academia's current problems are almost identical to those of past academia, and every other industry and time period, and its productive to acknowledge that in isolation'
comment in response to post
any idea why the russians keep demanding sanctions relief given they don't seem to be having an impact?
comment in response to post
the real issue bluesky has is that it really only appeals to people who are/want to read academics' hot takes...thas a relatively niche target audience...
comment in response to post
2 things i dont get: 1) how is 'likers' a proxy for DAU? 2) there's no way people are using twitter over bluesky because they got bullied off bluesky. that seems totally delusional.
comment in response to post
erm guys, THIS ISN'T NEWS.
comment in response to post
dan hodges' expert economic analysis available now, from the IFS.
comment in response to post
lol seems unlikely when he's basically halved immigration in a year. the real issue here is that, as proved similar with brexit; the british electorate appear to just have no grasp of immigration on anything beyond a vague half-emotive level