Elon Musk is allegedly planning to utilize artificial intelligence to manage the US government, a move that experts caution against as a “very bad idea.”
Musk has reportedly terminated numerous federal government employees through his Department of Government Efficiency (DOGE) and requires remaining workers to submit a weekly email outlining their accomplishments in five bullet points.
To handle the influx of emails at DOGE, Musk is turning to AI for assistance in determining who should stay employed and is considering replacing government workers with AI systems.
Experts express concerns about the lack of transparency regarding the design and validation of these AI systems, warning of potential mistakes, biases, and other issues that could arise from utilizing AI in such a critical capacity.
Despite these concerns, the current government, led by Musk, is moving forward with plans to use AI for various tasks, such as scanning social media accounts to identify potential threats.
Undetected harms
Experts emphasize the need for cautious and conscientious adoption of AI in government operations, citing past instances of AI errors leading to significant negative impacts on individuals.
From wrongly identifying cases of fraud to making biased decisions in law enforcement, the use of AI in government systems has raised serious concerns about its potential for harm.
AI does not understand anything
Replacing human workers with AI presents a challenge due to the diverse skill sets and knowledge required for various government roles, making it difficult for AI to effectively replicate human expertise.
Experts caution against indiscriminate job cuts without proper training and development of AI tools, emphasizing the limitations of AI in understanding complex issues and historical biases.
While a previous executive order focused on responsible AI use in government was issued under President Biden, it was rescinded by the Trump administration, raising concerns about the future of AI governance in the US.
Schellmann mengatakan hal ini membuat AI lebih tidak mungkin digunakan dengan bertanggung jawab di pemerintahan atau bahwa para peneliti akan dapat memahami bagaimana AI digunakan.
Semua ini mengatakan, jika AI dikembangkan dengan bertanggung jawab, ini bisa sangat membantu. AI bisa mengotomatisasi tugas-tugas yang berulang sehingga pekerja bisa fokus pada hal-hal yang lebih penting atau membantu pekerja menyelesaikan masalah yang mereka hadapi. Tapi memang perlu waktu untuk diberikan agar bisa diimplementasikan dengan benar.
“Bukan berarti kita tidak bisa menggunakan alat AI dengan bijaksana,” kata Coglianese. “Tapi pemerintah salah langkah ketika mereka mencoba terburu-buru dan melakukan hal-hal dengan cepat tanpa masukan publik yang tepat dan validasi serta verifikasi yang menyeluruh tentang bagaimana algoritma sebenarnya bekerja.”