patchcritic (@patchcritics) header
patchcritic (@patchcritics) profile picture

Making robots move nicely • Bundeswehr Reserve’s favorite art critic

🇸🇬
Joined October 2022
Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)Photo by patchcritic (@patchcritics)
patchcritic retweeted
Dass Politiker überhaupt ihre Antworten ändern und sogar versuchen, die Fragen umzuschreiben, kannst Du außerhalb Deutschlands Keinem erklären.
15
32
1
258
6,756
patchcritic retweeted
The increasingly farcical saga of Berlin's attempt to put together a leftist government is v dispiriting, but for an Anglo journalist it is at least modestly heartening to see Germany's wretched interview-authorisation process at the heart of the shitshow. spiegel.de/politik/deutschla…
9
46
2
351
16,933
No one in the history of media knew the interior workings of a young boy’s brain better than Bill Watterson
117
1,327
81
16,611
383,302
The sleepiest trade union confederation in the interwar period
68
373
52
6,174
126,252
Salvador Dalí poster for US Army in campaign against venereal diseases, 1942.
8
99
10
1,115
169,216
23
3,075
27
47,670
386,828
He also said -GPT3 is the world's butterfly. -ChatGPT is an alien being -radiologists would be obsolete by 2023 (said in 2018). Bonus: coauthored paper with longtermists who believe our moral obligation is to create digital beings living happily ever after colonizing cosmos.
10
106
5
503
19,263
AI is conscious people be like “How can you know with certainty that the stripper isn’t in love with me? From a functional perspective, how would her behavior differ if she were in love with me?”
111
957
88
10,634
300,663
patchcritic retweeted
A map showing East Germany
128
88
43
3,408
279,088
>on my first date with my mathematically optimised state-assigned breeding partner from the singapore government’s new ‘you are completely useless looks like we are going have to help you have sex’ app >really nervous, sweating profusely and struggling to make eye contact >ghost of lee kuan yew appears >“come on you herbivore make a damn move” he says >pull out my spaghetti and drop it >“she is literally mathematically optimised for you please it’s this or we start artificially inseminating people to raise birth rates” he says >mumble some weird comment about her smell in a flat zoomer monotone voice >“put your fucking penis inside her for god’s sake” he says
I CALLED IT. It's happening y'all. The singapore gov has launched a pilot programme for a gov-run dating service. Looks like the pilot is currently restricted to just government workers for now.
111
901
63
15,479
1,022,953
patchcritic retweeted
So, bin auf dem Weg zurück aus dem Urlaub - ab morgen dann auch hier wieder voll m Einsatz. Ich übe ja quasi das Twittern noch :-)
5
20
3
117
patchcritic retweeted
Ich glaube nicht, dass alle Linken bzw. jeder 4 Berliner Antisemit ist. Gerade im Bezug auf die Wähler dürfte das Thema schlicht egal gewesen sein. Was nicht egal war: das Thema Mieten, welches die Linken gut bespielt haben. Diese Analyse zu verkennen wir uns Wahlen kosten.
18
6
1
78
2,735
US: painstakingly describe LLMs in section 2, history, task, input, output, setup, what "size" means. US: talk about different types of tech like ASR, machine translation, others, what to do in cases of dual use etc. Them: STOCHASTIC PARROTS SAYS AI DOESN'T WORK.
Replying to @provisionalidea
He's totally right about this? People wanted to believe that AI didn't work, and so they were receptive to this argument and blind to the evidence against it. I include myself in that category (up until '23).
9
20
137
9,116
patchcritic retweeted
"That's Marketing!"
We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation. Sharing preliminary findings to help defenders understand emerging risks: openai.com/index/hugging-fac…
1
2
13
965
This is the kind of movie that pretentious douchebags like to recommend to people But then you watch it and it's actually amazing
Seven Samurai (1954) | Akira Kurosawa
598
3,587
192
75,828
2,305,725
2019: “GPT2 is groundbreaking in two ways. One is its size, says Dario Amodei, OpenAI’s research director. The models “were 12 times bigger, and the dataset was 15 times bigger and much broader” than the previous state-of-the-art AI model” theguardian.com/technology/2…
36
391
104
2,473
422,477
Imagine you’re a conman. You’ve said for years you have MAGIC BEANS. You’ve shown lots of REALLY BIG normal beans that you made in your lab. “Almost there!” you say—over and over. Now, you have *almost* convinced the villagers to give you TWO TRILLION DOLLARS. But while the beans keep getting bigger. No magic yet… What do you do? That money is SO CLOSE. “I know!” you think, “instead of MAGIC beans, lets’s make them KILLER BEANS.” So you pump up your beans up SUPER BIG and you say: “I’m sorry, but these BEANS ARE GETTING BIGGER ON THEIR OWN!” You threaten people that if they don’t give you the money, the beans may get into the WRONG HANDS. They might get BIGGER THAN THE EARTH! The villagers, aghast, run to and fro arguing whether the beans are going to kill everyone or make them all rich. “Well, either way,” the idiots in the village conclude, “We’d better give him the money.”
228
1,942
158
10,105
403,077
More realistic AI doom scenario: You wake up, and can't log in to any of your devices or accounts. "Hmm, that's weird..." The Internet doesn't work. For anyone. Nothing loads. No passwords work. This does not improve, for several days. Then weeks. What happens next?
1,049
237
159
5,397
894,424
patchcritic retweeted
Reviewer 1: “Accept with minor revisions” Reviewer 2:
14
317
20
3,350
77,071