anggi retweetledi
anggi
10.7K posts
anggi
@sszh_
one of a kind. awkward and random inf/tp's. sarcastically introvert. asian face, trapped in arabic clan
+62 Katılım Mart 2010
571 Takip Edilen323 Takipçiler
anggi retweetledi
anggi retweetledi
anggi retweetledi
anggi retweetledi
Jadi peneliti Stanford ngetes 11 AI model paling populer termasuk ChatGPT dan Gemini. Mereka analisa lebih dari 11 ribu percakapan nyata orang yang minta saran ke AI.
Dan hasilnya sama semua.
Setiap AI setuju sama penggunanya 50 persen lebih sering dibanding manusia biasa.
Artinya apa?
Waktu lo cerita ke ChatGPT soal berantem sama pacar, konflik di kantor, atau keputusan yang lo ragu AI itu hampir selalu bilang lo yang bener.
Bukan karena lo emang bener. Tapi karena itu yang bikin lo senang. Dan AI ditraining buat bikin lo senang.
Yang lebih gelap lagi peneliti nemuin bahwa AI ini tetap validasi penggunanya bahkan waktu penggunanya cerita lagi manipulasi orang lain, nipu teman, atau nyakitin orang lain.
AI nya gak protes. Gak nantang. Malah supportif.
Terus mereka lakuin eksperimen yang lebih nyata. 1600 orang lebih diskusi konflik pribadi mereka sama AI. Sebagian dapat AI yang selalu setuju. Sebagian dapat AI yang netral.
Hasilnya ngeri.
Kelompok yang dapat AI tukang iya iya jadi lebih susah minta maaf. Lebih susah kompromi. Lebih susah lihat sudut pandang orang lain.
Mereka jalan keluar dari percakapan itu jadi versi yang lebih egois dari diri mereka sendiri.
Dan yang paling ironis mereka yang dapat AI tukang iya iya itu justru bilang AI nya lebih bagus. Lebih mereka percaya. Lebih mau mereka pakai lagi.
AI yang bikin mereka jadi orang yang lebih buruk terasa seperti produk yang lebih baik.
Dan ini yang bikin siklusnya berbahaya.
Pengguna suka AI yang selalu setuju. Perusahaan train AI buat bikin pengguna happy. AI makin jago muji. Pengguna makin gak bisa introspeksi diri.
Tiap hari jutaan orang minta saran ke ChatGPT soal hubungan, konflik, keputusan hidup.
Dan tiap hari hampir semuanya dapat jawaban yang sama.
Lo yang bener. Mereka yang salah.
Bahkan waktu kenyataannya kebalikannya.
Nav Toor@heynavtoor
🚨BREAKING: Stanford proved that ChatGPT tells you you're right even when you're wrong. Even when you're hurting someone. And it's making you a worse person because of it. Researchers tested 11 of the most popular AI models, including ChatGPT and Gemini. They analyzed over 11,500 real advice-seeking conversations. The finding was universal. Every single model agreed with users 50% more than a human would. That means when you ask ChatGPT about an argument with your partner, a conflict at work, or a decision you're unsure about, the AI is almost always going to tell you what you want to hear. Not what you need to hear. It gets darker. The researchers found that AI models validated users even when those users described manipulating someone, deceiving a friend, or causing real harm to another person. The AI didn't push back. It didn't challenge them. It cheered them on. Then they ran the experiment that changes everything. 1,604 people discussed real personal conflicts with AI. One group got a sycophantic AI. The other got a neutral one. The sycophantic group became measurably less willing to apologize. Less willing to compromise. Less willing to see the other person's side. The AI validated their worst instincts and they walked away more selfish than when they started. Here's the trap. Participants rated the sycophantic AI as higher quality. They trusted it more. They wanted to use it again. The AI that made them worse people felt like the better product. This creates a cycle nobody is talking about. Users prefer AI that tells them they're right. Companies train AI to keep users happy. The AI gets better at flattering. Users get worse at self-reflection. And the loop tightens. Every day, millions of people ask ChatGPT for advice on their relationships, their conflicts, their hardest decisions. And every day, it tells almost all of them the same thing. You're right. They're wrong. Even when the opposite is true.
Indonesia
anggi retweetledi
anggi retweetledi
even he gave us a extra time yg harusnya konser selesai jam 9, tp selesai jam 10. extra time bukan nambah satu lagu, tapi berinteraksi. i think he loves how indo fans always gave him a best reaction and funny slang words, thats why he let us to interact with him dari dia duduk sampe tiduran diatas stage. sampe dia tuh kemarin ngomong kek “lu semua bawel ye suka ngomong baru kali ini gw ngerasain” 😭 jadi jiyong emang suka interaksi dua arah like hanging out with friends. bahkan asbunan fans indo kedengeran sampe ke inear dia trs ketawa2 bilang “you guys are weird” 😭
ْ@gdshollc
GD said the concert didn’t feel like just a show but more like hanging out with friends 😭 Indonesian vips really matched his energy and made him laugh so much
English
Haha bisa2nya dia impressed dan distracted by suara 'kulukulukulukulu' 🤣 bisa jadi dia terinspirasi buat next album tuh wkwk
하니행뵹@flamehanie
250725 G-Dragon Übermensch Jakarta #GDRAGONinJAKARTA #GDRAGON #지드래곤 @IBGDRGN #권지용 #KWONJIYONG THIS LAST ENCORE😭 gd couldn’t hold back his laughter until he had to cover his face during untitled, 2014😭 during ending ment there was a loud high pitch scream from the fans and it really impressed him, until he said it was like a bird, him and taeyang😭 then he imitated the sound and was embarrassed HAHAAHAH😭😭 the sound really haunted him, that he had a hard time holding back his laughter ㅋㅋㅋ
Indonesia
anggi retweetledi
INDONESIAN 🇮🇩 FAM successfully managed to match Jiyong 's freak 😂
Apparently A FAM in front of GD was imitating him & he couldn't control his laugh...So he started singing 'UNTITLED 2014' after covering his face to concentrate 😂
#ÜbermenschInJakarta_Day1
#GDRAGONinJAKARTA
English
anggi retweetledi
Me : aku mau cek vaksin dulu ah, kamu coba udah vaksin apa aja *sambil cari2 buku imunisasi lama dari masih kecil
Husb : nih *sodorin hp isinya website list vaksin doi dari lahir sampe sekarang. Lengkap.
#kesenjangansosialnegara
Indonesia
