Hebrew-language scholars debate whether AI-accelerated science will make human researchers obsolete

On September 3, Hebrew-speaking users including ShalitUri and Rebbetzin engaged in a debate on X about AI's impact on research and universities. One side argued that "for the science of machines, we are too dumb and too slow": human papers take 3–5 years from research, writing, and peer review to publication, while AI systems will vastly accelerate this process and ultimately make human academic roles redundant; even fields where AI currently underperforms will catch up quickly.

Confirmed

• This is a discussion of views built from multiple quote-tweets and replies on the same topic, with participants including ShalitUri and Rebbetzin; all views are personal speculation with no factual events involved
• Rebbetzin's summarized extreme picture: universities and science are heading toward a "civilization" that may contain no humans at all—humans won't be invited to academic conferences among AIs, nor able to publish in their journals
• A cited view holds that we are entering an extreme knowledge crisis, with vast amounts of "paradoxical knowledge" (unknown knowledge) emerging outside human hands, potentially accelerating the university's internal collapse
• ShalitUri added: the fastest acceleration will come in highly instrumental fields with few physical constraints, such as mathematical research

Controversy and disagreement

• ShalitUri proposed distinguishing two types of knowledge: "instrumental knowledge" (e.g., a cure for cancer) and "knowledge aimed at human understanding" (the bulk of the humanities and pure mathematics). He argued machines can greatly accelerate the former, but the latter's value lies in human understanding itself—so even if machines produce knowledge, it still must pass through the step of human understanding
• This distinction is effectively a correction to the pessimistic claim that "AI will make human researchers redundant"; the two sides disagree on "whether humans will be completely excluded from science"

Why it matters

• This debate reflects two typical academic reactions to AI-accelerated research: a "obsolescence thesis" based on the speed disadvantage of human academic workflows, and a "irreplaceability of human understanding thesis" appealing to the value of knowledge itself
• Concepts raised in the discussion—unknown knowledge, self-contained academic systems among AIs, universities collapsing from within—are representative formulations for understanding academic institutions' anxieties in the AI era

2026-09-03 ~ 2026-09-03 · 5 related posts

• Peer Review Takes 3-5 Years: Why Human Academia Can't Keep Up With Machine Science — _Rebbetzin_ · 2026-09-03
• [source] The Future of University and Science: A Civilization That May Not Include Humans — _Rebbetzin_ · 2026-09-03
• [source] Hebrew X thread: AI will accelerate science and make human researchers obsolete — ShalitUri · 2026-09-03
• [source] Distinguishing Instrumental vs Understanding Knowledge: Why AI-Accelerated Science Still Hits Physical Bottlenecks — ShalitUri · 2026-09-03
• Researchers debate AI-generated 'unknown knowledge' humans can't verify — ShalitUri · 2026-09-03