KatzOnEarth The most shocking thing, to me, about the Sanders/Warren story is that Donald Trump laundered money from the Islamic Revolutionary Guard Corps of Iran then ordered an illegal hit on its leader
Chekhov's confusables
JayCogi @perseveresther Marius von Raum für Notizen, ship's medic
JayCogi @perseveresther this relationship intentionally left blank
VQuaschning Geniale Idee: Hochgeschwindigkeits-#Nachtzug als Alternative für den EU-#Flugverkehr. Abends in Frankfurt einsteigen, morgens in Barcelona aufwachen. Die Trassen sind weitgehend schon da. Es fehlen nur die Züge. Wie wär's @AndiScheuer?
#FridaysForFuture
zeit.de/mobilitaet/202…
Luisamneubauer Wir klagen: Gemeinsam mit einigen anderen jungen Menschen werde ich vor dem Bundesverfassungsgericht eine #Klimaklage einreichen.
Morgen um 10 Uhr erklären wir uns zusammen mit @greenpeace_de, @Germanwatch & der @Umwelthilfe bei der Bundespressekonferenz. Bis morgen!✊
emilymbender IOW, you can't say/do harmful things, then try to turn it into a "political issue" and claim immunity on those grounds.
emilymbender But refusing to follow those policies and then getting reprimanded/disciplined/fired for it doesn't mean you're getting fired for your political views.
emilymbender The university's job is to provide educational opportunities for everyone. In light of existing systems of oppression, this can me instituting policies that seem strange/frivolous/irrelevant to people whose lived experience of privilege makes it hard to relate to their value. >>
@Dunkelmagier I feel ya twitter.com/texttheater/st…
gneubig Happy to release NN4NLP-concepts! github.com/neulab/nn4nlp-…
It's a typology of important concepts that you should know to implement SOTA NLP models using neural nets: github.com/neulab/nn4nlp-… 1/3
We'll reference this in CMU CS11-747 this year, trying to maximize coverage. 1/3 twitter.com/gneubig/status… pic.twitter.com/ILFeobZmPM
literalbanana denial-of-context attack
WatchedPotts It's "email to prospective dissertation/thesis supervisor" season! My inbox runneth over. Are you writing messages to academics, asking for advice or endorsement on your proposals? Maybe this thread will help.
jasoninmia The best sign of a good class is how students react to not you, the teacher, but each other. Are they talking when you walk in the door? If so, they’re probably glad to be there. I used to think their chatter was a sign of distraction. Now I see it as a readiness for engagement.
e_salvaggio In case anyone thinks I know what I’m doing, here’s how I tell Photoshop to batch resize 1400 images when I don’t want to manually confirm the JPEG quality options for 1400 images pic.twitter.com/IRzPcw2R21
TBSkyen Anything to muddle and obscure the fact of pollution and climate change that it is INDUSTRY that is most responsible for it, and industry which has the most real power to fix it.
TBSkyen Big Think is funded by the Koch foundation - and tbh literally any time you see articles implicitly or explicitly blaming consumers and ordinary people for climate change, it's safe to assume it is fossil-funded propaganda. pic.twitter.com/iG7pGxI7Wo
"If we change the tokenization after annotation has started, the web interface gets really angry." #OH
sciforfuture "#Klimahysterie" ist das Unwort des Jahres 2019 unwortdesjahres.net/fileadmin/unwo…
publictorsten Bühnennebel auf der Straße. twitter.com/polizei_nrw_k/…
@prooffreader @fmailhot You stole it from me
twitter.com/texttheater/st…
evanmiltenburg The problem is that quick-and-dirty methods are reaching a plateau. At this point, the limitations are fairly clear.
Since we know (roughly) where the data can take us, why bother doing more of the same?
evanmiltenburg At the same time, I guess it’s reasonable to want to see how far the data can actually take us.
But the incentives here are all against slow science, and actually putting in the work. So it’s rare to see these kinds of papers.
evanmiltenburg I’m getting tired of these excuses in #NLProc papers, saying that X is expensive/difficult/labor-intensive, and thus we should opt for some quick-and-dirty solution.
We’d probably make a lot more progress by actually putting in the work. This is just procrastinating.