Ford telah menghabiskan tiga tahun terakhir secara diam-diam membalikkan salah satu taruhan paling agresifnya terhadap kecerdasan buatan. According to a report from Bloomberg, the automaker has rehired roughly 350 veteran engineers — internally nicknamed "gray beards" — after the AI tools and automated systems meant to replace them failed to meet quality expectations.
The episode is emerging as one of the clearest real-world case studies yet of the gap between the promise of AI-driven automation and the messy reality of deploying it in heavy industry. For more context on this story, see our ongoing AI news.
Apa yang Salah
Ford's leadership had assumed that AI and automated diagnostic systems could absorb much of the work historically done by its most experienced engineers. Asumsi itu terbukti mahal. According to the report, the company's VP of vehicle hardware engineering, Charles Poon, said leaders had overlooked the deep experience of veterans who had survived many product cycles.
Poon was blunt about the outcome, admitting that simply replacing those engineers with AI was a "huge mistake." He was careful to add that AI remains "a fantastic tool," but cautioned that it is "only as good as the information you use to train it" — a recognition that institutional knowledge, the kind accumulated over decades on factory floors and in test labs, does not transfer easily into a training dataset.
The rehired engineers now run mandatory meetings to troubleshoot vehicles and reprogram the automated engineering software and AI tools that were underperforming. Their job is to hunt for failure points before parts ever reach the plant floor — the kind of preventive work that helps head off the massive recalls and defects that had previously cost the company billions.
Perubahan Haluan yang Terukur
Hasil yang membawa manusia kembali ke dunia ini sungguh luar biasa. Di tahun lalu J.D. Power Initial Quality Survey — an annual study measuring problems in a car during its first 90 days of ownership — Ford finished 10th among mainstream brands and scored below the industry average.
Namun tahun ini, J.D. Power menempatkan Ford sebagai merek mainstream teratas, menempatkannya di depan Toyota dan Honda. Ford attributed that dramatic improvement directly to the expertise of the returned engineers, a rare instance of a major corporation publicly crediting human rehiring for a quality rebound.
The turnaround is especially notable given the financial context: Ford has been working to cut roughly $1 billion in expenses this year, yet concluded that spending on veteran talent was the more reliable path to quality than continued reliance on AI.
Ford Tidak Sendirian
Pengalaman Ford cocok dengan pola yang lebih luas dan terdokumentasi dengan baik. A study by outplacement firm Careerminds, cited in coverage of the story, examined companies that conducted AI-driven layoffs and found that 35.6% of them had to rehire more than half of the employees they had previously let go. Another 32.7% rehired between 25% and 50% of those workers — suggesting that a substantial share of AI-driven staff cuts were reversed when the technology failed to deliver.
Kisah peringatan yang paling sering muncul bersamaan dengan kisah Ford adalah Klarna. In 2024, the Swedish fintech's CEO proudly announced that a new AI chatbot was doing the work of 700 full-time customer service agents, prompting the company to freeze hiring and cut hundreds of positions. Namun pada pertengahan tahun 2025 dan memasuki tahun 2026, Klarna berusaha merekrut agen manusia lagi setelah kepuasan pelanggan anjlok. Perusahaan menemukan bahwa meskipun AI menangani pertanyaan dasar dengan baik, masalah kompleks yang memerlukan nuansa dan empati secara konsisten mengalahkannya.
Batasan Otomatisasi
Taken together, these cases point to a recurring lesson that the technology industry has been slow to absorb: AI excels at narrow, well-defined tasks but struggles with the judgment, context, and institutional memory that experienced workers provide.
In Ford's case, that judgment manifests as an almost instinctive sense for where a vehicle design will fail — knowledge built across decades of seeing the same categories of defects recur across product cycles. No amount of training data, however large, easily reproduces that kind of hard-won intuition, particularly when the most valuable lessons are precisely the near-misses and edge cases that never made it into any database.
Semua ini tidak berarti bahwa AI tidak mempunyai tempat di pabrik. Ford sendiri terus menggunakan alat diagnostik otomatis — yang kini diawasi dan disempurnakan oleh para insinyur yang pernah ingin digantikannya. Pembingkaian yang lebih jujur adalah augmentasi, bukan substitusi: AI paling berharga ketika memperluas jangkauan manusia yang terampil, bukan ketika mencoba menghilangkan mereka.
Perhitungan untuk Siklus Hype
Kisah Ford tiba pada saat yang canggung bagi industri AI. Setelah bertahun-tahun janji-janji berani mengenai otomatisasi yang akan menggantikan banyak kategori pekerjaan, semakin banyak bukti yang menunjukkan bahwa perpindahan tersebut jauh lebih selektif – dan pembalikan ini jauh lebih umum terjadi – daripada yang disiratkan pada pernyataan awal.
Bagi para investor, eksekutif, dan pekerja, perubahan yang terjadi pada Ford memberikan sebuah pengingat yang mendasar: kesalahan paling mahal yang dapat dilakukan sebuah organisasi dengan AI mungkin bukanlah penerapannya yang terlalu lambat, namun mempercayakannya untuk menggantikan keahlian yang tidak pernah dapat ditiru oleh AI.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
