News for hackers. Stuff that matters.

A comment on Making Claude a Chemist. Parent comment

Re:Making Claude a Chemist by chaostheory on Sunday June 14, 2026 @10:05AM (#48527331)
It would make more sense to have specialized versions of Claude instead of having a universal one ie BioClaude, ProgrammerClaude, NuclearClaude, CyberSecurityClaude, et al

This would be both safer and less annoying to use.

[ Parent | on HN ]
Re:Making Claude a Chemist by OutOfHere on Sunday June 14, 2026 @11:40AM (#48528415)
That seems bad because it kills interdisciplinary science.
[ Parent | on HN ]
Re:Making Claude a Chemist by chaostheory on Sunday June 14, 2026 @03:41PM (#48531783)
sure, but how else would you better guarantee safety? How else can we avoid annoying guardrails?

Besides, it's not like you can't have different agents with skills collaborating

[ Parent | on HN ]
Re:Making Claude a Chemist by matheusmoreira on Sunday June 14, 2026 @04:11PM (#48532111)
> how else would you better guarantee safety?

You wouldn't. This safety nonsense is overrated.

[ Parent | on HN ]
Re:Making Claude a Chemist by OutOfHere on Sunday June 14, 2026 @04:46PM (#48532482)
Safety is a lie. It's always possible to divide a task into subtasks, to use abstractions, and the LLM will allow each of these individually.

For example, If you want it to write malware, break that down into a hundred high level steps, then implement each one individually, then stitch them together.

A dumber but uncensored agentic LLM will even do this for you to control the censored smarter LLM.

[ Parent | on HN ]
Re:Making Claude a Chemist by chaostheory on Monday June 15, 2026 @05:31PM (#48547282)
If safety is a lie, then the arguement for massive and overbearing regulation and controls gets a massive boost
[ Parent | on HN ]