Fish&Almond📯
@Zirconos
Followers
135
Following
65K
Media
647
Statuses
8K
Chemiker | sozialliberal | as a society we have to give every individual the opportunity to live up to its potential.
Joined June 2015
@smynt1967 @AlexSchaumburg Ich habe hier mal einiges „aus dem Zusammenhang gerissen“. Er hat wirklich viel Pech bei der Formulierung wichtiger Fragen.
***Thread*** Im Gegensatz zu vielen seiner Anhänger, habe ich Höckes Buch gelesen und insbesondere den politischen Teil einer Prüfung unterzogen. Dieser Teil beginnt auf S.257 seines Buches und ist mit „Krise und Renovation“ überschrieben. Ich zitiere hier Stellen aus seinem Buch
3
3
14
Wen ihr wählt, wenn ihr AfD wählt: Nur ein Beispiel.
168
153
557
Russians joke: "All that we found out today is that if, God forbid, NATO attacks us, the most we'll be able to do is to dig out the asphalt and call Lukashenko"
329
4K
23K
a rainbow shirt in target won’t turn your kids gay but Brad Pitt as Achilles will
a rainbow shirt in target won’t turn your kids gay but gillian anderson as dana scully will
45
632
8K
Wohin #Kulturkampf und „Kampf gegen #Woke“ führt: Beim #CSD #Hannover wurden zwei transidente Jugendliche brutal angegriffen und am Boden liegend auf ihre Köpfe eingetreten. Nein, jetzt bitte keine „thoughts and prayers“ 🤬. #LGBTQIA #Gewalt
https://t.co/I68wrjlPmm
@LSU_Bund
haz.de
Zwei junge Teilnehmende des Christopher-Street-Day wurden in Hannover Opfer einer queerfeindlichen Attacke – einer wurde schwer verletzt. Die Polizei hat weitere Übergriffe aufgenommen. Im Interview...
11
13
37
That's a really good point and gave me some ideas about the way those LLMs can used. A major strength of those tools is that you can kind of have a dialogue with them. So maybe they should rather ask you for context, casuality and logical connections than just doing that for you?
Why write an essay when you can type a few words and have AI generate one for you? Why write an email when AI can auto-respond for you with all the typical pleasantries and talking points? While AI doing these things for you is likely to happen, it’s not necessarily a good
1
1
0
Because, if it would detect the problem, it would avoid it, right? To circumvent that the model has to ask questions about any assumption/connection it's making, which is again pushing a lot of interaction but also avoids the issue that its taking the learning experience.
0
0
0
Especially as I found this to be a major issue. The way they phrase their sentences makes a very factual and "convinced" impression, even though they're completely false sometimes. To ask the "right" questions, the Model would have to detect it's own errors, which is a problem.
1
0
0
That's a really good point and gave me some ideas about the way those LLMs can used. A major strength of those tools is that you can kind of have a dialogue with them. So maybe they should rather ask you for context, casuality and logical connections than just doing that for you?
Why write an essay when you can type a few words and have AI generate one for you? Why write an email when AI can auto-respond for you with all the typical pleasantries and talking points? While AI doing these things for you is likely to happen, it’s not necessarily a good
1
1
0
Why write an essay when you can type a few words and have AI generate one for you? Why write an email when AI can auto-respond for you with all the typical pleasantries and talking points? While AI doing these things for you is likely to happen, it’s not necessarily a good
52
256
1K
Love is sharing a password.
5K
77K
529K
@schlotterbog Mats shirtless and with pink hair would've made you click as well, I'm sure about that 🦦
0
0
0
Based on the replies to this tweet, I've compiled a nice 32 bracket single elimination bracket of all the shit we've been served since the purchase of Twitter. Each day I'll make a new poll with citations and let the public decide. The seeding was random, the results? Who knows!?
what is *the* dumbest thing elon has done at twitter so far?
85
1K
6K