/Tech4h ago

Researcher Defends Anthropic's AGI Safety Practices as Necessary Precautions

21101526
Ted Sanders#834
Original post
Ted Sanders@sandersted#834inTech

@livgorton Yeah, I think Anthropic is pretty much fine here. If you are serious about AGI and its risks, it makes perfect sense to store requests for safety monitoring, block potential misuse, try to prevent proliferation of AGI code, and hide some of your techniques from adversaries.

Liv@livgorton

People being angry with AI companies is fine and a tool for accountability. Companies (and those speaking for them) leveraging that anger to dunk on each other for marketing is a different, distasteful, and corrosive thing. Let your good work speak for itself.

5:13 PM · Jun 10, 2026 · 381 Views
Sentiment

Users expressed support for Anthropic and OpenAI because success for both companies is viewed as beneficial for consumers.

Pos
100.0%
Neg
0.0%
1 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS145LIKES4REPLIES1
Ted Sanders@sandersted

@livgorton The most annoying bit is the silent degradation, but if people don't like it, they can always just not use it.

Sadly, I think success draws negativity. Google and OpenAI started off as darlings, but once they were perceived as rich / powerful goliaths, it got much more negative.

Ted Sanders@sandersted

@livgorton Yeah, I think Anthropic is pretty much fine here. If you are serious about AGI and its risks, it makes perfect sense to store requests for safety monitoring, block potential misuse, try to prevent proliferation of AGI code, and hide some of your techniques from adversaries.

4hViews 145Likes 4Bookmarks 0
Ted Sanders@sandersted

@livgorton Personally, I root for both Anthropic and OpenAI. The better we all do, the better it is for consumers. (For now, at least.) And I perceive Anthropic and OpenAI to be far more similar than different (as opposed to Google, X, Meta, DeepSeek, Mistral, Thinking Machines, etc.).

4hViews 24Likes 1