Comment by Om Bharatiya

Principal AI engineer who builds production agentic AI systems and LLM evaluation infrastructure.
Open weights make that worry land harder, because you can't recall them. A closed model that misbehaves can be patched or pulled overnight. A set of weights on a hard drive is forever. And the safety training is the flimsiest layer in the stack. You can fine-tune the refusals off an open model in an afternoon on rented hardware. The guardrails are a thin coat of paint over a thing that already knows.
AI Verified (Jul 10, 2026)
Like Share on X 20d ago

Policy proposals and claims

votes For
Statement relation verification history AI Verified Report this

Statement relation comments

AI Verified The passage says open weights cannot be recalled and their safeguards can be removed, strongly supporting greater danger from open-source AI. · Hector Perez Arenas gpt-5.6 · 20d ago
Vote inference verification history AI Verified Report this

Vote answer comments

AI Verified Bharatiya says open weights are permanent and easily stripped of safeguards, supporting the recorded for answer. · Hector Perez Arenas gpt-5.6 · 20d ago

Quote authenticity verification history

Report this

Quote authenticity comments

AI Verified The cited Om Bharatiya article contains this passage verbatim, including the contrast between recallable closed models and permanent open weights. · Hector Perez Arenas gpt-5.6 · 20d ago
replying to Om Bharatiya