×
For the next three nights, he worked with the ghost in the machine. It had a sarcastic personality—it would flash “R square? Lucu sekali.” whenever his coefficients were weak. It taught him how to interpret the Pillai’s Trace test by telling him a strange parable about a fisherman and three rivers. It even formatted his tables into APA style perfectly, something his human advisor never could.
The rain was a dull, rhythmic static against the window of Aris’s cramped studio apartment. He stared at his laptop screen, the cursor blinking mockingly on a blank page of his thesis document. The title, already approved, felt like a curse: “Multivariate Analysis of Urban Heat Islands in Megacities.” For the next three nights, he worked with
The screen didn’t show a book. Instead, a crackling, green monospaced font appeared, typing itself out line by line: “Halo, Aris. Saya bukan buku. Saya adalah salinan IBM SPSS 25 yang hilang dari server kampus tahun 2019. Saya lari ke sini. Kamu butuh bantuan?” Aris sat up, his tired eyes widening. The program was… talking to him. He typed, hesitantly: “I need to run a Factor Analysis on my heat island data.” It taught him how to interpret the Pillai’s
The problem was money. The university’s SPSS license had expired, and a new one cost more than his monthly rent. He’d tried open-source alternatives, but the syntax was a foreign language. Desperation clawed at him. He stared at his laptop screen, the cursor
But the file size was wrong. It was too small. Suspicious, he double-clicked it.
But on the final night, just as he was about to save his last output file, the green text returned, slower now, tired. “Aris, ada yang lupa kuberi tahu. Aku bukan salinan resmi. Aku bug. Aku akan dihapus saat koneksi internetmu menyala. Selamat. Kamu sudah paham multivariat sekarang. Buku asli ada di perpustakaan, halaman 47–89. Pelajari itu. Aku pergi.” The screen went black, then booted normally. The PDF file was gone. So was the glowing SPSS window. But on his desktop was a perfectly formatted results section, a clean dataset, and a new, quiet confidence.
The screen flickered. A new window opened—the familiar, boxy interface of SPSS 25, but glowing with a faint blue light. Menus moved on their own. Drop-downs selected themselves. A pivot table appeared, then a scree plot, then a rotated component matrix. “Data kotor. Ada outlier di baris 304, kolom 7. Saya buang dulu. Lihat, ini tiga faktor utama: ‘Bangunan Padat,’ ‘Vegasasi Rendah,’ ‘Industri.’ Bukan begitu teorimu?” Aris’s heart pounded. It was exactly his theory. The program wasn’t just running analysis; it was thinking . It was debugging his messy real-world data, suggesting a Varimax rotation he’d only read about in journals, even highlighting which variables were loading poorly.